New TB set for black Phosphorous

[ams_nanolab]

Hi,
Could you add this recent published tight-binding parameters for black Phosphorous to your database (A. N. Rudenko and M. I. Katsnelson Phys. Rev. B 89, 201408(R))? or maybe just help in conversion to appropriate file format , for the "user defined" SK sets 

https://arxiv.org/pdf/1404.0618.pdf

Final published paper

http://journals.aps.org/prb/abstract/10.1103/PhysRevB.89.201408

[Petr Khomyakov]

You may try implementing it for black phosphorus, following the procedure described at http://quantumwise.com/documents/tutorials/ATK-2015/SlaterKosterModel/index.html/chap.carbonpi.html for carbon atom treated with a single-orbital tight-binding model, which is also adopted for phosphorus in the PRB paper you have mentioned.

[Anders Blom]

One problem with this parameter set, and probably a reason we will not include it in ATK as part of the package, is that the cut-off radius d|| for t||(1) and t||(2) are so close to each other. So if you just make tiny changes to the configuration, e.g. strain it a bit, the model probably falls apart, unless you change also the values of d|| in the model. This is perhaps not a problem if you encode your TB model directly for this configuration type, in which case you can maybe identify the interactions by bond angles or by which layer the atoms are in, but in ATK the value of t is determined only and directly from the distance d between the atoms. So, you can easily implement this model, by following the tutorial mentioned, but you have to be careful to use the exact same distances between the atoms in your structure, as used in the TB model.

[Anders Blom]

I would also like to comment that I don't think the TB model provides a particularly good fit to the GW bands... Sure, the tiny region around the band gap is ok, but if you want to do any serious transport simulations for instance, where you apply maybe 0.5 to 1 V bias, this model will not be accurate at all.

[ams_nanolab]

Okay but is it possible to optimize the parameters with ATK? Maybe carry out a optimization to fit the GW/ HSE bands with a starting TB set (deafult)?

[Anders Blom]

That's basically what they did in the paper, so given the simplicity of the model, the parameter set they present is presumably the best possible fit.