So I am working with Quantum Espresso (6.5) and we have built a few square planar Fe complexes (to be specific $\ce{Fe(NH_2)_2(OH)_2}$ cis and trans), and we ran scf (self consistent calculation) and pdos (projected density of state) calculations on it to get info on the atomic orbital occupancy and the magnetic moments. But this still doesn't give the energy splitting (d orbital splitting) order of the system. Can someone suggest some methods where we can get the crystal/ligand field splitting of d orbital of Fe using DFT (preferably using QE but other methods work too)?

Also is there a way to get this order from the occupancy of orbitals only?

PS: I understand WIEN2K would be of use here but because of some issue we won't be able to use that, so if there are other methods, do tell us.

EDIT: I did some asking around and apparently the order of energy can be found by using tight-binding-method or by group theory but I'm not really sure if someone can explain how those methods would work here.

I'd still prefer using quantum espresso for this though if somehow possible

  • $\begingroup$ If you are only interested in single molecule, can't you use Gaussian or Orca or ADF to compute these properties? I know that you can compute Pados from Gaussian calculations using GaussSum. $\endgroup$ Commented Jun 20, 2023 at 12:32
  • 1
    $\begingroup$ @HemanthHaridas It's not exactly a single molecule, what we are trying to do is study a monomer which is what I've mentioned and compare it to a much larger but similar polymer whose experimental results we have. And our previous calculations are based on QE so it would be preferred to do it in QE (but is not a limitation) since the degauss and all those value can cause variation $\endgroup$ Commented Jun 20, 2023 at 13:00

1 Answer 1


You should take into account that for this type of square planar complex the set of 5 "d" orbilals of the iron atom are the ones that must be considered for the calculation. Usually, depending on the type of ligands (weak field or strong field) you can have a different orbital degeneracy for the "eg" and "t2g" orbitals, which are the energy levels that must be considered for the calculation of the field separation in a square planar complex (SP). Note that I have used the quotes for the label of the orbitals since they are the nomenclature that results from an octahedral complex from which 2 ligands of the z axis have been removed, the nomenclature for SP complexes does not follow these labels (e.g. they can be a, e, t2) The arrangement of these levels results in different configurations depending on the type of ligand and worsens a bit when there is metal-ligand hybridization (more covalent than ionic bonds). Example, array 1-2-2 , that is, the highest energy level "1", two degenerate levels "2" lowest in energy and "2" last two degenerate levels lowest in energy. In the literature there are many examples with 2-1-2 and 2-2-1 configurations as well. Your example has probably already been calculated. "Usually" the orbit that alone (1) is the dz2 and is not involved in the calculation of the ligand field separation delta. Finally, identify your "d" orbitals with a calculation of DOS in QE. Then order them in decreasing energy, look for the degenerate orbitals (same energy) and go to the formulas to make the calculation that you can find in the literature. Note: maybe someone has already developed a code that reads the output of QE and responds quickly, but be careful, check the occupancies of the orbitals, it is an extremely important theoretical aspect that you may need to consider based on the theory of LCAOs to ponder the energies of the orbitals.

  • $\begingroup$ We did a pdos (Projected/Partial Density of States) calculation using QE but that will also give us the occupancy of the system and not the orbital energy order and all $\endgroup$ Commented Jul 21, 2023 at 10:19
  • $\begingroup$ Try Chemissian or Chemcraft to read the wave function file. Maybe Multigenwfn has the option to search the energy levels from the XYZ structure. Also, you can explore codes that are in ASE code. You can obtain some guidance by using ChatGPT if you do not have experience. I hope I help you. Regards $\endgroup$
    – Deivi X
    Commented Aug 15, 2023 at 13:52

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .