can I add a new type classical potential in ATK?

[wot19920302]

Dear QuantumStaff:
         Recently, I read a paper that investigates Si-Ge  heterostructure nanoribbons with passivated hydrogen:http://journals.aps.org/prb/abstract/10.1103/PhysRevB.89.125403.  In this paper, interactions of atoms(Si-H, Ge-H, Si-Ge) can be draw using Keating potential(page 11). However,  I can't find Keating potential model in TremoloX potential classes..    So my question is :
          1. Can I impose this Keating potential to ATK to calculate thermal conductance using NEGF+ATK-Classical  ? If yes, how to do it? If no, how can I calculate thermal conductance of  Si-Ge heterosructure nanoribbon with passivated hydrogen in ATK?(DFT is a method, but it's too time-consuming)
          2. Now I am aware of that there probably two methods to calculate thermal conductance in ATK: NEGF and NEMD, and calculator includes DFT, SE, classical(potential).  DFT is time-consuming, and classical is limited. How about SE?
         3. Is there other way in ATK to calculate thermal conductance  except NEGF and NEMD?
          I am Looking forward to your reply.
         Best wishes

[zh]

1. Without implementation of the Keating potential into ATK, you cannot impose  the calculator to do simulation with such potential. This is quite obvious.   
Choose one of the available classic potentials especially for Si, Ge, and H.

2.  For thermal conductance calculations, you need to think about which contributions you want to calculate.  The NEGF and NEMD in ATK are suitable and efficient for the calculation of the phonon thermal conductance by using the classic potentials. In my personal point of view, the SE method may be suitable for the electronic properties (e.g, the electronic transport) of large system, and unsuitable for the phonon of lattice because the parameters in the SE model are mostly optimized from the electronic structures of some bulk materials.     If you want to simulate the  thermal conductance of electrons in a device by using NEGF,  the use  of NEGF+SE may be a good choice.

3. There are some open source tools (e.g., BoltzTraP and  ShengBTE) for the phonon or electron thermal conductance by post-processing the DFT output (such as forces or eigenvalues) and based on the Boltzmann Transport equations.  You can write your own interface between ATK and these tools for the calculations of thermal conductance. 

[wot19920302]

Thanks for your kindness!.....well, I have another  questions that confuses me:
         can I use Tersoff potential(C,H) to calculate thermal conductance for a molecular device instead of 1D or 2D graphen device where only includes  carbon atoms and hydrogen atoms using NEGF+ATK_classical?
           Best
             

[zh]

I think that you can use them .

[wot19920302]

thank you!