Talk:Org:sb807
Q1: Your calculated energies are good, although the endo dimer should be a little higher in energy. The dimerisation is correctly identified as being under kinetic control, backing it up with a reference was a nice touch. You have listed the different energy contributions, but a comparison between the data would have been useful. The discussion of kinetic and thermodynamic control is good and the diagram is helpful in conveying the point concisely; it should be noted that while formation of the higher energy product shows kinetic control, formation of the lower energy product could imply thermodynamic OR kinetic control -> the thermodynamic product could also be the kinetic product. In this case it may be possible to change the control you get by varying the hydrogenation protocol; typically, a transition metal catalyst is used (in the presence of hydrogen gas).
Q2: The calculations appear to be good and the reasoning about the relative stereocontrol is correct: the Grignard reagent can coordinate to the proximal carbonyl and aniline attacks from the opposite face to the carbonyl group due to electron pair repulsion. Discussion of the relevant dihedral angle was a nice clear way of explaining this. As a side point, there may be multiple minima on the energy surface, but when you talk about one of them you should refer to it in the singular form – e.g. minimum.
Q3: You have correctly identified the lowest energy conformer, although you should have found that the twist-boat form of the 6-ring is the most stable conformation in one case. It would have been nice to see some of the different iterations presented, the energies with the 6-ring in different conformations. The explanation of olefin stability is good, as it is indeed defined by the difference in energy between the olefin and the analogous alkane (specifically the strain energy)
Q4: Your MO diagrams all look good and are presented nicely and indeed the syn double bond is the most nucleophilic. The values you got for the IR stretches are all reasonable as is the observation that the C=C stretch doesn’t change much when you hydrogenate the other double bond, whereas the C-Cl bond is strengthened by the absence of the double bond anti to it. This donation of the pi bond density into the sigma* could be seen as half-way to forming a carbonium ion (non-classical carbocation), which would be the case if you donated enough density to break the C-Cl bond completely.
MINI-PROJECT- The discussion of the mechanism at the start is great, just the right amount of information for a concise introduction. It’s nice to see you have applied the computational chemistry to another problem not specifically asked for, namely the difference in energy between the quinoline and dihydroquinoline – not many students did any extra work of this kind. The discussion of the expected chemical shifts is good as it shows how you focussed the comparison on only the important different peaks. There is a chemist at Oxford (Prof. Jonathan Burton) who conducts NMR predictions in a similar way, focusing on the fused part of ring system in his case and using the 13C shifts as the distinguishing feature since these are accurately calculated. The way you presented the errors in the shifts was fine, but perhaps a better way is to show a bar chart from which it is easy to see where the major errors lie and how good the method is overall.