Multiwfn forum / Localized Orbital Energies

Re: Localized Orbital Energies

dummy@example.com (sobereva) — Tue, 26 May 2020 02:03:50 +0000

dcjskea wrote:

Hello again. I would like some advice please. I am successfully using Multiwfn to calculate localized orbitals and energies. However I am not sure if I am using the correct process for molecules that have an unpaired electron (unrestricted wavefunction) e.g. NO2. I have done localized energy calculations in such cases by combining the A Fock matrix with the B Fock matrix. Is this correct? The results show the A and B energies to be similar, but not the same. In some cases the A orbital is different from the corresponding B orbital. Can you give me any advice on the interpretation of the (different) A/B results? Thanks. Regards Donald.

This is very normal situation. For open-shell case, spin polarization effect makes energies of alpha orbitals different to that of beta orbitals.

Re: Localized Orbital Energies

dummy@example.com (dcjskea) — Mon, 25 May 2020 23:27:55 +0000

Hello again. I would like some advice please. I am successfully using Multiwfn to calculate localized orbitals and energies. However I am not sure if I am using the correct process for molecules that have an unpaired electron (unrestricted wavefunction) e.g. NO2. I have done localized energy calculations in such cases by combining the A Fock matrix with the B Fock matrix. Is this correct? The results show the A and B energies to be similar, but not the same. In some cases the A orbital is different from the corresponding B orbital. Can you give me any advice on the interpretation of the (different) A/B results? Thanks. Regards Donald.

Re: Localized Orbital Energies

dummy@example.com (sobereva) — Fri, 29 Nov 2019 22:57:50 +0000

Dear Donald,

Please pay attention to this statement in Section 3.22 of Multiwfn manual:

The matrix elements should be provided in lower-triangular form, namely in this sequence: F(1,1) F(2,1) F(2,2) F(3,1) F(3,2) F(3,3) ... F(nbasis,nbasis), where nbasis is the total number of basis functions, the format is free.

Obviously, you don't need to carry out any mathematical transformation on the origin Fock matrix, you just need to ignore all F(i,j) elements whose j>i.

Best regards,

Tian

Localized Orbital Energies

dummy@example.com (dcjskea) — Fri, 29 Nov 2019 17:22:37 +0000

Hi, this is my first post. I am quite new to ab initio electron calculations. I am using Psi4 with Multiwfn. I am interested in localized orbital calculations on small molecules, giving meaningful information on chemical bonding structures. I am first of all testing the programs on a methane molecule. The visual results using Psi4 / Multiwfn give 4 (beautiful) tetrahedral localized bonding orbitals, but I run into problems when I try to calculate the energy.
My question concerns the Fock matrix in lower triangular form. Examination shows the Fock matrix to be symmetric about the leading diagonal. Is "lower triangular form" the same matrix with just the upper triangular entries removed? Or is "lower triangular form" the matrix after a gaussian elimination to make the upper triangular entries equal to zero? I would value your help. Regards Donald.