Some Extended Huckel questions

[ewilson]

Hi,

I'm interested in studying interfaces with ATK's Extended Huckel and while working with it I came up with a few questions.

1.)  Not specifically relevant to interfaces, also to bulk calculations or what have you, but how is the Fermi level determined.  Say I do a bulk EHT calculation, how is the Fermi level set relative to say, the valence band maxima/conduction band minima?

2.)  How exactly is vacuum energy defined?  Is it actually the workfunction or is it a shift in the bandstructure to reproduce the correct workfunction?  Sorry if this is retreading old ground.

3.)  Related question.  How to approach band offsets on interfaces?  Suppose I know the band offsets from experiment, how should I (or should I) adjust the the vacuum energy?  How does that change if I want to use different parameter sets, say Hoffmann parameters for one material (which has no vacuum energy adjustment) and Cerda parameters for another?

Thanks!

[ewilson]

So I kept looking into my questions and I think I may have made some progress.  Perhaps you can tell me if I have gone astray.

I noticed if you export band structure data to text it appears that the E(k) and the Fermi level look to be defined relative to the vacuum.  Please correct me if I am wrong here.

Say I am doing an EHT calculation with the Hoffmann parameters where no vacuum energy (E_vac) has been set.  Suppose I know from experiment that the distance in eV from conduction band minimum (CBM) to vacuum and that is equal to phi.  Assuming CBM is measured from vaccuum (again correct me if I'm wrong)  then my correct E_vac is:

E_vac = CBM + phi

Sorry, if I'm being pedantic here, I just want to be painfully clear.

[Anders Blom]

This is somewhat tricky territory. Let me start with clarifying some basics.

1) The absolute Fermi level is arbitrary in a bulk system, but it will have meaning on a surface - you can read more about this in the minitutorial on workfunctions.

The Fermi energy is generally placed in the middle of the band gap, and then all Ek energies are always reported relative this level.

3) You can, in general, not mix Huckel parameters from different parameter sets for self-consistent calculations. This is not due to the vacuum level, but due to the occupations which describe the charge transfer in the reference system (for which the parameters were fitted).

[Anders Blom]

Did you check http://quantumwise.com/documents/manuals/latest/ReferenceManual/index.html/chap.atkse.html?

[ewilson]

1)  So just to be sure I understand, you could follow an analagous procedure to find the Fermi level with Huckel that you do with DFT in that tutorial?

3) I've read that portion of the manual pretty thoroughly but I've only glanced through the reference so far.  I understood the part about not using parameters in different sources for SCF calculations.  What I didn't understand was this:

The Hückel parameters have been fitted for non-self-consistent calculations. To use the parameters in self-consistent calculations, the self-consistent onsite shifts must be compensated by a reverse shift of the vacuum_levels.

To first-order non SCF calculations are probably good enough for me right now, but I am curious.

[Anders Blom]

1) Yes
3) Yeah this is tricky. Let's see if we later can make a good description of this, in a tutorial or something...