Molcas Forum

Nuno

Member

Registered: 2016-01-15

Posts: 14

MC-PDFT geometry scan

Hello all.

I am attempting to perform a relaxed scan with CMS-MC-PDFT and I got a strange error message with one of the scan geometries:

--- Start Module: mclr at Mon Oct 30 08:38:31 2023 ---
 
      Residual in Qaa Lagrange Multipliers:    4.334574129E-03
 
      ERROR: RESIDUAL(S) FOR INTERMEDIATE STATE TOO BIG!
 
      This may come from a linear molecular or a linear
      fragment.
      CMS-PDFT Lagrange multipliers are not solved.
 ###############################################################################
 ###############################################################################
 ###                                                                         ###
 ###                                                                         ###
 ###    ERROR: Residual in Lagrange Multipliers for Qaa Too Big              ###
 ###                                                                         ###
 ###                                                                         ###
 ###############################################################################
 ###############################################################################

I should say that the molecule is not at all linear.

Does anyone know anyway to get around this ?

Nuno

Member

Registered: 2016-01-15

Posts: 14

Re: MC-PDFT geometry scan

Just in case anyone comes across the same problem here's the solution.

You need to input draconian thresholds for the CMSI rotation matrix. The calculation keeps on going without a hiccup.

&RASSCF
    CMSI    
    CMTH = 1e-12
    CMMAX = 400
    THRS = 1e-10 1e-4 1e-4
...