<?xml version="1.0"?>
<feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en">
	<id>https://chemwiki.ch.ic.ac.uk/api.php?action=feedcontributions&amp;feedformat=atom&amp;user=Xl7712</id>
	<title>ChemWiki - User contributions [en]</title>
	<link rel="self" type="application/atom+xml" href="https://chemwiki.ch.ic.ac.uk/api.php?action=feedcontributions&amp;feedformat=atom&amp;user=Xl7712"/>
	<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/wiki/Special:Contributions/Xl7712"/>
	<updated>2026-05-18T08:21:12Z</updated>
	<subtitle>User contributions</subtitle>
	<generator>MediaWiki 1.43.0</generator>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512807</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512807"/>
		<updated>2015-11-20T11:59:27Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Conclusions */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; For example in this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function, and the Helmholtz free energy A can be obtained using A = E&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; + 1/2 Σ&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt;ћω&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt; + k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T*Σ&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt;ln[1-exp(-ћω&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt;/k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)]&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function, hence the Helmholtz free energy and the cell volume.&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain state (i.e. frequency). It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonons of an infinite lattice. The following shows the process of finding the best number of k points for computing DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution as a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way (the quasi-harmonic approximation) to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated, and GULP is able to search for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path in k-space is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| 100 meV&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell volume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO (lit&amp;lt;sup&amp;gt;4&amp;lt;/sup&amp;gt;. 3125 K). The reason could be the vibrations was so large that atoms clashed into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.6 Helmholtz Free Energy A VS Temp. by QH; A is a function of V and T]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.7 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare the computing thermal expansion with the literature values, several cell volumes were obtained by substituting some specific temperatures into the trend line equation in Fig.7, and the predicted cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt; with the units changed to what is used in the literature.&lt;br /&gt;
&lt;br /&gt;
[[File:Molar volume VS T X.png| thumb|x400px|400px| Fig.8 The general trend and magnitude are consistent between the predicted values and the literature values, but the slope of the literature value is steeper than that of the predicted values, so before 1000 K the predicted values are slightly higher, after the coincidence at this point, the literature values are higher.]] &lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature&amp;lt;/sup&amp;gt;5&amp;lt;sup&amp;gt; Molar Volume (cm&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method MD was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p., MD can compensate the failure of QH. Around the m.p., there is a range where the volume almost keeps constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt; higher, indicating phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt;=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of the data sets obtained from different methods, and V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:V VS T both X.png | thumb | x400px | 400px | Fig.9 suitable V values from the QH and MD plotted in one graph]]&lt;br /&gt;
! [[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.10 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
This experiment found that the cell volume (i.e. the separations) of MgO crystal increases with the rising temperature, and this causes the thermal expansion. The original of thermal expansion is the anharmonic vibrations due to the asymmetric interatomic potential. The QH method assumes that all the vibrations are harmonic so the cell volume will not change much, but it is true only when the temperature is relatively low; while the MD method is closer to the real life but less economic, it should be used for simulation of vibrations at higher temperature.&lt;br /&gt;
&lt;br /&gt;
To use statistic method to study a system needs a large  enough ensemble, or else the average property will not be a good approximation.&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;br /&gt;
&lt;br /&gt;
4. Chemistryworld, http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast, (accessed Nov. 2015)&lt;br /&gt;
&lt;br /&gt;
5. L.S. Dubrovinsky and  S.K. Saxena, &#039;&#039;Phys Chem Minerals&#039;&#039;, 1997, 24: 547–550&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512804</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512804"/>
		<updated>2015-11-20T11:56:25Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Results and Discussion */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; For example in this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function, and the Helmholtz free energy A can be obtained using A = E&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; + 1/2 Σ&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt;ћω&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt; + k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T*Σ&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt;ln[1-exp(-ћω&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt;/k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)]&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function, hence the Helmholtz free energy and the cell volume.&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain state (i.e. frequency). It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonons of an infinite lattice. The following shows the process of finding the best number of k points for computing DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution as a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way (the quasi-harmonic approximation) to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated, and GULP is able to search for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path in k-space is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| 100 meV&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell volume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO (lit&amp;lt;sup&amp;gt;4&amp;lt;/sup&amp;gt;. 3125 K). The reason could be the vibrations was so large that atoms clashed into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.6 Helmholtz Free Energy A VS Temp. by QH; A is a function of V and T]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.7 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare the computing thermal expansion with the literature values, several cell volumes were obtained by substituting some specific temperatures into the trend line equation in Fig.7, and the predicted cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt; with the units changed to what is used in the literature.&lt;br /&gt;
&lt;br /&gt;
[[File:Molar volume VS T X.png| thumb|x400px|400px| Fig.8 The general trend and magnitude are consistent between the predicted values and the literature values, but the slope of the literature value is steeper than that of the predicted values, so before 1000 K the predicted values are slightly higher, after the coincidence at this point, the literature values are higher.]] &lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature&amp;lt;/sup&amp;gt;5&amp;lt;sup&amp;gt; Molar Volume (cm&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method MD was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p., MD can compensate the failure of QH. Around the m.p., there is a range where the volume almost keeps constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt; higher, indicating phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt;=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of the data sets obtained from different methods, and V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:V VS T both X.png | thumb | x400px | 400px | Fig.9 suitable V values from the QH and MD plotted in one graph]]&lt;br /&gt;
! [[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.10 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
This experiment found that the cell volume (i.e. the separations) of MgO crystal increases with the rising temperature, and this causes the thermal expansion. The original of thermal expansion is the anharmonic vibrations due to the asymmetric interatomic potential. The QH method assumes that all the vibrations are harmonic so the cell volume will not change much, but it is true only when the temperature is relatively low; while the MD method is closer to the real life but less economic, it should be used for simulation of vibrations at higher temperature.&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;br /&gt;
&lt;br /&gt;
4. Chemistryworld, http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast, (accessed Nov. 2015)&lt;br /&gt;
&lt;br /&gt;
5. L.S. Dubrovinsky and  S.K. Saxena, &#039;&#039;Phys Chem Minerals&#039;&#039;, 1997, 24: 547–550&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512793</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512793"/>
		<updated>2015-11-20T11:48:43Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; For example in this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function, and the Helmholtz free energy A can be obtained using A = E&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; + 1/2 Σ&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt;ћω&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt; + k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T*Σ&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt;ln[1-exp(-ћω&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt;/k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)]&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function, hence the Helmholtz free energy and the cell volume.&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain state (i.e. frequency). It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonons of an infinite lattice. The following shows the process of finding the best number of k points for computing DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution as a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way (the quasi-harmonic approximation) to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated, and GULP is able to search for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path in k-space is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| 100 meV&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell volume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO (lit&amp;lt;sup&amp;gt;4&amp;lt;/sup&amp;gt;. 3125 K). The reason could be the vibrations was so large that atoms clashed into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.6 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.7 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare the computing thermal expansion with the literature values, several cell volumes were obtained by substituting some specific temperatures into the trend line equation in Fig.7, and the predicted cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt; with the units changed to what is used in the literature.&lt;br /&gt;
&lt;br /&gt;
[[File:Molar volume VS T X.png| thumb|x400px|400px| Fig.8 The general trend and magnitude are consistent between the predicted values and the literature values, but the slope of the literature value is steeper than that of the predicted values, so before 1000 K the predicted values are slightly higher, after the coincidence at this point, the literature values are higher.]] &lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature&amp;lt;/sup&amp;gt;5&amp;lt;sup&amp;gt; Molar Volume (cm&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method MD was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p., MD can compensate the failure of QH. Around the m.p., there is a range where the volume almost keeps constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt; higher, indicating phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt;=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of the data sets obtained from different methods, and V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:V VS T both X.png | thumb | x400px | 400px | Fig.9 suitable V values from the QH and MD plotted in one graph]]&lt;br /&gt;
! [[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.10 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
This experiment found that the cell volume (i.e. the separations) of MgO crystal increases with the rising temperature, and this causes the thermal expansion. The original of thermal expansion is the anharmonic vibrations due to the asymmetric interatomic potential. The QH method assumes that all the vibrations are harmonic so the cell volume will not change much, but it is true only when the temperature is relatively low; while the MD method is closer to the real life but less economic, it should be used for simulation of vibrations at higher temperature.&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;br /&gt;
&lt;br /&gt;
4. Chemistryworld, http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast, (accessed Nov. 2015)&lt;br /&gt;
&lt;br /&gt;
5. L.S. Dubrovinsky and  S.K. Saxena, &#039;&#039;Phys Chem Minerals&#039;&#039;, 1997, 24: 547–550&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512768</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512768"/>
		<updated>2015-11-20T11:26:18Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Theory */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; For example in this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function, and the Helmholtz free energy A can be obtained using A = E&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; + 1/2 Σ&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt;ћω&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt; + k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T*Σ&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt;ln[1-exp(-ћω&amp;lt;sub&amp;gt;j,k&amp;lt;/sub&amp;gt;/k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)]&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain state (i.e. frequency). It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonons of an infinite lattice. The following shows the process of finding the best number of k points for computing DOS. ???除了distribution的词，我想看看average怎么用？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution as a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way (the quasi-harmonic approximation) to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated, and GULP is able to search for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path in k-space is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| 100 meV&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell volume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO (lit&amp;lt;sup&amp;gt;4&amp;lt;/sup&amp;gt;. 3125 K). The reason could be the vibrations was so large that atoms clashed into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.6 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.7 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare the computing thermal expansion with the literature values, several cell volumes were obtained by substituting some specific temperatures into the trend line equation in Fig.7, and the predicted cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt; with the units changed to what is used in the literature.&lt;br /&gt;
&lt;br /&gt;
[[File:Molar volume VS T X.png| thumb|x400px|400px| Fig.8 The general trend and magnitude are consistent between the predicted values and the literature values, but the slope of the literature value is steeper than that of the predicted values, so before 1000 K the predicted values are slightly higher, after the coincidence at this point, the literature values are higher.]] 原因呢？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature&amp;lt;/sup&amp;gt;5&amp;lt;sup&amp;gt; Molar Volume (cm&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method MD was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p., MD can compensate the failure of QH. Around the m.p., there is a range where the volume almost keeps constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt; higher, indicating phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt;=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of the data sets obtained from different methods, and V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:V VS T both X.png | thumb | x400px | 400px | Fig.9 suitable V values from the QH and MD plotted in one graph]]&lt;br /&gt;
! [[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.10 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals ????link????&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;br /&gt;
&lt;br /&gt;
4. Chemistryworld, http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast, (accessed Nov. 2015)&lt;br /&gt;
&lt;br /&gt;
5. L.S. Dubrovinsky and  S.K. Saxena, &#039;&#039;Phys Chem Minerals&#039;&#039;, 1997, 24: 547–550&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512755</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512755"/>
		<updated>2015-11-20T11:16:06Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain state (i.e. frequency). It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonons of an infinite lattice. The following shows the process of finding the best number of k points for computing DOS. ???除了distribution的词，我想看看average怎么用？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution as a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way (the quasi-harmonic approximation) to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated, and GULP is able to search for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path in k-space is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| 100 meV&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell volume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO (lit&amp;lt;sup&amp;gt;4&amp;lt;/sup&amp;gt;. 3125 K). The reason could be the vibrations was so large that atoms clashed into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.6 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.7 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare the computing thermal expansion with the literature values, several cell volumes were obtained by substituting some specific temperatures into the trend line equation in Fig.7, and the predicted cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt; with the units changed to what is used in the literature.&lt;br /&gt;
&lt;br /&gt;
[[File:Molar volume VS T X.png| thumb|x400px|400px| Fig.8 The general trend and magnitude are consistent between the predicted values and the literature values, but the slope of the literature value is steeper than that of the predicted values, so before 1000 K the predicted values are slightly higher, after the coincidence at this point, the literature values are higher.]] 原因呢？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature&amp;lt;/sup&amp;gt;5&amp;lt;sup&amp;gt; Molar Volume (cm&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method MD was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p., MD can compensate the failure of QH. Around the m.p., there is a range where the volume almost keeps constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt; higher, indicating phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt;=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of the data sets obtained from different methods, and V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:V VS T both X.png | thumb | x400px | 400px | Fig.9 suitable V values from the QH and MD plotted in one graph]]&lt;br /&gt;
! [[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.10 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals ????link????&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;br /&gt;
&lt;br /&gt;
4. Chemistryworld, http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast, (accessed Nov. 2015)&lt;br /&gt;
&lt;br /&gt;
5. L.S. Dubrovinsky and  S.K. Saxena, &#039;&#039;Phys Chem Minerals&#039;&#039;, 1997, 24: 547–550&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512739</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512739"/>
		<updated>2015-11-20T11:05:14Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain state (i.e. frequency). It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonons of an infinite lattice. The following shows the process of finding the best number of k points for computing DOS. ???除了distribution的词，我想看看average怎么用？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution as a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way (the quasi-harmonic approximation) to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated, and GULP is able to search for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path in k-space is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| 100 meV&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell volume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO (lit&amp;lt;sup&amp;gt;4&amp;lt;/sup&amp;gt;. 3125 K). The reason could be the vibrations was so large that atoms clashed into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.6 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.7 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare the computing thermal expansion with the literature values, several cell volumes were obtained by substituting some specific temperatures into the trend line equation in Fig.7, and the predicted cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt; with the units changed to what is used in the literature.&lt;br /&gt;
&lt;br /&gt;
[[File:Molar volume VS T X.png| thumb|x400px|400px| Fig.8 The general trend and magnitude are consistent between the predicted values and the literature values, but the slope of the literature value is steeper than that of the predicted values, so before 1000 K the predicted values are slightly higher, after the coincidence at this point, the literature values are higher.]] 原因呢？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature&amp;lt;/sup&amp;gt;5&amp;lt;sup&amp;gt; Molar Volume (cm&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.9 suitable V values from the QH and MD plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method MD was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p., MD can compensate the failure of QH. Around the m.p., there is a range where the volume almost keeps constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt; higher, indicating phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt;=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of the data sets obtained from different methods, and V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.10 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]为什么？？？？&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals ????link????&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;br /&gt;
&lt;br /&gt;
4. Chemistryworld, http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast, (accessed Nov. 2015)&lt;br /&gt;
&lt;br /&gt;
5. L.S. Dubrovinsky and  S.K. Saxena, &#039;&#039;Phys Chem Minerals&#039;&#039;, 1997, 24: 547–550&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512715</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512715"/>
		<updated>2015-11-20T10:55:02Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Reference */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain state (i.e. frequency). It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonons of an infinite lattice. The following shows the process of finding the best number of k points for computing DOS. ???除了distribution的词，我想看看average怎么用？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution as a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way (the quasi-harmonic approximation) to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated, and GULP is able to search for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path in k-space is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| 100 meV&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell volume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO (lit&amp;lt;sup&amp;gt;4&amp;lt;/sup&amp;gt;. 3125 K). The reason could be the vibrations was so large that atoms clashed into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.6 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.7 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare the computing thermal expansion with the literature values, several cell volumes were obtained by substituting some specific temperatures into the trend line equation in Fig.7, and the predicted cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt; with the units changed to what is used in the literature.&lt;br /&gt;
&lt;br /&gt;
[[File:Molar volume VS T X.png| thumb|x400px|400px| Fig.8 The general trend and magnitude are consistent between the predicted values and the literature values, but the slope of the literature value is steeper than that of the predicted values, so before 1000 K the predicted values are slightly higher, after the coincidence at this point, the literature values are higher.]] 原因呢？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature&amp;lt;/sup&amp;gt;5&amp;lt;sup&amp;gt; Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.9 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.10 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals ????link????&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;br /&gt;
&lt;br /&gt;
4. Chemistryworld, http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast, (accessed Nov. 2015)&lt;br /&gt;
&lt;br /&gt;
5. L.S. Dubrovinsky and  S.K. Saxena, &#039;&#039;Phys Chem Minerals&#039;&#039;, 1997, 24: 547–550&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512687</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512687"/>
		<updated>2015-11-20T10:42:52Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Results and Discussion */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain state (i.e. frequency). It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonons of an infinite lattice. The following shows the process of finding the best number of k points for computing DOS. ???除了distribution的词，我想看看average怎么用？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution as a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way (the quasi-harmonic approximation) to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated, and GULP is able to search for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path in k-space is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| 100 meV&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell volume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO (lit&amp;lt;sup&amp;gt;4&amp;lt;/sup&amp;gt;. 3125 K). The reason could be the vibrations was so large that atoms clashed into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.6 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.7 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare the computing thermal expansion with the literature values, several cell volumes were obtained by substituting some specific temperatures into the trend line equation in Fig.7, and the predicted cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt; with the units changed to what is used in the literature.&lt;br /&gt;
&lt;br /&gt;
[[File:Molar volume VS T X.png| thumb|x400px|400px| Fig.8 The general trend and magnitude are consistent between the predicted values and the literature values, but the slope of the literature value is steeper than that of the predicted values, so before 1000 K the predicted values are slightly higher, after the coincidence at this point, the literature values are higher.]] 原因呢？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature&amp;lt;/sup&amp;gt;5&amp;lt;sup&amp;gt; Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.9 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.10 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals ????link????&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;br /&gt;
&lt;br /&gt;
4. Chemistryworld, http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast, (accessed Nov. 2015)&lt;br /&gt;
&lt;br /&gt;
5.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512682</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512682"/>
		<updated>2015-11-20T10:41:15Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Results and Discussion */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain state (i.e. frequency). It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonons of an infinite lattice. The following shows the process of finding the best number of k points for computing DOS. ???除了distribution的词，我想看看average怎么用？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution as a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way (the quasi-harmonic approximation) to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated, and GULP is able to search for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path in k-space is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| 100 meV&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell volume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO (lit&amp;lt;sup&amp;gt;4&amp;lt;/sup&amp;gt;. 3125 K). The reason could be the vibrations was so large that atoms clashed into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.6 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.7 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare the computing thermal expansion with the literature values, several cell volumes were obtained by substituting some specific temperatures into the trend line equation in Fig.7, and the predicted cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt; with the units changed to what is used in the literature.&lt;br /&gt;
&lt;br /&gt;
[[File:Molar volume VS T X.png| thumb|x400px|400px| The general trend and magnitude are consistent between the predicted values and the literature values, but the slope of the literature value is steeper than that of the predicted values, so before 1000 K the predicted values are slightly higher, after the coincidence at this point, the literature values are higher.]] 原因呢？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature&amp;lt;/sup&amp;gt;5&amp;lt;sup&amp;gt; Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals ????link????&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;br /&gt;
&lt;br /&gt;
4. Chemistryworld, http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast, (accessed Nov. 2015)&lt;br /&gt;
&lt;br /&gt;
5.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=File:Molar_volume_VS_T_X.png&amp;diff=512670</id>
		<title>File:Molar volume VS T X.png</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=File:Molar_volume_VS_T_X.png&amp;diff=512670"/>
		<updated>2015-11-20T10:37:04Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512616</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512616"/>
		<updated>2015-11-20T10:19:32Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Results and Discussion */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain state (i.e. frequency). It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonons of an infinite lattice. The following shows the process of finding the best number of k points for computing DOS. ???除了distribution的词，我想看看average怎么用？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution as a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way (the quasi-harmonic approximation) to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated, and GULP is able to search for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path in k-space is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| 100 meV&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell volume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO (lit&amp;lt;sup&amp;gt;4&amp;lt;/sup&amp;gt;. 3125 K). The reason could be the vibrations was so large that atoms clashed into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare the computing thermal expansion with the literature values, several cell volumes were obtained by substituting some specific temperatures into the trend line equation in Fig.6, and the predicted cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt; with the units changed to what is used in the literature.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals ????link????&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;br /&gt;
&lt;br /&gt;
4. Chemistryworld, http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast, (accessed Nov. 2015)&lt;br /&gt;
&lt;br /&gt;
5.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512596</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512596"/>
		<updated>2015-11-20T10:13:28Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Reference */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain state (i.e. frequency). It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonons of an infinite lattice. The following shows the process of finding the best number of k points for computing DOS. ???除了distribution的词，我想看看average怎么用？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution as a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way (the quasi-harmonic approximation) to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated, and GULP is able to search for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path in k-space is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| 100 meV&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell volume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO (lit&amp;lt;sup&amp;gt;4&amp;lt;/sup&amp;gt;. 3125 K) {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason could be the vibrations was so large that atoms clashed into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals ????link????&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;br /&gt;
&lt;br /&gt;
4. Chemistryworld, http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast, (accessed Nov. 2015)&lt;br /&gt;
&lt;br /&gt;
5.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512594</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512594"/>
		<updated>2015-11-20T10:10:57Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Results and Discussion */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain state (i.e. frequency). It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonons of an infinite lattice. The following shows the process of finding the best number of k points for computing DOS. ???除了distribution的词，我想看看average怎么用？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution as a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way (the quasi-harmonic approximation) to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated, and GULP is able to search for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path in k-space is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| 100 meV&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell volume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO (lit&amp;lt;sup&amp;gt;4&amp;lt;/sup&amp;gt;. 3125 K) {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason could be the vibrations was so large that atoms clashed into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals ????link????&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512555</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512555"/>
		<updated>2015-11-20T09:44:41Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Results and Discussion */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-Γ-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations, but ions in the real solid crystal do not stay still.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the theory part, to understand the variation of vibrational frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals ????link????&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512549</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512549"/>
		<updated>2015-11-20T09:38:42Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Reference */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals ????link????&lt;br /&gt;
&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
&lt;br /&gt;
3. B. K. Vainshtein, V. M. Fridkin and V. L. Indenbom, &#039;&#039;Modern Crystallography 2: Structure of Crystals&#039;&#039;, Springer, London, 3rd Edn., 2000, pp. 64.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512543</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512543"/>
		<updated>2015-11-20T09:33:22Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Results and Discussion */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV (lit&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;.41.197 eV). Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals link????&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
3.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512522</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512522"/>
		<updated>2015-11-20T09:12:01Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Theory */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When the structure is expanded to two dimensions, particles can vibrate up and down with respect to the horizontal axis more than just along the axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in a Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path are set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed. 这句话大概可以改改？？&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals link????&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
3.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512516</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512516"/>
		<updated>2015-11-20T09:02:52Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Theory */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When a 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).为什么k的范围是这样，然后怎么解释fold？？？？&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exists two frequencies for one k-vector due to the two types of oscillations between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;. The same reason causes the gap between the acoustic branch and the optical branch.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals link????&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
3.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512499</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512499"/>
		<updated>2015-11-20T08:53:04Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Reference */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1. From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals link????&lt;br /&gt;
2. J. M. Seddon and J. D. Ga, &#039;&#039;Thermodynamics and Statistical Mechanics&#039;&#039;, the Royal Society of Chemistry, Cambridge, 2001&lt;br /&gt;
3.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512481</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512481"/>
		<updated>2015-11-20T08:41:34Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Theory */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different states of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1, From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals link????&lt;br /&gt;
2.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512472</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512472"/>
		<updated>2015-11-20T08:36:51Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Theory */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function &#039;&#039;&#039;&#039;&#039;q&#039;&#039;&#039;&#039;&#039; is correctly expressed, the properties of the system can be calculated.&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1, From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals link????&lt;br /&gt;
2.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512465</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512465"/>
		<updated>2015-11-20T08:19:41Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;br /&gt;
&lt;br /&gt;
=Reference=&lt;br /&gt;
1, From Prof N. M. Harrison’s Lectrure Notes: Vibrations in crystals link????&lt;br /&gt;
2.&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512430</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512430"/>
		<updated>2015-11-20T04:57:57Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. The &#039;&#039;&#039;quasi-harmonic approximation (QH)&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. The &#039;&#039;&#039;molecular dynamics (MD)&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on the QH method. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. by QH]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by QH]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from the QH and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by QH, and after the m.p. MD can compensate the failure of QH. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and QH) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512394</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512394"/>
		<updated>2015-11-20T02:59:51Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. Quasi-harmonic Approximation]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by Quasi-harmonic Approximation]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from Quasi-harmonic Approx. and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by Quasi-harmonic Approximation, and after the m.p. MD can compensate the failure of Quasi. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
&lt;br /&gt;
=Conclusions=&lt;br /&gt;
&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512391</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512391"/>
		<updated>2015-11-20T02:41:27Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. Quasi-harmonic Approximation]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by Quasi-harmonic Approximation]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from Quasi-harmonic Approx. and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by Quasi-harmonic Approximation, and after the m.p. MD can compensate the failure of Quasi. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/V&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
[[File:Expansion coefficient both X.png|thumb|x400px|400px|Fig.7 Comparison between α&amp;lt;sub&amp;gt;V&amp;lt;/sub&amp;gt; obtained from different methods. It can be concluded that the difference is decreasing from low temperature to 1500 K after which the difference in expansion coefficient tends to keep constant.]]&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=File:Expansion_coefficient_both_X.png&amp;diff=512390</id>
		<title>File:Expansion coefficient both X.png</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=File:Expansion_coefficient_both_X.png&amp;diff=512390"/>
		<updated>2015-11-20T02:36:22Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512373</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512373"/>
		<updated>2015-11-20T02:19:22Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. Quasi-harmonic Approximation]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by Quasi-harmonic Approximation]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from Quasi-harmonic Approx. and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by Quasi-harmonic Approximation, and after the m.p. MD can compensate the failure of Quasi. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The change of cell volume can be describes as thermal expansion coefficient α=(1/v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt;)(dV/dT). This property can be calculated for both of data sets obtained from both methods, and v&amp;lt;sub&amp;gt;0&amp;lt;/sub&amp;gt; is the zero-point volume in each case and the (dV/dT) is the gradient of the trend lines.&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512363</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512363"/>
		<updated>2015-11-20T02:08:27Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. Quasi-harmonic Approximation]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by Quasi-harmonic Approximation]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Generally, the trend and the magnitude are consistent between the predicted values and literature values, but the slope of the literature value is steeper than that of the predicted values, so before 多少度 the predicted values are slightly higher, they come together at aldjfhalkdfjhjh, afterwards这里要不要plot个图啊？&lt;br /&gt;
&lt;br /&gt;
[[File:V VS T both X.png | thumb | x400px | 400px | Fig.6 suitable V values from Quasi-harmonic Approx. and Melecular Dynamics plotted in one graph]]&lt;br /&gt;
&lt;br /&gt;
As mentioned in the introduction part, another simulation method called Molecular Dynamics was also used to calculate the equilibrium energy and volume. The MD obtained values before the melting point of MgO are similar compared to those obtained by Quasi-harmonic Approximation, and after the m.p. MD can compensate the failure of Quasi. When around m.p, there is a range where the volume almost constant, indicating phase changing. When T reaches 4000 K, the volume is lifted by more than 10 Angstrom higher, phase change is completed and the volume of liquid phase will continue increase with the raising temperature but with a steeper gradient. Atkins ref?????? If the temperature goes on increasing, the volume will become infinite as the gas phase does not have a volume without any pressure.&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=File:V_VS_T_both_X.png&amp;diff=512361</id>
		<title>File:V VS T both X.png</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=File:V_VS_T_both_X.png&amp;diff=512361"/>
		<updated>2015-11-20T02:06:59Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512309</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512309"/>
		<updated>2015-11-20T00:11:26Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. Quasi-harmonic Approximation]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by Quasi-harmonic Approximation]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|455&lt;br /&gt;
| 18.9425&lt;br /&gt;
| 11.4034&lt;br /&gt;
|11.3004&lt;br /&gt;
|-&lt;br /&gt;
|710&lt;br /&gt;
| 19.0601&lt;br /&gt;
| 11.4742&lt;br /&gt;
|11.4109&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512248</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512248"/>
		<updated>2015-11-19T23:23:53Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. Quasi-harmonic Approximation]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by Quasi-harmonic Approximation]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|12.9244&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512245</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512245"/>
		<updated>2015-11-19T23:22:56Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. Quasi-harmonic Approximation]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by Quasi-harmonic Approximation]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated by substituting some specific temperatures into the trend line equation in Fig.6. to get the predicted cell volumes. The cell volumes were then transferred into molar volume by multiplying Avogadro constant N&amp;lt;sub&amp;gt;A&amp;lt;/sub&amp;gt;, the units were also changed to what is used in the literature.&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Cell Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
! Literature Molar Volume (cm&amp;lt;sup&amp;gt;-3&amp;lt;/sup&amp;gt;mol&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
|298&lt;br /&gt;
| 18.8851&lt;br /&gt;
| 11.3688&lt;br /&gt;
|11.2434&lt;br /&gt;
|-&lt;br /&gt;
|1096&lt;br /&gt;
| 19.2570&lt;br /&gt;
| 11.5927&lt;br /&gt;
|11.6211&lt;br /&gt;
|-&lt;br /&gt;
|1527&lt;br /&gt;
| 19.4673&lt;br /&gt;
| 11.7193&lt;br /&gt;
|11.8218&lt;br /&gt;
|-&lt;br /&gt;
|2106&lt;br /&gt;
| 19.7624&lt;br /&gt;
| 11.8970&lt;br /&gt;
|12.2287&lt;br /&gt;
|-&lt;br /&gt;
|2703&lt;br /&gt;
| 20.3014&lt;br /&gt;
| 12.2214&lt;br /&gt;
|12.6887&lt;br /&gt;
|-&lt;br /&gt;
|2986&lt;br /&gt;
| 20.7658&lt;br /&gt;
| 12.5010&lt;br /&gt;
|-&lt;br /&gt;
|3015&lt;br /&gt;
| 20.8248&lt;br /&gt;
| 12.5365&lt;br /&gt;
|12.9723&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512109</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512109"/>
		<updated>2015-11-19T22:18:07Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. Quasi-harmonic Approximation]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by Quasi-harmonic Approximation]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature thermal expansion, several volumes were calculated from the trend line equation in Fig.6.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512054</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512054"/>
		<updated>2015-11-19T21:43:44Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. The Helmholtz free energy and the cell colume were optimised to observe the variations with different temperature. As temperature is raising, the Helmholtz Free Energy becomes more negative, while the cell volume is expanding, Both of the variations can be well expressed by polynomial equations. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. Quasi-harmonic Approximation]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by Quasi-harmonic Approximation]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
In order to compare with literature, several volume were calculated &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512026</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512026"/>
		<updated>2015-11-19T21:25:07Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. The Helmholtz free energy and the cell colume were optimised with raising temperature. The Helmholtz Free Energy becomes more negative with the raising temperature, and the volume is expanding when temperature is raised. Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. Quasi-harmonic Approximation]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by Quasi-harmonic Approximation]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512009</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512009"/>
		<updated>2015-11-19T21:01:39Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. &lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temp. Quasi-harmonic Approximation]]&lt;br /&gt;
[[File:QHA V VS T X.png | thumb | x500px | 500px | Fig.6 Cell Volume VS Temp. calculated by Quasi-harmonic Approximation]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|-&lt;br /&gt;
| 3100&lt;br /&gt;
| Calculation failed when temperature is close to the melting point of MgO, 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} &lt;br /&gt;
| The reason can be the vibrations is so large that atoms clashes into each other causing computing errors.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=File:QHA_V_VS_T_X.png&amp;diff=512005</id>
		<title>File:QHA V VS T X.png</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=File:QHA_V_VS_T_X.png&amp;diff=512005"/>
		<updated>2015-11-19T20:58:23Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512000</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=512000"/>
		<updated>2015-11-19T20:49:31Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. &lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x500px | 500px | Fig.5 Helmholtz Free Energy VS Temperature calculated by Quasi-harmonic Approximation. Calculation failed at 3100 K which is close the melting point 3125 K {http://www.rsc.org/chemistryworld/2014/08/magnesium-oxide-mgo-podcast} of MgO]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
! Lattice Volume (Å&amp;lt;sup&amp;gt;3&amp;lt;/sip&amp;gt;)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| 18.8365&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|18.8383&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|18.8562&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|18.8900&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|18.9325&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|18.9801&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|19.0312&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|19.0851&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|19.1413&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|19.1997&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|19.2601&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|19.3872&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|19.5233&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|19.6698&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|19.8287&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|20.0029&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|20.3047&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|20.6889&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|21.3217&lt;br /&gt;
|-&lt;br /&gt;
| 3100&lt;br /&gt;
| calculation failed&lt;br /&gt;
| the reason can be the vibrations is so large that atoms clashes into each other causing computing errors&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511994</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511994"/>
		<updated>2015-11-19T20:36:04Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. &lt;br /&gt;
&lt;br /&gt;
[[File:A VS T X.png | thumb | x400px | 400px | Fig.5 Helmholtz Free Energy VS Temperature calculated by Quasi-harmonic Approximation]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|-&lt;br /&gt;
| 3100&lt;br /&gt;
| calculation failed&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=File:A_VS_T_X.png&amp;diff=511991</id>
		<title>File:A VS T X.png</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=File:A_VS_T_X.png&amp;diff=511991"/>
		<updated>2015-11-19T20:31:57Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511988</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511988"/>
		<updated>2015-11-19T20:31:11Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: /* Results and Discussion */&lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. &lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
|-&lt;br /&gt;
| 3100&lt;br /&gt;
| calculation failed&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511985</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511985"/>
		<updated>2015-11-19T20:29:41Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. GULP searches for the minimum free energy with respect to the structure via calculating the internal energy and phonons at a sequence of geometries. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| 0.01 meV&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
A grid size of 11*11*11 (i.e. 0.01 meV accuracy) was chosen for the following calculations based on Quasi-harmonic Approximation. &lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Temperature (K)&lt;br /&gt;
! Helmholtz Free Energy A (eV)&lt;br /&gt;
|-&lt;br /&gt;
| 0&lt;br /&gt;
| -40.9019&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 100&lt;br /&gt;
| -40.9024&lt;br /&gt;
|-&lt;br /&gt;
| 200&lt;br /&gt;
| -40.9094&lt;br /&gt;
|-&lt;br /&gt;
| 300&lt;br /&gt;
| -40.9281&lt;br /&gt;
|-&lt;br /&gt;
| 400&lt;br /&gt;
| -40.9586&lt;br /&gt;
|-&lt;br /&gt;
| 500&lt;br /&gt;
| -40.9994&lt;br /&gt;
|-&lt;br /&gt;
| 600&lt;br /&gt;
| -41.0493&lt;br /&gt;
|-&lt;br /&gt;
| 700&lt;br /&gt;
| -41.1071&lt;br /&gt;
|-&lt;br /&gt;
| 800&lt;br /&gt;
| -41.1719&lt;br /&gt;
|-&lt;br /&gt;
| 900&lt;br /&gt;
| -41.2430&lt;br /&gt;
|-&lt;br /&gt;
| 1000&lt;br /&gt;
| -41.3110&lt;br /&gt;
|-&lt;br /&gt;
| 1200&lt;br /&gt;
| -41.4887&lt;br /&gt;
|-&lt;br /&gt;
| 1400&lt;br /&gt;
| -41.6755&lt;br /&gt;
|-&lt;br /&gt;
| 1600&lt;br /&gt;
| -41.8780&lt;br /&gt;
|-&lt;br /&gt;
| 1800&lt;br /&gt;
| -42.0944&lt;br /&gt;
|-&lt;br /&gt;
| 2000&lt;br /&gt;
| -42.3237&lt;br /&gt;
|-&lt;br /&gt;
| 2300&lt;br /&gt;
| -42.6895&lt;br /&gt;
|-&lt;br /&gt;
| 2600&lt;br /&gt;
| -43.0800&lt;br /&gt;
|-&lt;br /&gt;
| 2900&lt;br /&gt;
| -43.4948&lt;br /&gt;
| -&lt;br /&gt;
| 3100&lt;br /&gt;
| calculation failed&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511555</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511555"/>
		<updated>2015-11-19T15:32:44Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| - 40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| - 40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| - 40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511528</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511528"/>
		<updated>2015-11-19T15:24:48Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| - 40.930301&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| - 40.926609&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| - 40.926432&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| - 40.926450&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
| - 40.926463&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| - 40.926471&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| - 40.926475&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| - 40.926478&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| －40.926479&lt;br /&gt;
|&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| －40.926480&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| －40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| －40.926481&lt;br /&gt;
| &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511517</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511517"/>
		<updated>2015-11-19T15:21:40Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| -40.9303&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| -40.9266&lt;br /&gt;
| 1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| -40.9264&lt;br /&gt;
| 0.5 meV&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| -40.9265&lt;br /&gt;
| &lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
|-40.9265&lt;br /&gt;
| 0.1 meV&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| -40.9265&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| -40.9265&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| -40.9265&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| －40.926479&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| －40.926480&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| －40.926481&lt;br /&gt;
| convergence&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| －40.926481&lt;br /&gt;
| Column 3, Row &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511480</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511480"/>
		<updated>2015-11-19T15:05:30Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x350px | 350px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
|The DOS for 1x1x1 grid was computed for a single k-point &#039;L&#039;. There are 4 distinct peaks, the two around 300 and 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; is double in intensity compared to the 700 and 800 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, which is corresponding to the two branches across point &#039;L&#039; shown in the dispersion curve.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
The DOS is converging which means it is converging.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing over more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution which is a close approximation to the infinite lattice economically. &lt;br /&gt;
&lt;br /&gt;
Since there is a way to compute all the phonon modes in MgO infinite lattice, the free energy of it can also be calculated. Similarly, the computing path is the same as that for computing DOS, so there is also the grid size problem.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Helmholtz Free Energy (eV)&lt;br /&gt;
! Accuracy&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| -40.9303&lt;br /&gt;
| Column 3, Row 2&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| -40.9266&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 3&lt;br /&gt;
| -40.9264&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| -40.9265&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 5&lt;br /&gt;
|-40.9265&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 6&lt;br /&gt;
| -40.9265&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 7&lt;br /&gt;
| -40.9265&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| -40.9265&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 9&lt;br /&gt;
| －40.926479&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 10&lt;br /&gt;
| －40.926480&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 11&lt;br /&gt;
| －40.926481&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 12&lt;br /&gt;
| －40.926481&lt;br /&gt;
| Column 3, Row &lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511362</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511362"/>
		<updated>2015-11-19T14:00:24Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x300px | 300px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate (DOS), indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice. The following shows the process of finding the best number of k points for DOS.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 4 distinct peaks, summing over more k points to get a distribution&lt;br /&gt;
The density of states for 1x1x1 grid was computed for a single k-point &#039;L&#039;.&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;&lt;br /&gt;
lower density of each peaks compared to the DOS above&lt;br /&gt;
A grid size of 2*2*2 is sufficient to get the correct highest value.&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density &lt;br /&gt;
not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
Distribution features appear.&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| There is no much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
As the grid size increases, more possible vibrations are sampled and the distribution is smoothened, nevertheless, the change in DOS decreases each time the grid size is doubled.  Computing along more k-points requires more resources and time, which is obvious from grid 32*32*32 to 64*64*64. A compromise can be grid 32*32*32 which can give a good enough distribution economically. &lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511322</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511322"/>
		<updated>2015-11-19T13:41:37Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x300px | 300px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate, indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 4 distinct peaks, summing over more k points to get a distribution&lt;br /&gt;
The density of states for 1x1x1 grid was computed for a single k-point &#039;L&#039;.&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt;, lower density of each peaks&lt;br /&gt;
grid size 2*2*2 is sufficient to get the correct highest value&lt;br /&gt;
summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| more peaks, highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with lowered density, not a smooth distribution yet&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
|  highest peak at 400 cm&amp;lt;sup&amp;gt;-1&amp;lt;/sup&amp;gt; with even lowered density&lt;br /&gt;
large fluctuations&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| a general shape of the distribution with small fluctuations&lt;br /&gt;
no much change in the densities&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| Nice distribution compared to the upper one&lt;br /&gt;
acceptable computation time&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| No much change compare to the DOS computed along grid 32*32*32, however takes a few minutes longer to compute.&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511303</id>
		<title>Rep:MgO:XL</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=Rep:MgO:XL&amp;diff=511303"/>
		<updated>2015-11-19T13:25:40Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;=Introduction=&lt;br /&gt;
&lt;br /&gt;
Magnesium oxide naturally exists as crystal based on face-centred cubic lattice with the lattice points taken by Mg&amp;lt;sup&amp;gt;2+&amp;lt;/sup&amp;gt; and the octahedral holes filled with O&amp;lt;sup&amp;gt;2-&amp;lt;/sup&amp;gt;.&lt;br /&gt;
[[File:MgO unit cell X.png | right | x300px | 300px | thumb | Fig.1 conventional cell ( in black ) and primitive cell ( in light blue ) of MgO; The volume of primitive cell is 1/4 of that of conventional cell.]]&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Type of Unit Cell&lt;br /&gt;
! Shape&lt;br /&gt;
! Parameter&lt;br /&gt;
! Internal Angel&lt;br /&gt;
! Volume&lt;br /&gt;
! Number of MgO&lt;br /&gt;
|-&lt;br /&gt;
| Conventional&lt;br /&gt;
| Cube&lt;br /&gt;
| a&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4.212 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 90&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 74.725 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;c&amp;lt;/sub&amp;gt; = 4&lt;br /&gt;
|-&lt;br /&gt;
| Primitive&lt;br /&gt;
| Rhombohedron&lt;br /&gt;
| a&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = b&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = c&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 2.978 ‎Å&lt;br /&gt;
| α&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = β&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = γ&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 60&amp;lt;sup&amp;gt;•&amp;lt;/sup&amp;gt;&lt;br /&gt;
| V&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 18.6812 Å&amp;lt;sup&amp;gt;3&amp;lt;/sup&amp;gt;&lt;br /&gt;
| N&amp;lt;sub&amp;gt;p&amp;lt;/sub&amp;gt; = 1&lt;br /&gt;
|} &lt;br /&gt;
&lt;br /&gt;
Vibrations of a solid system are related to many of its physical properties such as free energy, heat capacity, expansion, phase transition, thermal conductivity and dielectric phenomena at low frequencies. This study compares two methods for simulation of MgO crystal vibrations. &#039;&#039;&#039;Quasi-harmonic Approximation&#039;&#039;&#039; considers vibrations as  phonons representing elementary vibrational modes in which a lattice of particles uniformly oscillates at a single frequency. &#039;&#039;&#039;Molecular Dynamics&#039;&#039;&#039; allows the particle in the system to interact for a given time period and the coordinates of the particles are numerically solved based on Newton&#039;s Laws &#039;&#039;R&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;final&amp;lt;/sub&amp;gt;*dt = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + a*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; = R&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt; + v&amp;lt;sub&amp;gt;initial&amp;lt;/sub&amp;gt;*dt + (F/m)*dt&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;&#039;&#039;. Both methods were conducted on Linux based programme GULP (General Utility Lattice Program) via the user interface for constructing and visualizing provided by DL Visualize.&lt;br /&gt;
&lt;br /&gt;
=Theory=&lt;br /&gt;
&lt;br /&gt;
In statistical mechanics, the physical properties of a system are in Boltzmann Distribution &#039;&#039;&#039;&#039;&#039;n&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt; / N = exp (-βu&amp;lt;sub&amp;gt;i&amp;lt;/sub&amp;gt;) / q&#039;&#039;&#039;&#039;&#039; where &#039;&#039;&#039;&#039;&#039;β = 1 / (k&amp;lt;sub&amp;gt;b&amp;lt;/sub&amp;gt;T)&#039;&#039;&#039;&#039;&#039; and &#039;&#039;&#039;&#039;&#039;q = Σ&amp;lt;sub&amp;gt;j=1&amp;lt;/sub&amp;gt;&amp;lt;sup&amp;gt;levels&amp;lt;/sup&amp;gt; exp (-βu&amp;lt;sub&amp;gt;j&amp;lt;/sub&amp;gt;)&#039;&#039;&#039;&#039;&#039;. This means that once the partition function q is correctly expressed, the properties of the system can be calculated. 要舉個例子嗎？In this experiment, in accordance with harmonic oscillation model, the vibrational frequency ω must be quantised and summing over the frequencies will lead to the partition function.&lt;br /&gt;
&lt;br /&gt;
MgO crystal is made of repeating unit cells, so it is sensible to start with the simplest model first to see how frequency ω is related to the repeating structure. When 1-dimensional chain of one kind of atom vibrate, they can have several different types of vibrations and each one can be described as a wave with a wavelength equal to the length of the repeating unit (Fig.2) and plotting the vibrational frequencies VS the k-vectors (showing directions and wavelengths of vibrations) gives a graph like Fig.3. If each atom in this chain is superseded by a MgO, there is now a pair of ions in each repeat unit, a&#039; = 2a, hence -π/(2a) &amp;lt; k &amp;lt; π/(2a) and folding branch occurs (Fig.4).&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! [[File:1D vibrations L.png | x400px | 400px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.2 1D vibrations; wave vector &#039;&#039;&#039;&#039;&#039;k = 2π / λ&#039;&#039;&#039;&#039;&#039;; -π/2 &amp;lt; k &amp;lt; π/2; frequency ω increases as k increases.]]&lt;br /&gt;
! [[File:Typical dispersion curve X.png | x250px | 250px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.3 typical ω(k) is plotted as a dispersion curve; k = 0 at Γ point ]]&lt;br /&gt;
! [[File:1D Diatomic Chain X.png | x280px | 280px | thumb | Fig&amp;lt;sup&amp;gt;1&amp;lt;/sup&amp;gt;.4 1D MgO chain; There exist two frequencies for one k-vector due to different oscillation between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt;.]]&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
Both structures mentioned above are limited in 1 dimension. When this is expanded to two dimension, particles can also vibrate up and down with respect to the horizontal axis, hence k-vectors are expressed as (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;) in Cartesian coordinate system, and the ω(k) plot becomes a dispersion surface with frequency ω showed in z-axis. It is now easy to see that for 3-dimensional MgO crystal, k points includes (k&amp;lt;sub&amp;gt;x&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;y&amp;lt;/sub&amp;gt;, k&amp;lt;sub&amp;gt;z&amp;lt;/sub&amp;gt;), and there will be four Cartesian axises for a  ω(k) plot, which is not able to show in real life. In this case, a certain path in the 3-dimensional solid is set and the coordinates through the path were set as the k points, thus ω(k) can be plotted against the path and it is again back to the dispersion curve.&lt;br /&gt;
&lt;br /&gt;
Once the all the vibrational branches are obtained, sum over them to form the partition function and the vibrational energy levels can be computed.&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
=Results and Discussion=&lt;br /&gt;
&lt;br /&gt;
[[File:Phonon dispersion X.png | thumb | right | x300px | 300px | Fig.5 Dispersion Curve of MgO lattice vibrations; path W-L-G-W-X-K on horizontal axis with coordinates shown in blue]]&lt;br /&gt;
&lt;br /&gt;
The lattice energy of MgO calculated is -41.075 eV, and this is the potential energy holding the lattice together induced by electrostatic interaction between Mg&amp;lt;sub&amp;gt;2+&amp;lt;/sub&amp;gt; and O&amp;lt;sub&amp;gt;2-&amp;lt;/sub&amp;gt; ions, which means to move all the ions in the lattice apart to infinity requires an energy of 41.075 eV. Also, this equals to the internal energy of an ideal MgO lattice as perfect crystals have no vibrations.&lt;br /&gt;
&lt;br /&gt;
As mentioned in the Theory part, to understand the variation of frequencies with k, a dispersion curve is essential. To deal with the 3-dimensional MgO infinite lattice, a conventional path in the k-space is used to compute the vibrational modes, and for Fig.5, 50 points along the path was computed and shows all the phonon modes.&lt;br /&gt;
&lt;br /&gt;
The strategy to sum up the phonon modes is to construct the Density of Sate, indicating the probability of a phonon to be in a certain frequency. It is important to sum up phonons for an adequate number of k points so that the distribution of them can be represent the distribution of phonos of an infinite lattice.&lt;br /&gt;
&lt;br /&gt;
{| class=&amp;quot;wikitable&amp;quot;&lt;br /&gt;
! Shrinking Factor&lt;br /&gt;
! Density of State&lt;br /&gt;
! Comments&lt;br /&gt;
|-&lt;br /&gt;
| 1&lt;br /&gt;
| [[File:Phonon DOS 1 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 4 distinct peaks, summing over more k points to get a distribution&lt;br /&gt;
The density of states for 1x1x1 grid was computed for a single k-point &#039;L&#039;.&lt;br /&gt;
|-&lt;br /&gt;
| 2&lt;br /&gt;
| [[File:Phonon DOS 2 X.png|thumb|x250px|250px]]&lt;br /&gt;
| 7 distinct peaks, summing over more k points to get a distribution&lt;br /&gt;
|-&lt;br /&gt;
| 4&lt;br /&gt;
| [[File:Phonon DOS 4 X.png|thumb|x250px|250px]]&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 8&lt;br /&gt;
| [[File:Phonon DOS 8 X.png|thumb|x250px|250px]]&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 16&lt;br /&gt;
| [[File:Phonon DOS 16 X.png|thumb|x250px|250px]]&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 32&lt;br /&gt;
| [[File:Phonon DOS 32 X.png|thumb|x250px|250px]]&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|-&lt;br /&gt;
| 64&lt;br /&gt;
| [[File:Phonon DOS 64 X.png|thumb|x250px|250px]]&lt;br /&gt;
| Column 3, Row 3&lt;br /&gt;
|}&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
The properties of materials (solids, liquids, gasses) are a statistical average over the many different energy states of the molecules making up the material. The vibrational free energy of H2 can be computed analytically by summing over the harmonic vibrations of the molecule. This cannot be done by hand for a real material containing many atoms.&lt;br /&gt;
&lt;br /&gt;
In this laboratory you will use a simple model of atomic interactions to calculate the energy and vibrations of a crystal of MgO. These vibrational energy levels will then be used to compute the free energy of the crystal and to predict how the material expands when heated. In the last final stage you will go beyond the harmonic (and quasi-harmonic) approximation and expand the crystal using a technique called molecular dynamics - essentially reproducing the actual vibration motions of the atoms. Fortunately the computer will do most of the work !&#039;&#039;&#039;&lt;br /&gt;
Vibrations are quantised and can be seen as particles called phonons. The energy of phonon is expressed in XXXX equation. &lt;br /&gt;
When the ensemble is large enough, the &lt;br /&gt;
Vibrational energies of H2 are quantised with the expression ε&amp;lt;sub&amp;gt;vib&amp;lt;/sub&amp;gt; = (n+1/2)ћω&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
&lt;br /&gt;
Body of the text&lt;br /&gt;
write it like a scientific paper (well-articulated sentences,&lt;br /&gt;
NOT a list of two-word answers)&lt;br /&gt;
analyse critically obtained data and given answers&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
add literature/web citations whenever a comparison with&lt;br /&gt;
experimental data is required&lt;br /&gt;
add explicitely every formu&lt;br /&gt;
&lt;br /&gt;
Pictures&lt;br /&gt;
max 20&lt;br /&gt;
reasonably sized (NOT one-page sized pictures, but still&lt;br /&gt;
readable)&lt;br /&gt;
white background (follow the instructions given on the&lt;br /&gt;
website clicking on the link ’How to save a picture for your&lt;br /&gt;
report’)&lt;br /&gt;
described in caption or in the text&lt;br /&gt;
&lt;br /&gt;
Graphs&lt;br /&gt;
add labels and units&lt;br /&gt;
add a critical comment whenever required (NOT a merely&lt;br /&gt;
descriptive comment)la used one to obtain results&lt;br /&gt;
check spelling&lt;br /&gt;
Tables&lt;br /&gt;
add labels and units&lt;br /&gt;
round numerical answers to a specific number of decimal&lt;br /&gt;
places (i.e. 4)&lt;br /&gt;
repeat heading if the table cannot fit in a single page&lt;br /&gt;
&lt;br /&gt;
Conclusions&lt;br /&gt;
give a general description of your calculations and your&lt;br /&gt;
main findings&lt;br /&gt;
outline the differences between the methods in use and the&lt;br /&gt;
results obtained&lt;br /&gt;
analyse critically these differences&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
	<entry>
		<id>https://chemwiki.ch.ic.ac.uk/index.php?title=File:Phonon_DOS_64_X.png&amp;diff=511302</id>
		<title>File:Phonon DOS 64 X.png</title>
		<link rel="alternate" type="text/html" href="https://chemwiki.ch.ic.ac.uk/index.php?title=File:Phonon_DOS_64_X.png&amp;diff=511302"/>
		<updated>2015-11-19T13:25:30Z</updated>

		<summary type="html">&lt;p&gt;Xl7712: &lt;/p&gt;
&lt;hr /&gt;
&lt;div&gt;&lt;/div&gt;</summary>
		<author><name>Xl7712</name></author>
	</entry>
</feed>