Broken symmetry in HF

[Jonathan Lasham]

Hi,

I'm trying to set up a broken-symmetry calculation for a spin-up Fe / spin-down Fe system, similar to what was done here using DFT:

https://www.nwchem-sw.org/index-php/Special_AWCforum/st/id1680/DFT_Broken-symmetry_for_molecule---.html

I'm using the nopen keyword in HF in a similar way to how mult was used in the DFT example, and in general, it works as expected.

However, I'm running into a problem when combining fragments, specifically when the overall spin polarization is in the β direction.

I've attached input and output files for two cases:

• BS_HF_a: frag1 = Fe(III) spin up, frag2 = Fe(II) spin down → nopen = 1. This works fine.

• BS_HF_b: frag1 = Fe(III) spin down, frag2 = Fe(II) spin up → nopen = -1. This gives the error:

  scf: no. of closed-shell electrons is not even!                   4
  

The equivalent DFT systems using mult (including negative values) run without issue. Is there a known limitation in HF regarding negative nopen values, or is there an alternative way to represent this kind of broken-symmetry setup in HF?

Many thanks,
Jonathan

[Edoardo Aprà]

Yes, you cannot use a negative nopen value  in the SCF module.

Instead, you could run the Hartree-Fock method while using the DFT module by setting the  XC keyword as "hfexch" https://nwchemgit.github.io/Density-Functional-Theory-for-Molecules.html#exchange-correlation-functionals

dft

 xc hfexch

  ...

end

task dft

If you wish to use the molecular orbitals generated by the dft module, this is possible by using the vectors option as in this code snippet below

dft

  xc hfexch

  vectors input atomic output dft.mos

end

task dft

scf

  vectors input dft.mos output hf.mos

end

task scf