Difference in density of states for bulk and device density of states

[Arya]

Hi All,
I calculated density of states for a bulk configuration and then used device from bulk to get a device. Then I added some charges on electrode to get a p-i-n structure. I calculated device density of states and found it to be significantly different from bulk DOS.

Can anyone explain why should DOS change significantly from device to bulk configuration?

Thanks

[Anders Blom]

Bulk is bulk and a device is very different from that. A p- or n-doped piece of a semiconductor has no real functionality, but put them together and you have something that rectifies current - a diode. So naturally the DOS is very different, and quite easily explained from the basic theory of a p-n junction. Note, however, that the device DOS you compute is a composite DOS for the whole structure, which perhaps is not so useful, in this case. It is however useful when you project it on the left and right atoms, and you will then see how the DOS there is indeed similar to that of the original bulk structures, but shifted in energy to create a common Fermi level, plus with band bending. In short, as shown in our SISPAD 2013 paper, available here: http://in4.iue.tuwien.ac.at/pdfs/sispad2013/19-1.pdf