Background Charges and dipoles

Hi,
I am new to MRCC, and I am still learning about its capabilities. I am interested in analyzing excited states of relatively large systems (160 electrons, 350 basis set functions). My current interest is the analysis of the effect of external fields on those states. What attracted me of MRCC in the first place, was the availability of the reduced cost ADC(2) methods.
I was wondering if there is a way to introduce background point charges, and or point dipoles to the initial SCF calculations, to be used subsequently in the reduced cost ADC(2).
A first quick look at the manual show the following keywords:
- epert: This seems to allow introducing the effect of external dipoles. Can these dipoles be located at specific user locations, i.e, give the position and magnitude/orientation of the dipole, for several dipoles?
- Point charges seems to be included only via the qmmm=amber keyword, but I do not have Amber installed. Is there any other way to add background charges? Thanks!

You can add external point charges. You should just set qmmm=amber, even if you do not use Amber, and give the point charges at the end of the input file, see the example at the description of keyword qmmm in the manual.
The epert keyword introduces external electric fields. It is not sufficient for your purpose. External dipole moments can be modelled using the corresponding point charges.

Thanks Prof. Kallay for your answer.
May I ask you, please, how to model dipole moments using point charges? I am attempting to include the atomic dipole moments provided by some force fields as the effect of the surroundings on a QM system.
Thanks

A dipole moment can be defined by two point charges of opposite polarity and their relative displacement vector. I am not familiar with the force filed methods, but probably you can extract this information from the force field.

Dear Edums,
calculations with 350 basis functions do not require the reduced-cost approximation. These calculations take less than half an hour on a simple laptop with a few GB of main memory and a 4-core processor.

If you still want to use the approximation, please note that the default thresholds for the cutoff parameters are primarily defined for the aug-cc-pVTZ basis set. In the case of greatly different (smaller or larger) basis sets, the utilization of the numerical results requires more caution.

Best regards,
Dávid

Thanks very much David. After reading your post, I run an ADC(2) calculation without the reduced-cost approximation in my system, and I found it run much faster than I expected. Congratulations for a great code!

Now I am considering to apply ADC(2) to several configurations extracted from a MD simulation, but I am concerned about the requirement of using the aug-cc-pVTZ basis set. Does the redcost approximation works with other basis sets, or is only optimized to run with this particular basis set. What can be expected, for example, if used with a smaller basis set, for example, def2-svp? Thanks!