1D-like charge system

[paulanatalia_abufager]

Hello, I am working with an infinite carbon nanotube (along z-direction) containing a charged atom inside. To model the charged system, I modify the total number of electrons using the NELECT tag in VASP. However, I am unsure whether dipole corrections are applicable or implemented in this case. According to the VASP documentation on electrostatics, dipole corrections are not available for 1D-like systems. Do you have any suggestions on how to deal with the artefacts introduced by charging such a system? Thanks, Paula

[marie-therese.huebsch]

Hi Paula,

Thank you for your question. Indeed, we currently do not have a feature that provides electrostatic corrections for 1D charged systems. We are working on something that will be in a future release, but for now I am afraid I cannot provide a better answer.

Marie-Therese

[paulanatalia_abufager]

Hello, If I compute relative energies for a charged system (by modifying NELECT) without applying dipole corrections, do you think the results are still reliable? Thank you!

[marie-therese.huebsch]

I am unfortunately not an expert in this area, but I asked around a bit, and here is what I got as a response:

The nuance here is what does compare mean? Are you taking an energy difference? If you are taking an energy difference, then you would need to align the potentials between the calculations (so-called potential alignment) to get meaningful energies.

Moving to a supercell with reduce the impact of the charge. But both the potential alignment and the monopole-monopole interaction energy (along the aperiodic dimensions) decrease very slowly with cell size, making convergence very challenging (in vacuum it is made more difficult imo) - extrapolating is also not very easy in lower dimensionality systems.

Note: we already have implemented the 0D version for charged systems where the energies are absolutely comparable - so ignoring the nanotube would get you a 0D system.

Maybe a user has experience removing electrons from tubes and can share it here?

All the best,
Marie-Therese