Regarding Initial Guess

Questions
#1

Dear Q-Chem Developers,

How to generate the initial guess without performing the SCF iterations for an unrestricted DFT calculation so that the resulting alphas and betas comes out to be the same, however, the occupations of alphas and betas can differ.

Thanks

#2

Not clear what you are looking for, but you can run a prior restricted SCF calculation using the (N-1)- or (N+1)-electron system. Then, you can use these orbitals as guess for your N-electron calculation using “SCF_GUESS = read”. You may consider $occupied section along with “MOM_START” and “MOM_METHOD” REM variables, if you want a specific SCF solution. I don’t think that “UNRESTRICTED = true” calculation can give alpha and beta orbitals that are energetically equivalent, if your query is about that. Please elaborate on your original post if this is not what you are looking for.

#3

If you have a closed-shell system then calculation may not break alpha/beta spin symmetry even if unrestricted, because the guess has same orbitals for alpha and beta. Use SCF_GUESS_MIX (as described in the manual) to break the spin symmetry of the initial guess.
https://manual.q-chem.com/latest/sect_occupied.html

#4

I have a open shell (s=3/2) system.

#5

Step 1. Perform a quartet RODFT (i.e. ROKS) calculation.
Step 2. Perform a quartet UDFT (i.e. UKS) calculation using RODFT orbitals obtained above. Almost all programs will copy beta MOs from alpha MOs in this case, and thus the initial alpha/beta MOs are the same, but with different occupation numbers.

Is this what you want?