Hi all!

I write with this problem I am facing for few weeks now. I am doing ROHF calculations for enzyme's active center that contains iron ion as cofactor. I use ORCA software. I try to get energy for iron's triplet and quintet spin states.

I am following CAS/NEVPT procedure.

I did ROHF/CAS/NEVPT calculations in cc-pVDZ/cc-pVTZ basis set and everything was fine: in triplet spin state, "mulliken population" section in output show approximately 1.9 on iron and in quintet spin state 3.9. After looking at orbitals, they were looking legit as well.

But I'm trying now to increase basis set and use cc-pVTZ/cc-pwCVQZ (so quadruple-dzeta basis set only for iron atom). And in this case, i can't obtain proper spin mulliken populations on iron during calculations for quintet state (triplets are calculating properly).

In output, mulliken charge/ population for iron looks like that:

35 Fe: 2.575985 2.858409

So i get only almost three unpaired on iron when i need four.

My question is, is there any trick to enforce unpaired electrons to sit on the iron ion?

Head of my input:

| 4> %pal nprocs 24

| 5> end

| 6>

| 7> %maxcore 4000

| 8>

| 9> ! ROHF NoIter

| 10> %basis basis "cc-pVTZ"

| 11> newGTO N "cc-pVTZ" end

| 12> newGTO O "cc-pVTZ" end

| 13> newGTO Fe "cc-pwCVQZ" end

| 14> end

| 15>

| 16> * xyz 1 5

| 17>here goes the geometry

I will be extremely grateful for any tips and advices.