Molcas Forum

sruta

Member

Registered: 2023-10-25

Posts: 1

MCPDFT combined with RASSI

I am trying to use the results from MC-PDFT in the RASSI and WFA modules. The MC-PDFT is running without problems and printing out the corrected energies, but it does not create the JOBPDFT and MCDENS files nor update the current JobIph file. So the RASSI and WFA modules take the information only from the RASSCF run.
I am using OpenMolcas v22.10, the input file is bellow:

&GATEWAY
  title = RASSCF
  coord = fa.xyz
  RICD
  basis = ANO-RCC-VDZP
  SYMT = 0.0

&SEWARD
  grid  input
  grid=ultrafine
  end  of  grid  input

&SCF

&RASSCF
  Charge = 0
  RAS2 = 6
  NActEl = 4
  Spin = 3
  CIroot = 2 2 1

&MCPDFT
  KSDFT = T:PBE
  WJOB

&RASSI
TRD1

&WFA
h5file = $Project.rassi.h5
NXO
H5ORbitals

asauza

Member

Registered: 2023-12-07

Posts: 3

Re: MCPDFT combined with RASSI

When using RASSI, you have to include NROF to specify the number of JobIph files you are going to use, also, with the EJOB keyword you specify the MC-PDFT energies you are going to substitute instead of the RASSCF energies.

Something like this:

>>>COPY $CurrDir/triplet_calculation.JobIph JOB001

&RASSI  &END
NROF = 1 2 
1 2 

EJOB
MC-PDFT energy1
MC-PDFT energy2

emiranda

Member

Registered: 2024-05-28

Posts: 5

Re: MCPDFT combined with RASSI

I am encountering the same issue!

From what I understand, when using .JobIph, you are relying on the RASSCF results. However, if you intend to calculate the Oscillator Strengths, as exemple, using MCPDFT, you should use the .JobPDFT file before running RASSI, like this:

>>>COPY $CurrDir/triplet_calculation.JobPDFT JOB001

Unfortunately, in the current version (v24.06), the .JobPDFT file does not seem to be generated. Has anyone found a workaround or can confirm this behavior?

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,085

Re: MCPDFT combined with RASSI

I don't see any issue. With the following input file:

&GATEWAY
  Coord =    9
    angstrom
    C     0.01123794     0.11528521     0.03174185
    H     0.45737989    -0.83268891    -0.22911283
    Cl    0.52115838     1.26427094    -1.22533826
    C    -1.50853461    -0.01013742     0.08144770
    H    -1.96369097     0.94050810     0.32698742
    H    -1.87932294    -0.33976623    -0.88019620
    C     0.56185030     0.52431144     1.33873927
    N     0.97353254     0.82084915     2.38274348
    H    -1.78681974    -0.73865819     0.83488757
  Basis = ANO-RCC-MB
  Group = NoSym
  RICD

&SEWARD

&RASSCF
  Spin = 1
  Charge = 0
  NactEl = 8
  RAS2 = 7
  StAv = 6

&RASSI
  EJob

&MCPDFT
  KSDFT = t:PBE
  WJob

&RASSI
  EJob

I get the following ("grep ::" on the log file):

::    RASSCF root number  1 Total energy:   -631.18461081
::    RASSCF root number  2 Total energy:   -630.88389586
::    RASSCF root number  3 Total energy:   -630.88076101
::    RASSCF root number  4 Total energy:   -630.82899212
::    RASSCF root number  5 Total energy:   -630.79937643
::    RASSCF root number  6 Total energy:   -630.79815680
::    RASSI State    1     Total energy:    -631.18461081529756
::    RASSI State    2     Total energy:    -630.88389585784205
::    RASSI State    3     Total energy:    -630.88076100768421
::    RASSI State    4     Total energy:    -630.82899212207803
::    RASSI State    5     Total energy:    -630.79937643133997
::    RASSI State    6     Total energy:    -630.79815680096556
::    Total MC-PDFT energy for state     1                               -632.52297840
::    Total MC-PDFT energy for state     2                               -632.25221210
::    Total MC-PDFT energy for state     3                               -632.24694129
::    Total MC-PDFT energy for state     4                               -632.20877005
::    Total MC-PDFT energy for state     5                               -632.21052113
::    Total MC-PDFT energy for state     6                               -632.20541158
::    RASSI State    1     Total energy:    -632.52297840273684
::    RASSI State    2     Total energy:    -632.25221209514507
::    RASSI State    3     Total energy:    -632.24694129459317
::    RASSI State    4     Total energy:    -632.20877005261946
::    RASSI State    5     Total energy:    -632.21052112540406
::    RASSI State    6     Total energy:    -632.20541158390256

Clearly, the first RASSI is reading the RASSCF energies, and the second RASSI is reading the MC-PDFT energies.

I think the OP's original problem was the missing EJOB keyword. Without EJOB, RASSI will compute all elements of the Hamiltonian matrix, that includes the state energies. But RASSI doesn't "know" anything other than plain (RAS)SCF, so it will compute just the RASSCF energies. If one wants to use other post-RASSCF (MCPDFT, CASPT2, etc.) energies, one must use EJOB or some other similar keyword.