Transport routines within the kubo approach

[apezol]

Dear developers, I'm sharing some routines to calculate transport properties using SIESTA and sisl. I'm not sure if this submit is done the properly nor I'm an actual developer, so I've tried to comment the lines within the codes to highlight how they work.

This routines allow to calculate some transport quantities using an approach base on the Kubo formalism. The velocities are constructed using both derivatives of the Hamiltonian and Overlap matrices. Within this formalism, there are three examples showing how to calculate the spin Hall conductivity in Pt, the Orbital Hall conductivity in MoS2 and the Orbital Rashba Edelstein effect in Cu/O.

Example cases SHC in bulk Pt (reference https://arxiv.org/abs/1810.07637) - routine -> shc.py OHC in 2d MoS2 - routine -> ohc.py Orbital Edelstein effect in oxidized Cu (Cu/O) (used in https://pubs.aip.org/aip/apm/article/12/5/051105/3287821/Quantifying-the-large-contribution-from-orbital) - routine -> job_surface_orbital ( In this case the quantity is calculated as a projection on the layers forming the system)

Comments: routines_sisl_kubo.zip

*The main advantage of using sisl as a post-processing tool to siesta is that we don't need to rely in any wannierization procedure since the orbitals are already localized.

• The SHC and OHC are the intrinsic contributions to the Hall response and there's an issue appearing when Spin is set to SOC+onsite. I still don't know if this is related to the version of Siesta under which the SCF simulation was performed.
• Finally, if the routines aren't well explained through the comments, I'll be more than happy to help with any doubt.

• [x] Author : Armando Pezo
• [x] Which version of sisl it was originally developed and tested with sisl version: 0.0.0.dev+eacb8351ff416ad9f7f93732777c3d7931c5f711 routines_sisl_kubo.zip