Molcas Forum

dsb99

Member

Registered: Yesterday

Posts: 1

PREX function in SINGLE_ANISO

Dear Molcas developers,

I have a query in relation to the PREX function in SINGLE_ANISO.

I came across the PREX function whilst reading the 2023 JCTC OpenMolcas Web paper (https://doi.org/10.1021/acs.jctc.3c00182) and associated supporting information (specifically S19 in the pdf, and the attached supporting files DyMn_vtzp_vdzp_396r_PREX_SINGLE_ANISO.inp, DyMn_vtzp_vdzp_cas_14in12_396r_aniso_cas.aniso, Mn_in_DyMn_vtzp_vdzp_cas_5in5_aniso_cas.old.aniso, Dy_in_DyMn_vtzp_vdzp_cas_9in7_aniso_cas.old.aniso).

I am using OpenMolcas v26.02.

I have tried to implement the PREX function on a Ln-radical system in SINGLE_ANISO, following the whole molecule / fragment methodology described in the above paper. I have been unable to get the PREX function to work - it is silently ignored in my SINGLE_ANISO calculations, producing no "DERIVATION OF ANISOTROPIC MAGNETIC EXCHANGE..." block in the output file, and also producing no errors. The pseudospin Hamiltonian tensors and magnetic susceptibility were successfully calculated.

For reference, this was the input used:

>>COPY $FileDir/wholemolecule.aniso       $WorkDir/
>>COPY $FileDir/ln_fragment.old.aniso    $WorkDir/
>>COPY $FileDir/rad_fragment.old.aniso    $WorkDir/

&SINGLE_ANISO
DATA
wholemolecule.aniso
PREX
4 ln_fragment.old.aniso
2 rad_fragment.old.aniso
TINT
0 300 31
MLTP
4
2 2 2 2
End Of Input

I also tried using the .aniso files instead of the .old.aniso files for the fragments, but this did not solve the problem.

To see if the issue was specific to my Ln-rad system, I tried reproducing the results published in https://doi.org/10.1021/acs.jctc.3c00182 as seen in the supplementary file DyMn_vtzp_vdzp_396r_PREX_SINGLE_ANISO.out, using the identical input and .aniso/.old.aniso files. The exact same issue arose, where the PREX function was silently ignored, and no exchange parameter block was computed in the output file.

For reference, this was the input used:

>>COPY $FileDir/DyMn_vtzp_vdzp_cas_14in12_396r_aniso_cas.aniso       $WorkDir/
>>COPY $FileDir/Dy_in_DyMn_vtzp_vdzp_cas_9in7_aniso_cas.old.aniso    $WorkDir/
>>COPY $FileDir/Mn_in_DyMn_vtzp_vdzp_cas_5in5_aniso_cas.old.aniso    $WorkDir/

&SINGLE_ANISO
DATA
DyMn_vtzp_vdzp_cas_14in12_396r_aniso_cas.aniso
PREX
2 Dy_in_DyMn_vtzp_vdzp_cas_9in7_aniso_cas.old.aniso
6 Mn_in_DyMn_vtzp_vdzp_cas_5in5_aniso_cas.old.aniso
TINT
0 300 31
MLTP
6
2 2 2  2 2 2
End Of Input

I am wondering what might be causing this issue with PREX?

I have looked through GitLab OpenMolcas src/single_aniso and couldn't find any information/documentation on PREX. The closest thing I could find was pr_ito_int.F90 in src/poly_aniso. Is the ab initio calculation of anisotropic exchange parameters now available only in poly_aniso?

Many thanks in advance for your help!

Kind regards,
Dom

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,214

Re: PREX function in SINGLE_ANISO

As far as I can see, the PREX keyword is undocumented, and in the code it's a no-op (it does nothing).