Is it possible to plot the pDOS or visualize the molecular orbitals from GAMESS outputs?

Question

I'm using the GAMESS software and want to try to plot the pDOS for my systems and visualize the molecular orbitals. I was thinking I could get the data from one of the NPRINT commands, although which one is not so clear. And then once I have the data (if possible), is there a straight-forward way to plot/visualize it?

Answer 1

Yes, you can get what you want. If you have a Gaussian .fch file, you can visualize MOs using GaussView/Multiwfn/... any GUI you like. And you can plot PDOS by loading the .fch file in Multiwfn(http://sobereva.com/482). Here is one solution on how to generating a .fch file, or exporting MOs into it:

Let us take the water(H2O) molecule as an example. Firstly, we perform a one-second job using Gaussian. The input file h2o.gjf is like

%mem=1GB
%nprocshared=1
%chk=h2o.chk
#p RHF/cc-pVTZ nosymm int=nobasistransform guess(only,save)

title

0 1
 O                  1.39830507    0.29275809    0.00000000
 H                  2.35830507    0.29275809    0.00000000
 H                  1.07785048    1.19769392    0.00000000

This job takes less than 1 second since the keyword guess(only) tells Gaussian program not to perform the SCF computation, but only generate the SCF initial guess.

Then you obtain the file h2o.chk. Run

fch2inp h2o.chk

or equivalently,

formchk h2o.chk h2o.fch
fch2inp h2o.fch

you will get the GAMESS input file h2o.inp, in which the charge, spin, Cartesian coordinates, basis sets and initial guess MOs of Gaussian are written. There is no need to change them. You can modify keywords within this file and perform computations you want.

After your GAMESS job is accomplished, run

dat2fch h2o.dat h2o.fch

this will export MOs from h2o.dat into h2o.fch. Now you obtain a .fch file which includes MOs from GAMESS.

(Here I assume that you perform single-point calculations in GAMESS, i.e. the geometry is not changed, so the same h2o.fch file is used.)

fch2inp and dat2fch are two utilities in MOKIT, free and open source. You can simply install MOKIT via conda or download Linux pre-built executables.