Unexpected behaviour: core orbitals are not frozen during MP2/MP3 step in ADC calculations

Questions
#1

Hello,

I encountered that core orbitals are included (not frozen) by default during the MP2/MP3 step in ADC calculations (in particular ADC(3)), which is strange, as one would expect them frozen by default, similarly to EOM-CCSD calculations. And if they are not frozen at the MP2/MP3 step they are apperently not frozen at the ADC step too. Has this been fixed in Q-Chem 6.3.1? Also, it would be useful if information about frozen orbitals is printed out, similarly to EOM-CCSD calculations.

#2

I don’t know why you would “expect” that, but you can control it with N_FROZEN_CORE in $rem.

#3

Many thanks for your reply! One would expect for MP2/MP3 ADC etc that by default

N_FROZEN_CORE = FC

which is at the moment not the case. It is stated explicitly in the online manual that for a post-HF method, which MP2/MP3/ADC belong to, N_FROZEN_CORE = FC. Here is a copy of the text from the Q-Chem web-page:

N_FROZEN_CORE
Sets the number of frozen core orbitals in a post-Hartree–Fock calculation.
TYPE:
INTEGER
DEFAULT:
FC
OPTIONS:
FC Frozen Core approximation (all core orbitals frozen). n Freeze n core orbitals (if set to 0, all electrons will be active).
RECOMMENDATION:
Correlated calculations calculations are more efficient with frozen core orbitals. Use default if possible.

It would be useful to have information about frozen orbitals in the output of ADC calculation before the MP2/MP3 step, I mean the following (as an example):
…
Point group: Cs (2 irreducible representations).

                       A'   A"    All

All molecular orbitals:

  • Alpha 42 15 57
  • Beta 42 15 57

Alpha orbitals:

  • Frozen occupied 0 0 0
  • Active occupied 10 2 12
  • Active virtual 32 13 45
  • Frozen virtual 0 0 0

Beta orbitals:

  • Frozen occupied 0 0 0
  • Active occupied 10 2 12
  • Active virtual 32 13 45
  • Frozen virtual 0 0 0

…

which is the case in the output of CCSD, MP3, EOM-CCSD calculations, but it is somehow missing in the MP2 and ADC calculations.

#4

You can send a feature request to Q-Chem.