Talk:Mod1:david90
Q1: The energy values you have found for the two isomers of the dimer and the hydrogenated endo products are spot on and the reaction is indeed under kinetic control. As you have identified, the product distribution is therefore dependent on the relative energies of transition states – the endo-transition state is in fact more stable because it involves a secondary orbital interaction not possible in the exo-transition state. As you say, bending strain is the major factor in the different energies of the two mono-hydrogenated endo-dimers due to deviation from ideal sp2 bond angles. Although the thermodynamic product is formed, this does not necessarily mean that hydrogenation is under thermodynamic control: If the product formed is the higher energy isomer (as in the dimerization of cyclopentadiene) it can be said for certain that the reaction proceeds under kinetic control (by the process of elimination since it can’t be under thermodynamic control). If the lowest energy isomer is formed, the reaction could be proceeding under thermodynamic OR kinetic control and it is not possible to say for certain without closer consideration of the reaction conditions and transition states. Alkene hydrogenation is usually carried out with a metal catalyst (e.g. palladium on charcoal) and hydrogen gas; under these conditions, reactions are usually under kinetic control (i.e. the products are not in equilibrium with the starting materials) and selective hydrogenation of alkenes is due to the steric effects that allow coordination of one alkene more easily than another. You should be careful when comparing MM2-calculated energies of different molecules – it is only fair to make direct comparison between isomers because for non-isomeric molecules, the energy scales are different.
--
Q2: You’re energy values are about right for both cases. It would have been nice for this problem to see a jmol of your lowest energy conformations. You say that you used different starting points to come across different conformations, but it would have been good to discuss the specific features that differ in each conformer you found. The rational for the selectivity of nucleophilic attack in both cases is correct. The mechanism of methylation in the first example is likely to proceed through an intermediate with the Grignard reagent coordinated to the carbonyl oxygen. The addition is likely to relieve the positive charge of the pyridinium which is the driving force of the reaction (i.e. as you have drawn it in the second example).
--
Q3. The energy of your “down” isomer is good and this is indeed the lowest energy conformation. In the case of the “up” isomer, the geometry of the double bond is incorrect. The hydrogen should be on the same side as the geminal dimethyl methylene bridge (the -CMe2- group). It is easy to get the wrong structure when trying to find different conformers so it is important to check after minimisation that nothing has fallen out of place, for things like alkene geometry such as this case or the configuration of stereocentres. As for comparison of the conformer energies, it would have been good to see a comparison of the various contributions to strain and a corresponding
rationalisation. You have correctly labelled this type of alkene as a hyperstable alkene, but the definition is not quite right. A hyperstable alkene is one for which strain is less than would be the case for the parent alkene that would be formed by hydrogenation; this phenomenon can occur for double bonds at bridgehead positions of medium rings.
--
Q4. The calculations and MOs look good, it was interesting to see how the MOs change for the various molecules under study – not many people thought to report these. The distribution of the HOMO does indeed show that the syn-alkene is the most reactive towards electrophiles. You are right to attribute the strength of the anti alkene to the pi-sigma* orbital, as a corollary the C-Cl bond is weakened by this interaction, which is which the C-Cl stretch is higher in energy for the anti-hydrogenated compound. Although the high electronegativity of oxygen means it is electron withdrawing by induction, it is also capable of donating electron density by resonance. Overall the resonance wins out in an enol and the alkene is in fact more nucleophilic than the alkene with hydrogen in place of the hydroxy group. The impact this has on this tricyclic system is that the alkene is more electron-rich, there is more donation into the sigma* and the C-Cl bond is weakened further.
--
MP. Your mini project choice is good, as it is a relatively constrained molecule and you have found a reaction which gives different diastereoisomers under different conditions. You have taken the right approach in analysing the most flexible part of the molecule at first and comparing all of the possible conformations (which goes to show how much more difficult it is to study systems with any flexibility) – for a complete picture you would have to calculate the amount of each conformation via the Boltzmann distribution and average out all of the properties accordingly. The errors you have seen are no doubt due to this flexibility as you suggest. Your catalytic cycle for the dihydroxylation is good the cyclic osmium intermediate is though to originate by cycloaddition onto the double bond – different models have been proposed for the diastereomeric transition states that are involved in the case with chiral ligands. As you say it is difficult to say which isomer you will get without a model to compare to. You have compared the calculated and experimental NMR shifts directly which is fine and they do seem to be a good fit. Another way to do this comparison would be to compare the calculated data for one isomer to the experimental data for both isomers and see if the actual experimental data fits better than the false data. This would show you whether you can distinguish between the isomers using computational methods.