Bandstructure in real space

[Hassan]

Hi,

My question is that can ATK help me plot the bandstructure of a device/bulk configuration in REAL space? A diagram where I can see band-splitting etc. I am talking about something like this:

Help is greatly appreciated.

[Anders Blom]

That's not a band structure in real space, it's a cartoon to convey an important conceptual principle. So no, you can't immediately just calculate and get a picture like that. But you can try in 1D to make an infinite chain of atoms, say Si, then modify the lattice constant from small to large, and plot the density of states in some clever way as function of the lattice spacing. That might look cool! Might work in 3D too, actually, provided the calculations converge when you reach large lattice constants.

[Anders Blom]

I had a bit of fun with this and actually I could get some quite cool plots. Attached is the corresponding figure for bulk Silicon, as computed with DFT/LDA. The most interesting thing, I think, is that the semiconducting nature of Si is almost a fluke - the band gap only exists for a very limited range of lattice constants.

There are a couple of different ways to plot it, but the simplest version is shown here, with scripts that are easily modified for any other structure - would be nice to try with a metal too!

First run "si_3d_a.py", then "si_3d_a_plot.py".

Note that because ATK uses pseudopotentials, we don't see the core levels 1s etc, the lowest level plotted is 3s, then 3p, etc.

[Anders Blom]

Same plot for gold fcc.

[Hassan]

Thanks a lot Anders!