Molcas Forum

Alessandro

Member

Registered: 2024-09-24

Posts: 8

RASSCF Non adiabatic couplings ERROR

Hello,
I'm trying to compute the NACs between some excited states of a water dimer (geometry slightly displaced from equilibrium) following a RASSCF calculation. Unfortunately I keep getting the error: ERROR: Numerical nonadiabatic coupling not implemented

I am using version 22.10

This is my input:

&gateway
coord
6
wd_0.9
 O 1.3890808675 0.0005842735 -0.1129206998
 H 1.7531189116 0.7540006016 0.3579437378
 H 1.7508101582 -0.7589431599 0.3496672054
 O -1.5078156225 -0.0009491908 0.1193463536
 H -1.8983618924 0.0037547164 -0.7535878337
 H -0.6187999259 0.0051630336 -0.0250914259

basis
cc-pvtz

group
c1

ricd

angm
0 0 0

&SEWARD

&SCF
****************the two 2a1 to Ras1 and the HOMO-4 and HOMO-5 to inactive
&RASSCF
cionly
alter
2;1 3 6; 1 4 7
inactive
5
ras1
2
ras2
6
ras3
0
nactel
10 1 0
spin
1


&RASSCF
inactive
5
ras1
2
ras2
6
ras3
0
nactel
10 1 0
spin
1
supsym
1
2 6 7;0


*******************Inner-valence-excited state
&RASSCF
inactive
5
ras1
2
ras2
6
ras3
0
nactel
10 1 0
spin
1
hexs
1
1
ciroot
5 5 1
supsym
1
2 6 7;0
PRWF
0.0
*PRSD

&ALASKA
NAC
1 3

And the end of the output I get:

  Input file to MOLDEN was generated!


  Input file to MOLDEN was generated!


  Input file to MOLDEN was generated!


  Input file to MOLDEN was generated!


  Input file to MOLDEN was generated!


  Input file to MOLDEN was generated!

      Average orbitals are written to the RASORB file
      Natural orbitals for root   1 are written to the RASORB.1 file
      Natural orbitals for root   2 are written to the RASORB.2 file
      Natural orbitals for root   3 are written to the RASORB.3 file
      Natural orbitals for root   4 are written to the RASORB.4 file
      Natural orbitals for root   5 are written to the RASORB.5 file
      Spin density orbitals for root   1 are written to the SPDORB.1 file
      Spin density orbitals for root   2 are written to the SPDORB.2 file
      Spin density orbitals for root   3 are written to the SPDORB.3 file
      Spin density orbitals for root   4 are written to the SPDORB.4 file
      Spin density orbitals for root   5 are written to the SPDORB.5 file

 ###############################################################################
 ###############################################################################
 ###                                                                         ###
 ###                                                                         ###
 ###    WARNING: RunFile label BirthCertificate                              ###
 ###    was used       49 times                                              ###
 ###                                                                         ###
 ###                                                                         ###
 ###############################################################################
 ###############################################################################
--- Stop Module: rasscf at Tue Sep 24 16:22:02 2024 /rc=_RC_ALL_IS_WELL_ ---
--- Module rasscf spent 25 seconds ---

--- Start Module: alaska at Tue Sep 24 16:22:02 2024 ---

()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()

                                              &ALASKA

                                   only a single process is used
                       available to each process: 60 GB of memory, 16 threads
                                            pid: 4088051
()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()

 ###############################################################################
 ###############################################################################
 ###                                                                         ###
 ###                                                                         ###
 ###    ERROR: Numerical nonadiabatic coupling not implemented               ###
 ###                                                                         ###
 ###                                                                         ###
 ###############################################################################
 ###############################################################################
--- Stop Module: alaska at Tue Sep 24 16:22:02 2024 /rc=-6 ---

.########################.
.# Non-zero return code #.
.########################.

    Timing: Wall=70.45 User=57.85 System=4.59

Do you know if it's possible to compute the NACs between these states and in case how can I do it?

Thank you in advance for the answer

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: RASSCF Non adiabatic couplings ERROR

You could try with the latest version. With an older version, you may need to add "DoAnalytical" in SEWARD. But HEXS and/or SUPSYM may prevent analytical differentiation, in which case I don't think there's any solution at the moment.

Alessandro

Member

Registered: 2024-09-24

Posts: 8

Re: RASSCF Non adiabatic couplings ERROR

Thank you very much Ignacio for the quick answer,

unfortunately the DoAnalytical keyword in SEWARD doesn't change the situation.
Adding &ALASKA alone in the same input as above, manages to compute the Numerical Gradients but not the NACs.

Different calculations on a different system (with a totally different input) without SUPSYM manage to compute the NACs (in the same Molcas version), so I guess it might be that what's preventing the NAC evaluation.

Thank you again for the help.

Alessandro

Member

Registered: 2024-09-24

Posts: 8

Re: RASSCF Non adiabatic couplings ERROR

Hi Ignacio,

I would have another question. I actually retried with the same system but simplifying the input, removing that rotation of orbitals and increasing the RAS2 space, such as to avoid SUPSYM but still the NACs calculation doesn't work.
Input:

&gateway
coord
6
wd_1.0
 O 1.3832027274 0.0005305073 -0.1113010031
 H 1.7431113123 0.7547042488 0.3616213140
 H 1.7415847054 -0.7598653852 0.3527901748
 O -1.5136862649 -0.0010516718 0.1210592194
 H -1.8809444243 0.0041975492 -0.7638747932
 H -0.5598832934 0.0051329031 -0.0286024265
basis
cc-pvtz

group
c1

ricd

angm
0 0 0

&SEWARD

&SCF
*******************Inner-valence-excited state
&RASSCF
cionly
inactive
2
ras1
2
ras2
9
ras3
0
nactel
16 1 0
spin
1
ciroot
5 5 1
PRWF
0.5

&ALASKA
&ALASKA
NAC
1 4

Output error:

()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()

                                              &ALASKA

                                   only a single process is used
                       available to each process: 60 GB of memory, 16 threads
                                             pid: 35869
()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()()

 ###############################################################################
 ###############################################################################
 ###                                                                         ###
 ###                                                                         ###
 ###    ERROR: Numerical nonadiabatic coupling not implemented               ###
 ###                                                                         ###
 ###                                                                         ###
 ###############################################################################
 ###############################################################################
--- Stop Module: alaska at Thu Sep 26 16:14:08 2024 /rc=-6 ---

.########################.
.# Non-zero return code #.
.########################.

At this point it's not the supsym or hexs keyword the problem. Any idea?

Thank you in advance for the answer

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: RASSCF Non adiabatic couplings ERROR

I don't think RASSCF (i.e. not CASSCF) analytical gradients (or couplings) are implemented, at least' they are not tested.