Property fails with cphf_solve2:SCF residual greater than 1d-2

41 views
Skip to first unread message

Szabolcs Góger

unread,
Jan 30, 2024, 10:23:21 AMJan 30
to nwchem...@googlegroups.com
Dear Community,

I am trying to run a TD-DFT property calculation (optical rotation tensor, but I imagine the problem is relevant for any property). In particular, the SCF input to the CPHF module is obtained from a PBE+X2C calculation, e.g., the input is

'''
dft
 xc xpbe96 1.0 cpbe96 1.0
end

relativistic
 x2c on
end

property
 response  1 0.0000  
 velocity            
 orbeta              
 giao               
 bdtensor         
 analysis        
 convergence 1e-4      
end
'''

The same calculation seems to work without the x2c correction. This, together with the text of the error message, leads me to believe that the Hamiltonian in the CPHF module is not exactly the same as in the SCF calculation - is this assessment correct? If yes, is there a way to run this PBE+X2C calculation - or, an even more important application, is the XC kernel itself inherited in the property calculation, or is it actually ''just'' a CPHF run for the properties? This seems a bit unclear to me from the manual and the output file.

Thank you in advance,

Szabolcs Góger

unread,
Feb 8, 2024, 9:39:56 AMFeb 8
to NWChem Forum
Dear Community,

I realize I did not upload a sample output file last time; which might help to facilitate the discussion. I am attaching one now.

Best regards,
Szabolcs
3AlaD_x2c_avdz.inp.out

Szabolcs Góger

unread,
Mar 12, 2024, 5:51:28 PMMar 12
to NWChem Forum
Dear Community,

Wile the issue is still not fully resolved, I am almost convinced that the underlying calculation for the optical response is just RHF, even when a DFT run is called beforehand.

Best,
Szabolcs
Reply all
Reply to author
Forward
0 new messages