VPT2 missing intensities

[Ruzica Markovic]

Dear Molpro Community,

I am trying to compute the VPT2 transition energies and intensities at the CCSD(T)-F12a level of theory, for the open-shell species (so I am using the unrestricted formalism). I would like to have both the energy and intensity information at the CCSD(T)-F12a level of theory (or CCSD(T) if F12a is not available).  I have the following input according to the manual, for H2O-Ar radical cation system:

gthresh,energy=1d-8,orbital=1d-7,coeff=1d-7
basis=vdz-f12
geometry={
 O           0.0345875283        1.9196946195        0.0000000000
 H          -0.0121048368        0.8941900098        0.0000000000
 H          -0.8874690875        2.2961640152        0.0000000000
 Ar          0.0350283959       -1.0376366445        0.0000000000
}

set,charge=+1

basis=vdz
uhf

basis=vdz-f12
uhf
uccsd(t)-f12a
optg,gradient=1.d-8,method=diis

frequencies,symm=auto
{frequencies,analytic,sym=auto;save;thermo;print,thermo,Hessian}

label1                   !the label1 block defines the electronic structure refernce when generating the PES
{uhf,
start,atden}
{uccsd(t)-f12a,
 cphf,1}

{xsurf,start1D=label1,sym=auto           !(3) generate a QFF
 vtaylor,type=qff,
 intensity,dipole=2,
 disk,where=home,dump='combi-new.pot'}
{vibstate,combi=2,ubound=8000}
poly,type=qff,vam=0                      !(4) transform the PES to polynomials
vpt2

However, in the output I get all 0.00 values for intensities

(example: 

 Mode         Harmonic     Anharmonic     Intensity

1^1 A'           215.56          204.63               0.00 )

In the poly program I also get the following:

 Dipole 1D    0.00E+00  0.00E+00  0.00E+00  0.00E+00     0.0    1
        2D    0.00E+00  0.00E+00  0.00E+00  0.00E+00     0.0    1  2
        3D    0.00E+00  0.00E+00  0.00E+00  0.00E+00     0.0    1  2  3

Does that mean there is no way to get the intensities at this level of theory or am I just missing some input information?

Thanks in advance!

Best,

Ruzica

[rauhut]

Dear Ruzica,

your input requests the calculation of UCCSD(T)-F12a dipole moments, which have not been implemented into Molpro. Note, that the XSURF program requests the availability of analytical dipole moments. Once you do not need high accuracy, you may simply use the UHF dipole moments, which you have computed anyway. For that, your input needs to be slightly modified, i.e.

label1                   
{uhf,
start,atden}
uccsd(t)-f12a,

{xsurf,start1D=label1,sym=auto           
 vtaylor,type=qff,
 intensity,dipole=1,
 disk,where=home,dump='combi-new.pot'}
{vibstate,combi=2,ubound=8000}
poly,type=qff,vam=0                      
vpt2

Best,

    Guntram