Molcas Forum

Support and discussions for Molcas and OpenMolcas users and developers

You are not logged in.

Announcement

Welcome to the Molcas forum. You can choose an avatar and change the default style by going to "Profile" → "Personality" or "Display".

Pages: 1

#1 2025-08-31 01:56:35

Lukhmanul Hakeem k
Member
Registered: 2024-02-05
Posts: 45

Assistance with RASSCF Optimization Using Floating Center

Dear Sir,

I am currently attempting to optimize the πσ* state (symmetry A"). For the 1πσ* excited-state optimizations, I followed the procedure described in the literature where the basis set is supplemented with an additional set of s and p diffuse Gaussian functions (exponent 0.02), localized at a floating center. This floating center is intended to provide extra flexibility in describing the diffuse σ* orbital, and its position should be optimized during the geometry optimization. Using below input file

&GATEWAY
Title
 Benzenemolecule.
Symmetry
  Z
Basis set
C.STO-3G.
C1	  -2.899818	  1.863052	  0.00000000	angstrom
C2	  -1.813673	  2.714537	  0.00000000	angstrom
C3	  -0.498277	  2.198637	  0.00000000	angstrom
C4	  -0.287677	  0.831234	  0.00000000	angstrom
............
End of basis
Basis set
O.STO-3G.
O11	   0.978864	  0.291367	  0.00000000	angstrom
..........................................................................................
End of basis
Basis set
H.STO-3G.
H13	  -3.909458	  2.25805	    0.00000000	angstrom
.................................................................................
End of basis
Basis set
X....1s1p / Inline
  0.00 1
  * s functions
    1 1
    0.02
    1.0
  * p functions
    1 1
    0.02
    1.0
    X      -1.54216915   -0.02400126   0.00000000 angstrom
End of basis
>>> DO WHILE <<<
&SEWARD
>>> IF (ITER = 1)
 &SCF
  Occupied = 47 10
  Iterations = 40
>>> ENDIF
&RASSCF
    Title = SA-RASSCF over 3 singlets, follow S1
    Symmetry = 2
    Spin = 1
    NACTEL = 8 0 0
    INACTIVE = 47 6
    RAS2 = 2 6
    LEVSHFT=1.0
&ALASKA
&SLAPAF
Cartesian
EndofInput
>>> ENDDO <<<


However, when I include the floating center in my input file (see below), the calculation fails with the following error message:

[ process     14]: xquit (rc =     64): _CONTINUE_LOOP_
[ process     21]: xquit (rc =     64): _CONTINUE_LOOP_
[ process     10]: xquit (rc =     64): _CONTINUE_LOOP_
[ process      6]: xquit (rc =     64): _CONTINUE_LOOP_
[ process      5]: xquit (rc =     64): _CONTINUE_LOOP_
[ process      3]: xquit (rc =     64): _CONTINUE_LOOP_
[ process      2]: xquit (rc =     64): _CONTINUE_LOOP_
[ process     12]: xquit (rc =     64): _CONTINUE_LOOP_
[ process     15]: xquit (rc =     64): _CONTINUE_LOOP_
[ process     11]: xquit (rc =     64): _CONTINUE_LOOP_
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_DIVIDE_BY_ZERO IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
--- Stop Module: slapaf at Sun Aug 31 02:48:37 2025 /rc=_RC_CONTINUE_LOOP_ ---
*** files: molcas_cas_s3_opt_test1.geo.molden molcas_cas_s3_opt_test1.slapaf.h5 molcas_cas_s3_opt_test1.structure molcas_cas_s3_opt_test1.Opt.xyz xmldump
    saved to directory /home/kavya.vnd20/PKLu/allo/PES/opt/new-pes/CASPT2/s3/opt-s3
--- Module slapaf spent 7 seconds ---

>>> END DO

.#######################.
.# Convergence problem #.
.#######################.

    Timing: Wall=2850.85 User=39787.02 System=4545.88

The RASSCF module stops with a non-convergence error, while the same state optimization without the floating center converges successfully.

Could you please advise me on how to correctly include the floating center in the optimization and resolve this convergence issue?

Thank you very much for your help.

Best regards,

Last edited by Lukhmanul Hakeem k (2025-09-12 18:33:31)

Offline

#2 2025-09-01 09:48:27

Ignacio
Administrator
From: Uppsala
Registered: 2015-11-03
Posts: 1,204

Re: Assistance with RASSCF Optimization Using Floating Center

Parallelization for this calculation is probably overkill, it will not give you any significant gain and will be more likely to be problematic. You don't show the actual reason for the calculation stopping, I guess it's just that the number of iterations reached the maximum (which is not unexpected given you have a few molecules with weak interactions). Also, using RICD will probably improve performance.

To increase the maximum number of iterations in the "do while" loop, add "> export MOLCAS_MAXITER=500" (or whatever number, default is 50) before the loop, or define the environment variable in your script.

The main problem, however, is that the ghost atom X will move away from your aromatic molecule, I don't think that's what you want. Perhaps you want to define some constraints for the X position.

Offline

#3 2025-09-07 13:12:24

Lukhmanul Hakeem k
Member
Registered: 2024-02-05
Posts: 45

Re: Assistance with RASSCF Optimization Using Floating Center

Dear Sir,

Thank you for your response. I tried the optimization again, but it still fails. I think the issue is with adding the ghost atom. I am considering constraining the ghost atom inside a circle with a 5 Å radius so that it can move only within this region. Is such a constrained optimization possible in OpenMolcas?

Best regards,
Lukhmanul Hakeem K.

Offline

#4 2025-09-08 08:25:13

Ignacio
Administrator
From: Uppsala
Registered: 2015-11-03
Posts: 1,204

Re: Assistance with RASSCF Optimization Using Floating Center

Not that I know. You can define constraints as linear combinations of internal or cartesian coordinates, but you must assign them a value, not a range.

Offline

Pages: 1

Links: OpenMolcas | Molcas | Manual | Facebook

Board footer

Powered by FluxBB 1.5.11

Last refresh: Today 10:50:40