script for calculating Seebeck coeffient, thermal conductance, and ZT

[marmotte]

So we don't need electrodes for the transmission spectrum ?

[Anders Blom]

The comment about computing the transmission from bulk should not be taken out of context. It is only valid for perfectly periodic systems. If you are looking to compute the influence on the transport from a single defect (or several, but not a whole periodic arrangement of them), then you need to set it up as a device.

[marmotte]

Dear ATK team,

I'd like to be sure that with ATK, we can calculate the thermoelectric properties : Seebeck coefficient, the thermal conductance, and then ZT etc.... Following your advice, I have to follow the steps of http://quantumwise.com/publications/tutorials/mini-tutorials/167 in order to get the transmission then calculate the thermoelectric properties using the script. I'd like to know if I have to follow all the steps, such as :

 
   1. Remove one layer, leaving only 3 atoms in the unit cell.
    2. Then perform the same transformation to the orthorhombic structure as for graphene above.
   3.  As an additional step, open Coordinate Tools>Center and center the atoms in all directions.

My system is somehow lamellar ( layers of atomes) but in Bulk. are these steps mandatory ?

I will appreciate much your help

Best

[Anders Blom]

Hard to say, without knowing the precise system. The steps are not a universal recipe, they are a method to achieve the necessary structure, which needs to follow the universal rule of having the AB plane perpendicular to C which in turn is parallel to Z, the transport direction.

[perfetti]

Dear Everyone,
        I am just wondering if it's capable for ATK to calculate the thermal conductance for system with phonon transport now? Then it should be more realistic.
     

[Anders Blom]

We are working on this feature, some part will be available in 13.8 coming out during the summer 2013.

[wuxing]

Please find an updated script which is more stable than the previous version.

Dear Dr. Stockbro,

Does the script for calculating Seebeck coeffient, thermal conductance and ZT not consider the influence of electron chemical potential? How to get the ZT with different chemical potential?

Thank you!

Wu-Xing Zhou

[kstokbro]

The plugin use pertubation theory, i.e. it is correct for linear responce, i.e. at zero bias where the is a common electrochemical potential.

At finite bias you may calculate the current directly, inputing the temperature of the electrodes. Using finite difference you can then get
dI/dT, dI/DV, and D2I/Dv/dT.