We are using gaussian G-16 to calculate molecular properties. When I specigy in the rout line def2qzvpp ( #P B3LYP/def2qzvpp opt output=pickett) with RU (Ruthenium atom in the atom list) -- Does the Gaussian do an ALL ELECTRON calculation -OR- use a core potential (ECP) for the Ru atom?
The def2 basis sets are ECPs, they are defined for H-Rn, So if you choose this theory level for your calculation, Ru core electrons will be described by an ECP. See: https://doi.org/10.1039/B508541A
I'd advise you to consider relativistic effects for your calculation, since Z>36 for Ru (a good rule of thumb for relativistic effects in DFT). Think about employing ZORA or DKH.
Lucas, Many thanks for clearing this up - did not get answer from Gaussian website... Steve K
Stephen Kukolich Now I better understand your question, and yes, Gaussian Inc. isn't very clear about the way they implement basis sets/functionals. Usually I read the paper that proposes the theory level and trust that it was implemented in the same way. Hard to know what actually goes behind the code, but that is common when using proprietary software.
Lucas - THANKS AGAIN! I did get one calc to run using ano-rcc-vdzp.Ru-H-C-O basis . Tried one --DK basis set (aug-ccpVTzxx - crashed = will look for DKH
Still have not succeeded trying different basis sets for different atoms using googled recs. Lots of stuff not spelled out in Gaussian help pages, but Fernendo Clemente and Doug Fox have been very helpful.
For gaussian, when I want to use mixed basis sets or basis sets that are not implemented in the code I usually use the GENECP keyword (https://gaussian.com/gen/) and I specify the basis sets after the coordinates in the input file. You can find the syntax in the example part of the website.
You can get the basis sets not implemented using the BSE (basis set exchange) https://www.basissetexchange.org/. Just select the atoms to include, the basis set and choose the format as "gaussian" in the box bellow the periodic table. You may then be able to click "get basis set" and copy the correct data to paste in the input file. I'll attach prints for reference.
Also, I'm attaching a simple water input where I imported the aug-cc-pV(T+d)Z for H and O from the BSE website, so you can check the correct syntax.
Hope it helps for you to figure out an option to deal with this.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics