Molcas Forum

FAguilar

Member

From: San Sebastián, Spain

Registered: 2021-04-16

Posts: 5

Website

ALTEr and SUPSym

Dear all,

I'm trying to optimize at CASSCF level the ground state of a complex with a transition metal cation, but I always get a rotation of the active space (one of the d orbitals goes out of the active space and some sigma* orbital of the ligand gets in). I want to keep the d orbitals of the metal and the sigma orbitals (but not the sigma antibonding, due to the huge computational effort). So I tried to use SUPSym to create a new symmetry group that contains all my desired active space (and only these orbitals)

After a SCF single point to get the starting orbitals, I realize that I want to put in the active space orbital #35 and I don't want #32. My first question is, when I'm going to define the orbitals in the new symmetry, should I refer the number to the "old" label of the orbital (same number than if I don't ask for alter, so in this case, #35) or the number should be the one after the alter? (so, in this example, #32)
Here is the example of my input (considering that the label is affected by alter)

 &RASSCF
   LumOrb
   Title
   primer casscf
   Spin
     4
   Symmetry
     1
   Nactel
     15 0 0
   Inactive
     19
   Ras2
     13
   CiRoot
     1 1 1
   Alter
   1
   1 32 35
   SUPSym
   1
   13 20 21 22 23 24 25 26 27 28 29 30 31 32
   ORBAppear
   FULL
   ORBListing
   ALL

And, regarding this extra symmetry group... Is it created locally, right? I mean, if in a later job where I use the orbitals that I'm getting with SUPSym I don't ask for this keyword, my orbitals are going to be able to mix with the others, is it correct? I need to "freeze" the active space when there is a long metal-ligand distance, but at shorter distances, I don't need to impose any restriction, the active space is stable...

Thanks in advance,

Fernando

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,079

Re: ALTEr and SUPSym

My first question is, when I'm going to define the orbitals in the new symmetry, should I refer the number to the "old" label of the orbital (same number than if I don't ask for alter, so in this case, #35) or the number should be the one after the alter? (so, in this example, #32)

I think it should be after alter, i.e. first the orbitals are reordered and then the supsym is applied. But you should double check and try with a fast calculation first

And, regarding this extra symmetry group... Is it created locally, right?

Yes, it's only active for that particular RASSCF run as far as I know. I believe the supsym information is saved in the .h5 file, but I don't think it's read (in case you were planning to use .h5 files).

FAguilar

Member

From: San Sebastián, Spain

Registered: 2021-04-16

Posts: 5

Website

Re: ALTEr and SUPSym

Thank you very much, Ignacio!!!

Yes, I think it would be a nice idea to check it. Since the calculation with SUPSym won't be the last one, the most important thing is having the correct active space, so it "doesn't matter". For the later CASPT2 single point, I won't impose the extra symmetry, so it won't be a real problem.

Thank you again!

Fernando