Molcas Forum

MaxParadiz

Member

Registered: 2019-09-01

Posts: 22

Mapping localized orbitals into a different geometry

Hello,

I have calculated the SCF orbitals for biphenyl in a geometry in the planes of the phenyl groups are perpendicular to each other. I then localized these orbitals using the Pipek-Mezey method.

I would like to take these localized orbitals and map them into a "flat" biphenyl, in which both phenyls are in-plane. Is there a way to do this? If I try to simply load the orbitals into the new geometry, the orbitals will not rotate with the rotating group. I am interested in learning how to do these mappings properly in order to set up energy surface calculations.

Thanks!

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,079

Re: Mapping localized orbitals into a different geometry

I don't think there's any practical way. Mathematically, it would require defining a rotation for each atom, and the orthogonality between orbitals would be lost. Keep in mind that the MOs are just a list of coefficients for, say, px, py, pz... and as you rotate things, px, py, pz keep their absolute orientation in space.

The same thing happens if you try to apply the orbitals from a geometry to the same geometry, rotated as a rigid body. But in this case the same rotation applies to all atoms, so it is possible to define a well-behaved transformation that keeps all orbitals orthogonal, although I don't think there's (currently) any tool for that either.

MaxParadiz

Member

Registered: 2019-09-01

Posts: 22

Re: Mapping localized orbitals into a different geometry

Ah, I understand. Thank you for your reply