Highest Occupied Crystalline Orbital Visualisation

[ryan_dempsey]

Dear Users,

I recently came across works that display the highest occupied crystalline orbital (HOCO). I am trying to compare some results I find in VASP to previous molecular calculations done with atom-centered basis sets, and it would be nice to be able to compare the HOCO to the molecular HOMO orbital. For those familiar with VASPKIT, it is possible to produce real space wave function isosurfaces from the VASP WAVECAR file (command 511), where the HOCO is at the Valence Band Maximum (VBM).1 I understand that this is not a VASPKIT forum, but I was hoping someone could explain to me whether it is fair to compare these constructions to atom-centered molecular orbitals, or whether I should not consider this as an orbital at all. Finally, what would the units of this isosurface be? The VASPKIT paper, nor any paper I have found with such figures, present numerical isosurface values...

[1] Section 3.7, eqn. 29 -- https://www.sciencedirect.com/science/a ... 5521001454

Thanks,
Ryan

[manuel_engel1]

Hi Ryan,

I am not familiar with VASPKIT, but I can comment on visualizing orbitals in general. The easiest thing to do would be to not plot the orbital but the partial charge density attributed to that orbital. You can read about the partial charge-density decomposition here. In essence, you can select the states you are interested in and obtain a partial charge density corresponding to that state. The output file has the same format as the CHG file and can be plotted using software like VESTA. The biggest disadvantage of doing this is that you loose all phase information, which might be necessary to identify bonding/anti-bonding orbitals.

Plotting directly from the WAVECAR file is in principle possible, but the information contained is given in the plane-wave basis and is thus not really suitable for plotting. There seem to be a few projects out there that try to translate the WAVECAR information to real-space for plotting purposes, but I cannot comment on whether they work or work correctly.

Finally, there is one more caveat. Internally, VASP calculates the orbitals from the projector-augmented-wave formalism, meaning that each orbital is split into different contributions. To the best of my knowledge, there is currently no easy way to plot the on-site contribution to these orbitals (because they are represented on radial logarithmic grids around the atoms), which might or might not be relevant to you. But it is a thing to be aware of.