Well, I wanted to know are there ways to add and remove gd3 empirical dispersion when using ONIOM layers. For example, I'm using 2-layered system, and I use B3LYP functional for my high layer, and I want to add empirical dispersion to b3lyp functional, but my lower layer is calculated on semiempirical PM6 level of theory and "program" is adding empirical dispersion on that layer (output says: PM6D3), I want to know how I can remove empirical dispersion from that level of theory and add it just on a high layered functions.
Thank you
Marko
Gaussian Inc. recommend the following procedure:
Use the "External" keyword to call Gaussian as an external program that would run the ONIOM calculation. Something like the following:
oniom(external="mygau o g09 b3lyp/empiricaldispersion=gd3 6-31g\(d\)":pm6)
"mygau" is a perl script that is included in the Gaussian distribution ("g09/bsd/mygau"). It is just an example script that shows how the "External" keyword can be used to run another Gaussian job as the external program. You may write your own program or make a modified copy of the "mygau" script, one in which you would embed the "EmpiricalDispersion=GD3" so the B3LYP job is done as B3LYP-D3.
Right now I am doing some tests on energy of the system, in the output I get E(PM6D3) and E(B3LYP) which confuses me, but when I check the energy criteria, the empirical correction is added to b3lyp too even though the output does not says so. I am doing the same calculations with m062x and wb97xd functionals, and im not adding dispersion correction because they already have one, and there in the output says E(PM6) without D3, can I even compare these three?
Gaussian Inc. recommend the following procedure:
Use the "External" keyword to call Gaussian as an external program that would run the ONIOM calculation. Something like the following:
oniom(external="mygau o g09 b3lyp/empiricaldispersion=gd3 6-31g\(d\)":pm6)
"mygau" is a perl script that is included in the Gaussian distribution ("g09/bsd/mygau"). It is just an example script that shows how the "External" keyword can be used to run another Gaussian job as the external program. You may write your own program or make a modified copy of the "mygau" script, one in which you would embed the "EmpiricalDispersion=GD3" so the B3LYP job is done as B3LYP-D3.
Hi Marko,
did you find a solution about dispersion in a ONIOM calculation?
Let me know if you have any bullets about
I guess you are talking about basis set superposition error, this can be done in Gaussian with keyword BSSE
- Badges
- Science topic