I have no clue what is doing this. I am running DFT/B3LYP/GenECP, with 6-31G** for the main group elements and SDD for the Ru. I tried it before with LanL2DZ for Ru but this same thing happened, and I thought it was just an issue with the basis set. The ligands just get pulled all the way into the center; the last calculation stopped when the carbon of the CO ligand ended up at the same position at the Ru. Super weird.
I do experimental bond lengths from crystal structure data, so I can lock the ligands in place, but I would rather not. Anybody seen this before or know what might be the issue? I run a ton of Au stuff with LanL2DZ and this has never happened.
Edit 4/20/2017: It appears you were all correct, thank you. The definition I was using for Au had the ECP appended, but when I tried to roll my own for Ru I did not include the ECP definition at the end. Thank you all for your answers!
If you have Single crystal x-ray structure start optimization with it. You can change bond-length(bond angle ) like 0.1A(1-2 degree). If you dont have the single crystal X-ray structure then try to choose the very good initial guess(bond length and bond angle) for optimization. I have face many problem like this and a good initial guess of geometry can save a lot of time
It looks like a problem with improper definition of the basis set for Ru. Are you sure that you have the correct command line in Gaussian? E.g. for using the standard LANL2DZ basis set you should have something like:
# b3lyp/gen pseudo=read
and then at the end of file:
Ru 0
LANL2DZ
****
Ru 0
LANL2DZ
becuse first you specify the basis set for Ru and then the ECPs for Ru. If you omit the second definition you won't get any errors, but also you won't have any ECPs and the structure will converge to something strange, with other parts of moelcule overlapping the Ru atom.
Dear Max,
I totally agree with Bartosz.
In addition to what was mentioned by Bartosz, please read the publications in the following links:
http://joaquinbarroso.com/2011/11/02/gen-keyword-gaussian/
http://www.gaussian.com/g_tech/g_ur/k_pseudo.htm
Rafik
- Badges
- Science topic