Notes while viewing the model of CH3CHSCH2

(Advice: open two browser windows alongside each other. Go to this page in one, and the 'Chime model of methyl thiirane' in the other.)

I was interested in using this molecule as a source of sulfur atoms because it will lose sulfur to leave propene

I could calculate energies for starting materials and products, for several possible sulfur acceptors, and successfully predict which would be the best to try experimentally
Unfortunately, there was an alternative mode of reaction: a nucleophilic iodine atom in my reactant could open the three-membered ring, breaking one of the C-S bonds. To help assign the product to a particular isomer, I needed to predict which C would be attacked.

(Do the following using the 'Chime model of methyl thiirane')

(In your other window, go back to the index, and open the 'Chime total electron density..' model)

This takes a little while because a 2 Mb data file of electron density information has to be downloaded and uncompressed
The picture shows a contour of the total electron density
In three dimensions, a contour is a surface
Make it transparent by

In the .spt script file, I chose the value of the electron density for which the contour was to be plotted, by trial and error
Too low a value gives a bland, far-out surface in which detail is lost; too high a value risks missing out important bits, because there is not that much density in them
Turn the model to look at the sulfur, with the plane of the ring behind it
You can just about imagine the bulges of the lone pairs on sulfur
You can clearly see that there is electron density where the bonds are supposed to be: this may be useful to know in less obvious molecules
Other than that, an electron density picture looks like a space-filling model with VdW radii, which can be got without doing electronic calculations
Test this by

Where a nucleophile attacks is partly controlled by where there is a positive electrostatic potential to attract it
One way to display this is to map electrostatic potential energy onto a total electron density surface, using colours to represent different levels of potential

These are potentials calculated by Chime, from electronegativities, not from the electronic structure calculation
The white and blue end of the scale is negative potential; the red end is positive potential
Turn the model so that the sulfur is away from you
You can see in to a carbon with positive potential, which you can guess is where a nucleophile will attack

A nucleophile must attack an empty MO
It will probably attack the Lowest Unoccupied Molecular Orbital
In the MO modelling methods, the distribution of individual MOs is also part of the model

(In your other window, go back to the index, and open the 'Chime LUMO for..' model)

This is the LUMO superimposed on the space-filling model
All lobes are colored the same (red) and do not show the sign of the wavefunction. This is only a limitation of the Chime program: the data is present in the model.
This is clearly a very antibonding orbital because it has several antibonding nodes
Toggle the surface to transparent
Find an antibonding nodal surface cutting through the S-C bonds
Display sticks
Find another node cutting through sulfur and cutting the C-C ring bond
Molecular orbitals are delocalised over all or most of the atoms
The biggest lobes come from a p orbital on sulfur. We can show that our nucleophile will not attack there because that is the end of the molecule with negative potential
Map the electrostatic potential, with Rasmol colouring, onto the LUMO surface, in the same way that you did using the electron density surface
Of the warm colour, attractive to a nucleophile, you can see that biggest accessible LUMO lobe is underneath the ring CH₂ carbon
This finding guided my further experiments, and was eventually confirmed by consistent assignment of the carbon and proton NMR spectra of the products