If Gaussian can do it in 20 mnts, I suppose many other
programs could do it as well, some of them even will be faster. For a 2D
surface, one can simply do a series of constrained geometry optimizations
(on the grid points)
and export the energy data to gnuplot. You can simply write a shell script
to automate the whole calculation (this way at-least one knows what is
really going on) As a free program, Gamess can do it! All other program
packages
you mentioned could also do it. About robustness: well, it is a different topic
in ccl and I better not comment on that ;-)
Sent to CCL by: "Peter Damian Jarowski" [p.d.jarowski _ surrey.ac.uk]
Dear CClers:
I was brought up with Gaussian as my main computational chemistry software
package during my Ph.
D. and postdoc. Now I am an independent and can no longer afford it. So
obviously I have been
investigating all kinds of freeware developed by academics. I want to do
something very simple that
Gaussian could get done in 20 minutes but have not found a stable implementation
of it in any
software package yet. I simply want to do a semi-empirical relaxed surface scan
of a dihedral and a
bond length to generate a 2D surface. I have installed NWChem/python, GAMESS,
and Firefly.
Apparently NWChem and GAMESS can not do this kind of thing and Firefly does not
have the right
implementation as of yet but will in version 8.0. I looked through the manuals
of Q-chem and mopac
and there appears to be no implementation there either. Turbomole might do it,
but it looks
complicated and its proprietary anyway.
Any suggestions as to some freeware that can do this kind of thing easily and
with stability. I am tired
of learning the nuances of each program's input file only to find that it
doesn't really do what I want to
do.
Thanks,
P.S. Gaussian is a robust program, but what I have found is that many of these
other programs seem
faster. Kudos. I appreciate the effort of the developers and I look forward to
continuing developments.