Dipole moment

Dear All,

I would like to ask for your help regarding the computation of dipole moments.
Since dipole moments are orientation dependent (and for ionic species also origin dependent), I wanted to double check which coordinate frame the computed dipole components are expressed in. For this reason, I started comparing computed dipole values with those obtained with Gaussian (using same method, basis, symmetry), and was surprised to see that although the energies agree very well, the dipole moments are completely different for most of my test calculations.
For example, in the case of HeH; Energies: -3.4004227 au vs. -3.400422734555 au, Dipole: 0.0049 Debye vs. -10.00010396 Debye.
Do you have any suggestions on what I could be doing incorrectly, or what the issue could be?
Attached you can find Gaussian and MRCC input and output files for HeH.

With warm regards,
Tamas

Dear Tamás,
What is printed at the end of the mrcc output is the unrelaxed electronic contribution to the dipole moment. To get the total dipole moment, first you should add the nuclear contribution, but the dipole moment will still slightly differ from that obtained by Gaussian (unless you use full CI) since the orbital relaxation is missing. I know that the output is misleading and I have changed it in the development version. (Please also note that you are considering the wrong number in the Gaussian output. It is a polarizability calculation and I guess the last dipole moment in the output is calculated with some electric field. I think you should consider the first one, 0.0001 Debye).

Dear Professor Kállay,

thank you very much for the clarification and help!

With warm regards,
Tamás

Is it possible, that at the end of the mrcc output, (-1) x the unrelaxed electronic contribution to the dipole moment is shown?
The reason I ask, is because in order to get a dipole moment agreeing in absoulte value (but opposite in sign) with the Gaussian dipole, I needed to subtract the dipole contribution of the nuclei.

You are right, it should be multiplied by -1. And I have chacked, the core contribution is also not added, so if you have frozen core electrons, it won't be correct.
This was made for the Cfour interface, where the integrals are stored with opposite sign, the core contribution is stored, etc, but it has not been tested in standalone mode. I have fixed these things now.

Dear Professor Kállay,

thank you for the confirmation and the additional information!

With warm regards,
Tamás