Exercise: the Diels-Alder reaction, part 2


In addition to the 'one-dimensional' calculations earlier in this list, we offer the possibility to do a two-dimensional grid calculation. This is an experiment for us too, to explore the scope and limitations of such an exercise in a web environment.
The procedure is not so different. Use the editor to add substituents to either combination of reactants. The aldehyde can be placed in two orientations, denoted by a (+) and (-), which gives the possibility to create either the s-cis or s-trans vinyl aldehyde. This is a matter of trial and error; we cannot predict which H you will select for substitution and hence cannot tell whether (+) corresponds to s-cis or the other way around.
Generally speaking, the user is responsible for creating sensible and meaningful structures. It is probably very easy to construct compounds that will create fatal errors during the calculation.

A grid calculation takes more time than a reaction path of course. A 7x7 grid will be finished in about two minutes though, which is to our taste just acceptable for being called 'interactive'.

The result is presented as a two-dimensional clickable plot, next to a Chime display. In the process the data have been analyzed to some extent: the plot shows iso-energy contours, an estimated saddle point and a line showing the direction of the reaction coordinate at the saddle point.
Below the plot the saddle point data are shown in editable fields, which can be used to do a subsequent calculation. The default values will place the theoretical saddle point at the center of the grid. Decreasing the range is equivalent to zooming in on this point.

If a point sufficiently close to the transition state (TS) is obtained, the energy and the geometry could be used for comparison with other orientations of the substituents, in order to study the effect on product formation.
Maybe one would like to show the complete reaction path from starting material to product, and/or to look at the orbitals at various stages of the reaction.
That first has been accomplished: an automatic procedure that describes the complete reaction path. A link under the grid output will take you to Paderborn, where this feature has been implemented.
See the 'Note on IRC calculations'.

Note: We use MOPAC AM1 calculations just because they produce a 'reasonable' result within an acceptable 'interactive' time.
However, the AM1 Diels-Alder transition state bond lenghts are not 'correct'; ab initio calculations arrive at higher values for the TS, close to 2.3 Å for the unsubstituted reactants. As long as we use the results for comparison of similar compounds, this is not a serious problem.
It should be noted that even with sophisticated ab initio methods, it is hard to reproduce say endo/exo selectivity.

The SN2 reaction is added as another example of a grid calculation, although there is not so much to vary here. The default input file concerns chlorine/bromine substitution in methane. The Br and Cl are forced on a straight line, which is of course not correct for asymmetrically substituted systems. This is done to prevent the anions from abstracting protons from added methyl groups.
Unfortunately, it is hard to study SN2 vs. SN1 mechanisms by computation, as this requires inclusion of solvent effects. This would increase the computation time to such an extent that it is not practical yet in an interactive web environment.
We consider including some pre-calculated results in this page.