I have calculated fermi surface using fs.x executable implemented in Quantum espresso, which generate bxsf file. To view, I have used the command xcrysden --bxsf file.bxsf. After that two windows are displayed. One is the bar graph where we need to select the bands whose fermi surface we want to visualize. My question is, on what basis can we select the required bands?. I have tried by clicking on any one bands but I am unable to view the fermi surface.
Is there some alternate way to calculate fermi surface of a magnetic material which generate two bxsf file , or is it same as a non magnetic material?
So the problem with Xcrysden is that if you include way too many bands it may fail! make sure you restrict the bands you use to the range of interest in terms of energy. For a conductor, that will most likely be the Fermi Level. And for insulators, you may choose an energy level of interest.
Nonetheless, xcrysden is an older program and there is a very high chance you run into errors and compatibility issues when you run it. So below are two convenient alternatives:
1) IFermi Python package: This is a very powerful package for everything you may need regarding plotting and understanding the Fermi surface. And the code is open source so you can edit it and tailor it for your needs. Here is a link to it.
https://fermisurfaces.github.io/IFermi/
2) FermiSurfer: This is probably a more convenient option because you do not have to write Python code to set it up and because it has a nice user interface that you can interact with. Among both options, this one is more similar to xcrysden. Here is a link to that package.
https://mitsuaki1987.github.io/fermisurfer/index.html
In my experience, you just select the bands that cross the Fermi level. The bands that barely touch the Fermi level are often hard to see. Next, once the Fermi surface is plotted, it is worth rotating the image a bit and playing with the smoothing. For some structures the Fermi surface is almost parallel to the initial line of sight, so it may look like nothing is being plotted.
Ivan I Yakovkin Thankyou Sir for your help.
How can we visualize the Fermi surface of a half-metallic system? Can we only calculate the Fermi surface for non spin-polarized systems (non-magnetic), or is there a specific approach or keywords to consider for magnetic systems?
Well, the ultimate question is whether your bands cross the Fermi level. I never worked with half-metals, but I would expect that if you can somehow divide your bands by their spin (not sure if it is possible), then for one spin orientation (metal) there would be a well-pronounced Fermi surface, while for the other (insulator) there would be no such surface. If such division is not made, then I would expect the Fermi surface to look similar to such in metals.
Ivan I Yakovkin, Geetimallika Das
You can divide Fermi surface by their spins. As a matter of fact, you should. Ferromagnets typically have different Fermi surfaces for spin up and spin down. On the other hand, antiferromagnets have degenerate spin-up and spin-down bands and would typically show no difference in the Fermi surface in either channel.
Also, half metlals cross the band surface so you should be able to see the Fermi surface crossing and plot the Fermi surface. Just keep in mind that the Fermi energy value you get from a bandstructure calculation may be different from the Fermi energy value you get from a density of states calculation. The latter is more accurate for a Fermi surface. Just make sure you use a denser mesh and include points on the edge of the Brillouin zone
Ahmed Elrashidy Thankyou Sir for your reply, I have tried for ferromagnetic materials. I have performed the nscf calculation using a k mesh of 30*30*30. But still no fermi surface is observed for the semiconducting channel as the bands didn't cross the fermi level. While for the metallic channel fermi surface is observed. So, do we need to focus on the nbnd provided or should I try by increasing the k mesh ?
- Badges
- Science topic
- Similar topics
- Engineering
- Chemical Engineering