QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: baizq on July 12, 2013, 16:38

Title: Slater-Koster TB model for a self-defined system
Post by: baizq on July 12, 2013, 16:38

Dear ATK developers,

I would like to make sure whether we can use the Slater-Koster method as implemented in ATK to model a somewhat complex system, e.g., the Ni/Graphene contact system in your paper PRB 85, 165442 (2012). The complexity may include the either contacted or freestanding regions of graphene, the metal edge, and the atomic position change due to relaxation, etc...

Best,
baizq

Title: Re: Slater-Koster TB model for a self-defined system
Post by: Anders Blom on July 13, 2013, 00:03

Technically all the models in ATK-SE can be used to model this system, but finding parameters will be a different story. Defining a Slater-Koster model for a combined metal-covalent system might be rather difficult. The Extended Huckel model could be used, there are Cerda parameters for Ni and graphite that may fit reasonably together. DFTB would be a good choice but there are no published parameters for Ni-C that I'm aware of.

Title: Re: Slater-Koster TB model for a self-defined system
Post by: baizq on July 13, 2013, 04:22

Hi Anders,

Thanks! If no exsiting TB parameters, is it possible to exact from a preceding DFT calculation?

baizq

Title: Re: Slater-Koster TB model for a self-defined system
Post by: Anders Blom on July 14, 2013, 15:44

In principle yes, although it's a lot of work. I have no immediate experience myself but we are working on some tools that can make it easier in the future. You wouldn't do it for the actual system, of course, but you would attempt to parameterize Ni and C separately, and then their interaction.