[molpro-user] rs3 opt

[Toru Shiozaki]

Dear Benjamin,
this seems to me a problem in the code: when one computes the single-state RS3 for the nth excited state (i.e., using "state,1,n"), the RS3 energy seems to be set to ENERGY(n), instead of ENERGY(1) as in RS2. That is why optg picks the RS2 energy.
I am going to fix this problem and patch it soon; but for the time being, you could run your calculation with the following workaround.

! this is just a workaround
r=2
geometry={c;h,c,r;}
{hf;}
{multi;state,2}
{rs3;state,1,2}
{optg,numerical,variable=energy(2)}

Please note that this won't work once the code is fixed!
Regards,
Toru


> Date: Mon, 20 Dec 2010 15:14:45 -0500
> From: "Mintz, Benjamin J." <bm0037 at gmail.com>
> Subject: [molpro-user] rs3 opt
> To: molpro-user at molpro.net
> Message-ID: <95E47A08-536D-471F-8052-9340F0C0C29D at gmail.com>
> Content-Type: text/plain; charset=us-ascii
> 
> Dear Users,
> 
> I am currently trying to perform a CASPT3 optimization, and I am running into a problem.  Specifically, I am trying to optimize the first excited state for my molecule.  The optimization converges, but the output looks like the optimization is utilizing the RS2 energies and not the RS3 energies.  I pasted part of the output below, which lists the RSPT2 and RSPT3 energies and the energies from the optimization.  I also pasted part of my input.  Why is the optimization not optimizing the RS3 energies?  Any help would be appreciate.  Let me know if I need to provide any more information. 
> 
> {hf;wf,35,4,1;orbprint,4}
> {multi;occ,9,3,7,2;frozen,3,0,2,0;closed,8,0,6,0;wf,35,4,1;orbprint,4
> select
> con,2,2,2,0,0,1,0
> con,0,2,2,0,2,1,0
> con,2,2,0,0,0,2,1
> con,0,2,0,0,2,2,1
> }
> {rs3;state,1,2}
> {optg,numerical;
> inactive,d1;
> inactive,d2;
> inactive,d3;
> inactive,d4}
> 
> 
> 
> 
> Reference energy                    -192.00677670
> Nuclear energy                       130.22929885
> Kinetic energy                       192.49019540
> One electron energy                 -515.10328409
> Two electron energy                  192.28496840
> Virial quotient                       -1.00051338
> Correlation energy                    -0.58224014
> !RSPT2 STATE 2.4 Energy             -192.589016841281
> 
> !RSPT2 STATE 2.4 Dipole moment         0.00000000     0.00000000     0.56683718
> Dipole moment /Debye                   0.00000000     0.00000000     1.44066205
> 
> !RSPT expec            <2.4|H|2.4>  -192.490833418072
> 
> Correlation energy                    -0.60081552
> !RSPT3 STATE 1.4 Energy             -192.607592225598
> 
> .
> .
> .
> 
> END OF GEOMETRY OPTIMIZATION.    TOTAL CPU:     53836.6 SEC
> 
> ITER.   ENERGY(OLD)    ENERGY(NEW)      DE          GRADMAX     GRADNORM    GRADRMS     STEPMAX     STEPLEN     STEPRMS
>   1  -192.58901684  -192.59615035    -0.00713351  0.12323135  0.12793186  0.01421465  0.08500958  0.09811296  0.01090144
>   2  -192.59615035  -192.59670839    -0.00055804  0.02893302  0.03046462  0.00338496  0.02788328  0.03482585  0.00386954
>   3  -192.59670839  -192.59673641    -0.00002801  0.00508719  0.00606747  0.00067416  0.00630698  0.00935194  0.00103910
>   4  -192.59673641  -192.59673813    -0.00000173  0.00063479  0.00121156  0.00013462  0.00214935  0.00300464  0.00033385
>   5  -192.59673813  -192.59673824    -0.00000011  0.00021685  0.00036503  0.00004056  0.00049652  0.00066890  0.00007432
> 
> 
> 
> Thanks,
> Benjamin J. Mintz
> Research Associate, Postdoctoral
> Oak Ridge Leadership Computing Facility (OLCF)
> Oak Ridge National Laboratory
> Phone: (865) 574-6639
> Email: mintzbj at ornl.gov
> Gmail: bm0037 at gmail.com
>