# CASPT2 calculation

Peter Knowles P.J.Knowles at bham.ac.uk
Mon Sep 17 08:24:05 BST 2001

```The principal problem with your input is that the active space that
you have chosen breaks D3h symmetry. You would get on better with
closed,3,1 in the rhf, and occ,5,2,1 in the CASSCF. Note that the
difference depends on the coordinate axes chosen by the program, and
that you can get some hints on what to do by looking at the output
produced by the integral program, where the constitution of the
default core and valence spaces is printed.

There is in fact code that prevents orbital rotations that break full
point-group symmetry, and this code is switched on automatically by
default. Its control is described in the manual (see
http://www.tc.bham.ac.uk/molpro/molpro2000.1/doc/manual/node212.html )
but would not be relevant for this problem where in the Hartree-Fock
you have already enforced broken symmetry.

Peter

At Fri, 14 Sep 2001 18:24:31 -0500,
Sahasranaman Mahalakshmi wrote:
>
>
>  Dear MOlpro-users,
>
>  I am trying to do a caspt2 calculation on ch3 with d3h
>  symmetry converted to c2v symmetry as non-abelian groups
>  cant be handled. I assume there are lot of rotations
> between orbitals as the energy doesnt seem to right. HOw
>  do I restrict orbital rotations ?
>
>  The input is as follows :
>
>  memory,15,m
>  r1=1.079
>  geometry={C;
>            H1,C,r1;
>            H2,C,r1,H1,120.0;
>            H3,C,r1,H1,120,H2,180.0}
>  basis=VDZ
>  int;
>  PRI,0;
>  rhf
>  closed,3,0,1,0;
>  wf,8,1,0
>  multi
>  accuracy,1.0d-10,energy
>  maxiter,40;
>  occ,5,1,2,0;
>  closed,2,0,0,0;
>  wf,8,1,0
>  canonical,2140.2,cidump
>  rs2;
>  option,ifdia=2;
>  ---

--
Prof. Peter J. Knowles
Email P.J.Knowles at bham.ac.uk  Phone +44-121-414-7472  Fax +44-121-414-7471
School of Chemical Sciences, University of Birmingham, Birmingham, B15 2TT, UK
WWW http://www.tc.bham.ac.uk/~peterk/

```