[molpro-user] polari and quadrupole-quadrupole polarizabilities

[cong.wang]

Dear Glen,

 It seems the polarizability calculated in MOLPRO is
- <<A;B>> =   \sum_n  ( <0 | A |n > <n | B | 0 > +  <0 | B |n > <n | A
| 0 > ) / (E_n - E_0)

  The << .. >> notation follows J. Olsen and P. Jørgensen, JCP, 82, 3235

 While the common used Cartesian, quadrupole-quadrupole polarizability
adopted the Buckingham convention, Adv. Chem. Phys. 12, 107-143,
(1967) or Stone, the theory of intermolecular forces, P21

C_{abcd} =  1/3 \sum_n  ( <0 | Q_ab |n > <n | Q_cd | 0 > +  <0 | Q_cd
|n > <n | Q_ab | 0 > ) / (E_n - E_0)

That takes the factor of 1/3

best regards
Cong Wang

Ph. D. Student
Department of Chemistry
Laboratory for Instruction in Swedish
University of Helsinki
A.I. Virtanens plats 1
P.O. Box 55
FI-00014 University of Helsinki
FINLAND



On Fri, May 28, 2010 at 4:37 PM, Glen Jenness <grj3 at pitt.edu> wrote:
> Dear Molpro users,
> I recently used the polari keyword for both HF and MP2 to calculate
> quadrupole-quadrupole polarizabilities.  I opted to use water as a simple
> system with aug-cc-pVDZ basis set.
>
> However, I noticed some inconsistencies with the values from MOLPRO2009
> and other methods/literature values.  If we consider the Czz,zz term we
> get (in a.u.):
>
> MOLPRO:  24.13
> Bishop Method:  7.79
> Dehez et al.:  7.87
>
> For some references:
> Bishop method is outlined in Theor Chim Acta, 71, 247-253 (1987)
> Dehez et al. is literature value from J. Phys. Chem. A, 104, 1293-1303 (2000)
>
> It's clear that the MOLPRO values are off by about a factor of 3.
>
> Glen Jenness
>
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user
>