From owner-chemistry@ccl.net Fri Jul  6 09:27:00 2007
From: "Brian Salter-Duke brian.james.duke+/-gmail.com" <owner-chemistry*server.ccl.net>
To: CCL
Subject: CCL:G: CASSCF not giving ground state even though orbitals are correct
Message-Id: <-34678-070706062203-23799-Lm5wtjObSnVrqOX+nYP6fA*server.ccl.net>
X-Original-From: "Brian Salter-Duke" <brian.james.duke . gmail.com>
Content-Disposition: inline
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset=UTF-8; format=flowed
Date: Fri, 6 Jul 2007 19:26:35 +1000
MIME-Version: 1.0


Sent to CCL by: "Brian Salter-Duke" [brian.james.duke%x%gmail.com]
I have been doing a simple CASSCF(6,8)/STO-nG calculation on the
triplet groundstate of B2. I wanted to confirm that Gaussian agreed
with GAMESS(US) and with the CASVB(6,8) method in the VB program
VB2000. Gaussian kept giving me what was clearly an excited state.
Finally I added FullDiag to the CASSCF command and it worked but the
time in one case went from 3 secs to 3 mins. In every case I was doing
a double job - ROHF followed by CASSCF and the ROHF had no problem
with the correct orbitals - 2-sigma^2, 4-sigma^2, pi-x^1, pi-y^1 for
the 6 electrons. Clearly without the FullDiag (which is default for 6
or less orbitals but not 8), the initial vector was wrong. Has anyone
found this before and is there any solution to it? This occurred with
G98 and G03. It sort of confirmed my preference for GAMESS for CASSCF,
but I would like to understand the Gaussian use as well.

Brian.

-- 
Brian Salter-Duke (aka Brian Duke)
Brian.James.Duke#,#gmail.com