[molpro-user] Multipole Moments via GEXPEC and CCSD(T)

[Jacek Antoni Klos]

There is no ccsd(t) densities to get gexpec values at this level in Molpro. You can use finite field method  for ccsd(t) to calculate
dipole or higher moments at this level of correlation. Here is molpro manual section on finite field calcs:
 http://www.molpro.net/info/current/doc/manual/node457.html

Best
Jacek

On Jun 27, 2012, at 4:58 AM, Christian Cioce wrote:

Hi, I'm having trouble computing multipole moments at the CCSD(T) level of theory with the GEXPEC keyword. Using a similar sample input for H2O (within the Molpro User's Manual), I have the followingproperly working input file:

geometry={
2
N2 Monomer
N          0.000000      0.000000     -0.549000
N          0.000000      0.000000      0.549000
}
geomtyp=xyz

gexpec,dm,sm,qm                       !compute dipole and quadrupole moments

$methods=[hf,multi,ci]                !do hf, casscf, mrci
basis=AV5Z
do i=1,#methods                       !loop over methods
$methods(i)                           !run energy calculation
e(i)=energy
dip(i)=dmz                            !save dipole moment in variable dip
quadxx(i)=qmxx                        !save quadrupole momemts
quadyy(i)=qmyy
quadzz(i)=qmzz
smxx(i)=xx                            !save second momemts
smyy(i)=yy
smzz(i)=zz
enddo
table,methods,dip,smxx,smyy,smzz        !print table of first and second moments
table,methods,e,quadxx,quadyy,quadzz    !print table of quadrupole moments

This input file renders appropriate (normal) output:

 METHODS   DIP      SMXX          SMYY          SMZZ
 HF        0.0   -7.71965855   -7.71965855   -8.64863534
 MULTI     0.0   -7.52595656   -7.52595656   -8.71606123
 CI        0.0   -7.59202157   -7.59202186   -8.72499509

 METHODS        E           QUADXX       QUADYY        QUADZZ
 HF        -108.9928938   0.46448839   0.46448839   -0.92897679
 MULTI     -109.1415178   0.59505234   0.59505234   -1.19010467
 CI        -109.3997302   0.56648690   0.56648647   -1.13297337


But when I then alter the methods in the original input file to calculate the moments following a CC calculation, the results aren't as expected...

Truncated input file:

$methods=[hf,mp2,ccsd,ccsd(t)]

and the resulting output:

 METHODS   DIP      SMXX          SMYY          SMZZ
 HF        0.0   -7.71965855   -7.71965855   -8.64863534
 MP2       0.0   -7.71965855   -7.71965855   -8.64863534
 CCSD      0.0   -7.71965855   -7.71965855   -8.64863534
 CCSD(T)   0.0   -7.71965855   -7.71965855   -8.64863534


 METHODS        E           QUADXX       QUADYY        QUADZZ
 HF        -108.9928938   0.46448839   0.46448839   -0.92897679
 MP2       -109.4037581   0.46448839   0.46448839   -0.92897679
 CCSD      -109.3945766   0.46448839   0.46448839   -0.92897679
 CCSD(T)   -109.4155011   0.46448839   0.46448839   -0.92897679


It seems as though the CC routines aren't generating a new wavefunction or dumps for the properties routine to use and so when requested, the multipoles are using the original SCF wavefunction to calculate the multipoles, but I'm not sure of that. Am I doing something wrong in my input file? All I want to do is calculate the quadrupole tensor for the N2 monomer at the CCSD(T) level of theory.

Thanks in advance for your help!

Best,
Chris
_______________________________________________
Molpro-user mailing list
Molpro-user at molpro.net<mailto:Molpro-user at molpro.net>
http://www.molpro.net/mailman/listinfo/molpro-user




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20120627/cf46f1d6/attachment.html>