FNO Approach for the Non Hartree-Fock Orbital

Questions
#1

I would like to use the Non Hartree-Fock Orbital method (e.g. DFT) and I also would like to investigate the natural orbital occupation number (NOON) in this case.
Take a glance on FNO, it seems like it assumes Brillouin condition. However as I know, DFT type orbital doesn’t satisfy the Brillouin condition in general. Therefore, I would like to make sure:
Is it valid (or Is there a way) to use FNO with DFT reference orbital?

Below is my script:
(I use CCMAN because it would print out NOON)
(or Could Q-Chem print out 1-RDM such that I can obtain NOON by myself?)

$rem
METHOD           BLYP
BASIS            cc-pVTZ
$end

@@@

$rem
SCF_GUESS        read
MAX_SCF_CYCLES   0
BASIS            cc-pVTZ
METHOD           CCSD(T)
N_FROZEN_CORE    0
CC_FNO_THRESH    9900
CC_FNO_USEPOP    1
CCMAN2           FALSE
$end

Thanks in advance.

#2

I don’t think there’s a user-facing way to perform natural orbital analysis for arbitrary densities in Q-Chem. Each correlated method has it’s own implementation of NO analysis. If you specify which method you’d like to look at I will be able to provide more concrete instructions for either evaluating NOs or obtaining the correlated density.

That said, natural occupations for any SCF density are 1 for occupied MOs and 0 for virtual MOs by construction P=CoccCoccT.

#3

This is correct. The Kohn-Sham orbitals are natural orbitals, with the occupations of 1 or 0, just like the Hartree-Fock orbitals are. However, you can use Kohn-Sham orbitals in a correlated calculations and compute natural orbitals for such non-standard reference.