QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: wot19920302 on April 10, 2016, 12:07

Title: The mechanism of thermoelectric coefficient plugin
Post by: wot19920302 on April 10, 2016, 12:07

Dear staffs:
           I noticed that we can get thermoelectric coefficients (seebeck coefficient,S, conductance, G, phonon contribution to conductance Kph and electron  contribution to conductance Ke) as the functions of fermilevel or temperature in the thermoelectric coefficient plugin. It seems that thermoelectric coefficient plugin can calculate these thermoelectric coefficients in linear response regime using  eq1 (fig1),  I think eq1 is derived from eq2 (fig2) .  If I want to shift fermilevel (assume E=0.3) in  thermoelectric coefficient plugin, does it mean Ef^L in eq1 is 0.08? IF I want to adjust temperature (assume T=500K)  in thermoelectric coefficient plugin , does it mean T_R in eq1 is 500K?  Here is the case in fig3. By the way, how  do you define an energy window for finite T_R=T_L,  a certain constant or a variant as T?
            Another question: in general , we can calculate the current using trasmission spetrum at 0K, the bias window is from -eV/2 to eV/2 . When at a finite temperature , however , the range of integration  extends because of the change of the fermi distribution, how do we set the bias window at the finite temperature?
            Looking forward to your replies
            Best regards
         

Title: Re: The mechanism of thermoelectric coefficient plugin
Post by: Jess Wellendorff on April 11, 2016, 11:04

1) I don't quite understand your question about shifting the Fermi level, but it essentially corresponds to doping the device.
2) The specified temperature is the temperature that enters the Fermi distribution for the electrons in the electrodes. The default temperature is the average electrode temperature.
3) For the energy window, I think you simply need it to be wide enough to accommodate the temperature dependence. You set the energy window in the Transmission widget.

Title: Re: The mechanism of thermoelectric coefficient plugin
Post by: wot19920302 on April 11, 2016, 14:52

Code

1) I don't quite understand your question about shifting the Fermi level, but it essentially corresponds to doping the device.
2) The specified temperature is the temperature that enters the Fermi distribution for the electrons in the electrodes. The default temperature is the average electrode temperature.
3) For the energy window, I think you simply need it to be wide enough to accommodate the temperature dependence. You set the energy window in the Transmission widget.

  well, I mean 1).if I want to get conductance G  in fig1 (△E_f=0.3eV ,T=500K),   ATK will  calculate conductance G using eq1(fig2) when E_F^L=0.3eV , T_R=500K ,is that correct?
                           2). in this link: http://docs.quantumwise.com/tutorials/graphene_nanoribbon_device.html?highlight=linear%20response  it says "At TL=TR=0TL=TR=0, the conductance is determined by the transmission coefficient at the Fermi Level, while for finite electrode temperatures, the conductance depends on the value of the transmission coefficient in an energy window around the Fermi level.", how do you define  an energy window around the Fermi level  for  finite electrode temperatures?
                            by the way, is eq1 derived from eq2 (fig3) ?