Molcas Forum

jcardool

Member

Registered: 2026-03-26

Posts: 1

RASSI HFC Infinity values in A-tensor for Ho(III) with RX2C/Gen1Int

Dear all,

I am currently trying to calculate the hyperfine coupling constants (HFCs) of a Ho(III) molecular system using the RASSI module in OpenMolcas v25.06, and I am running into a numerical issue that I have not encountered with other lanthanide systems (Yb, Dy, Pr).

---

Problem description

All HFC tensor terms (FC, SD, PSO, and combined terms) return `Infinity` values for every single matrix element. Below is a representative extract from the Spin-Dipole (SD) term output:

   SO-STATE    1   E =        -0.02953360
   ----------------------------------------------
     A_TENS_TERM(ic1,ic2):
                  Infinity             Infinity                Infinity
                  Infinity             Infinity                Infinity
                  Infinity             Infinity                Infinity
 no twofold degeneracy
   =========================================
   A (SD)-Matrix for center:                    1
   =========================================
 Looking for ASD    1
   SO-STATE    1   E =        -0.02953360
   ----------------------------------------------
     A_TENS_TERM(ic1,ic2):
                  Infinity             Infinity                Infinity
                  Infinity             Infinity                Infinity
                  Infinity             Infinity                Infinity
 no twofold degeneracy

This pattern repeats for every term and every SO state.

---

Computational setup

Software: OpenMolcas v25.06, compiled with the Gen1Int submodule for relativistic integrals.

Basis sets:
- Ho: CC-pVZ-X2C
- First coordination sphere: CC-pVTZ-DK
- Remaining atoms: CC-pVDZ-DK

SEWARD input (skipping Basis set):

&SEWARD  &END
FINIte
cholesky
BSSHOW
Basis set
...
End of basis
RX2C
clight
  137.036
Angmom
  0.0 0.0 0.0
FLDG
  0
AMFI
MXTC
End of input

RASSI input:

Three JobIph files from RASSCF calculations with multiplicities 5, 3, and 1 (35, 210, and 196 roots respectively). The full state list is included and the following property keywords are requested:

&RASSI &END
NrofJobIphs
3 35 210 196
[all states listed]
SPIN
EPRA
0.0
AFCC
ASDC
FCSD
APSO
ATSA
PROPerties
  9
'ASD    1' 1
'ASD    1' 2
...
'PSOP   1' 3
End of input

---

What I have already checked

- The same methodology works correctly for Yb, Dy, and Pr systems with no issues and the RASSCF calculations converge normally.

---

Questions

1. Is there a known issue with RASSI HFC calculations specifically for Ho(III), possibly related to the number of roots (441 SO states in total) or the spin multiplicities involved?
2. Is there a recommended way to diagnose which intermediate quantity is producing the `Infinity` (e.g., a specific matrix element of the property integrals or the SO-coupling matrix)?

Any hints or suggestions would be greatly appreciated. I am happy to provide additional output files or the full input decks if needed.

Thank you in advance.

Last edited by jcardool (2026-03-31 11:01:39)

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,203

Re: RASSI HFC Infinity values in A-tensor for Ho(III) with RX2C/Gen1Int

Without a full input that reproduces the problem, it's difficult to say anything. "Infinity" is probably a division by zero somewhere...