QuantumATK Forum
QuantumATK => General Questions and Answers => Topic started by: wntjr1024 on March 24, 2025, 12:36
-
I read the manual on polarization, but it only covers calculations for primitive cells, and I don't know if the polarization term can calculate polarization on the interface.
1st question : Is it possible to calculate polarization at the interface of a heterostructure in QuantumATK?
2nd question : How is the area for calculating polarization in QunatumATK polarization? ex) entire bulk configuration
-
It should work for a slab configuration too, and I guess even a heterostructure.
As for the normalization, the absolute value is anyway typically not so accurate or relevant, what you are looking for are changes in the polarization with e.g. stress or geometry. As long as those calculations are run on the same volume/area, the overall normalization factor cancels out when looking at the trend.
-
I have an additional question. When materials A and B are combined in a hetero structure, the polarization of materials A and B are different. Then, the polarization should be different in two areas, but the result is only one polarization. Then, I wonder in which area this polarization was calculated (ex. A, B, A-B).
-
Yes, interfaces are not a single material, so like for any quantity in this situation, you really have to think about what you are looking to compute. The same question arises for e.g. a band gap or density of states - we cannot really talk about the band gap of an interface, but we can look at the alignment of the valence and conduction bands across the interface. So probably it makes no real sense to compute the total polarization for a system with an embedded interface, but that was your original premise, so I just responded technically to that. Whether or not the calculation is useful is up to you :-)