UT theoretical chemistry code forum

Hi. I am really glad to see further development in Bader's topological analysis. And any work that makes the analysis faster is a miracle!

Anyhow, I would like to know if you could increase capabilities somewhat. I published a couple of papers characterizing groups by their AIM properties (Arturo and Knox, Journal of Molecular Structure: THEOCHEM Volume 770, Issues 1-3, 29 September 2006, Pages 31-44; Arturo and Knox, Journal of Molecular Structure: THEOCHEM Volume 857, Issues 1-3, 30 May 2008, Pages 78-88) Besides volume and charge, other important properties include moments of electron density, exposed surface area, and average distance of nucleus to the exposed surface area. The integrations I did were to the classical 0.001 au isodensity surface. I developed a molecular force field in my dissertation and continue to find applications of these properties.

So I ask you this: Can you compute the other properties with your code? Can you compute the properties to the 0.001 au isodensity surface? Would you do it?

Thanks. Again, this work is an incredible help, and I wish you the best in your development of the method.

Steve

These are some very good suggestions. The reason our group is using the Bader analysis is primarily to decompose charges, density of states, and energy into local contributions per atom. We do not have any expertise in the AIM theory and this is the only reason that we have not calculated some of the properties that you mention which are of interest to that community.

I would be very happy to see these additions in the code, and I think there are a few ways this could be done.

1) We can add simple properties, such as the average distance from the nucleus to the surface, if they are valuable to someone. Some guys in the group just set up a system for making requests and reporting bugs:
http://theory.cm.utexas.edu/redmine/projects/bader

2) If you want to add a feature and have an algorithm, you can certainly do that. The code is open source and as long as new code works and basically conforms to the rest of the code, we will add the authors and the relevant references and make the improved version available.

3) A third option is to collaborate. It would be much easier for us to work with someone who can validate the output of an algorithm, than to implement something ourselves from scratch. If something is implemented in another code, for example, we could help translate it and compare results between the codes.

I guess the bottom line is that it would be great to have the code do more, but we would need some motivation or help to make it happen. Our current efforts are to improve the precision of the charge partitioning and not on implemented new features (except for critical points, which is on our list).