I’m trying to finish up some of the work on p-dimethyl aminocinnamaldehyde (DMAC) that started several years ago. My goal is to refine the theoretical calculations that Diane performed last summer by including Ag atoms[1]. Hopefully the calculations will help us understand how DMAC adsorbs at the Ag surface, and maybe support the vibrational assignments that Diane made.

I’m using Gaussian 03W for the calculations, and GaussView for building the input file. Initially, I thought I could just “add” the Ag atoms to the molecule and click “submit”. No. All that happened when I did that was a series of crashes (I kept trying – for some reason I thought that if I kept submitting the job, Gaussian would eventually give up and just do the work).

I contacted my quantum-mechanic colleague (who is on sabbatical in China now), and he suggested that for the Ag I might have to use pseudopotentials – it just has too many electrons. Not wanting to believe him, I did as much digging in the literature as I could about these types of calculations. Damn it, he knew what he was talking about.

Since lots of published authors supported Mr. quantum-mechanic, it was obvious that I’d have to learn how to do these types of calculations. I read the Gaussian documentation, and could find things about the PSEUDO keyword, that told me if I used it, Gaussian would look to the end of the input file for the basis sets to use. Poking around on the internets, I found some more references, figured out how to edit my input file, and now have Gaussian chugging away. Hopefully, it will work! (Looks like this is fairly computationally expensive. I submitted the job yesterday afternoon, and it is still running today. Thankfully).

[1] See, for example, Chis, Venter, Leopold and Cozar; Vibrational Spectroscopy 48, 210-214 (2008).