Molcas Forum

Ada

Member

Registered: 2021-03-19

Posts: 6

Term symbol issue for free ions

Hi, I'm trying to use RASSI to evaluate the multiplet splitting in free lanthanide ions. To begin with, Ce(III), so a 4f^1 state with 2F7/2 and 2F5/2 terms. Input file is below. The issue I have is that there is a warning in RASSCF: "WARNING: Supersymmetry may have failed.", so presumably the ATOM keyword isn't doing what it's meant to. This then means that the 7 RASSCF roots aren't degenerate:

::    RASSCF root number  1 Total energy:  -8852.04608778
::    RASSCF root number  2 Total energy:  -8852.04608087
::    RASSCF root number  3 Total energy:  -8852.04602572
::    RASSCF root number  4 Total energy:  -8852.04601124
::    RASSCF root number  5 Total energy:  -8852.04599755
::    RASSCF root number  6 Total energy:  -8852.04598384
::    RASSCF root number  7 Total energy:  -8852.04593308

And in the RASSI output, although the L-values are correct (L_eff = 3.0 for all states, indicating the active space just contains the 4f orbitals), the J-values aren't correct and there is a small (~20cm-1) splitting in both the 2F5/2 and 2F7/2 terms:

 SO State   Relative EMIN(au)    Rel lowest level(eV)    D:o, cm**(-1)   J-value
     1            -0.0062566065        0.0000000000              0.0000         2.1
     2            -0.0062566065        0.0000000000              0.0000         2.1
     3            -0.0062283624        0.0007685601              6.1989         1.2
     4            -0.0062283624        0.0007685601              6.1989         1.2
     5            -0.0061682012        0.0024056306             19.4027         2.4
     6            -0.0061682012        0.0024056306             19.4027         2.4
     7             0.0047303768        0.2989710455           2411.3641         2.6
     8             0.0047303768        0.2989710455           2411.3641         2.6
     9             0.0047852104        0.3004631440           2423.3987         2.0
    10             0.0047852104        0.3004631440           2423.3987         2.0
    11             0.0047876713        0.3005301090           2423.9388         2.9
    12             0.0047876713        0.3005301090           2423.9388         2.9
    13             0.0048443054        0.3020712002           2436.3685         3.4
    14             0.0048443054        0.3020712002           2436.3685         3.4

I assume that this is all symmetry related, is there anything that can be done to control this and get the expected J-values?

Thanks in advance!

Input file
---

&GATEWAY

 TITLE
 Ce(III)

 BASIS SET
 Ce.ANO-rcc.Roos.25s22p15d11f4g2h.8s7p4d3f2g
  Ce         0.000     0.000      0.000 /Angstrom
 END OF BASIS

 AMFI

END OF INPUT

&SEWARD

 CHOLESKY

END OF INPUT

&SCF

 CHOLESKY

 IVO

 CHARGE
  4

 PRORBITALS
  1 3

END OF INPUT

&RASSCF

 CHOLESKY

 ATOM

 PRWF
  0.1

 SPIN
  2

 SYMMETRY
  1

 CIROOT
  7 7 1

 ORBLISTING
  ALL

 ORBAPPEAR
  FULL

 NACTEL
  1  0  0

 INACTIVE
  27

 RAS1
   0

 RAS2
   7

 RAS3
   0

END OF INPUT

&RASSI

 CHOLESKY

 SPINORBIT

 J-VALUE

END OF INPUT

Last edited by Ada (2026-02-11 10:57:34)

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,204

Re: Term symbol issue for free ions

1. ANO-RCC basis, with no further contraction, is huge (12s11p8d7f4g2h). You probably want something like ANO-RCC-VTZP (8s7p5d3f2g1h) or something like that instead (check with "pymolcas help_basis Ce"). AMFI is already activated by default when using ANO-RCC.

2. We don't recommend the CHOLESKY keyword. We recommend RICD instead. Among other things, the CHOLESKY keyword may mess with symmetries, which seems to be the problem in this case...

3. The ATOM keyword is probably not necessary in this case, maybe because of the state averaging.

Try this input:

&GATEWAY
  Coord = 1

    Ce 0.0 0.0 0.0
  Basis = ANO-RCC-VTZP
  Group = NoSym
  RICD

&SEWARD

&SCF
  Charge = 4

&RASSCF
  Atom
  Charge = 3 
  Spin = 2
  NActEl = 1
  RAS2 = 7
  StAverage = 7
  PrWf = 0.1

&RASSI
  Spin-orbit
  J-value

Ada

Member

Registered: 2021-03-19

Posts: 6

Re: Term symbol issue for free ions

That works, thank you!