QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: ruyam on October 7, 2014, 19:37

Title: Electron Temperature Significance
Post by: ruyam on October 7, 2014, 19:37
Dear Sir,

I read from the forum that the electron temperature as no physical significance and is only there for smearing the fermi function, etc. Is my understanding correct?

If yes then, there should not be any change in current on changing this electron temperature. But, upon changing electron temperature for my device and calculating current at different temperature's keeping all other parameters same, I found the current changing on changing electron temperature. PFA a plot of it.

What does this electron temperature mean ? How should I understand its importance in my calculations?
Title: Re: Electron Temperature Significance
Post by: Anders Blom on October 7, 2014, 22:30
The temperature has one more effect, but only in a transmission spectrum calculation. If you increase the electron temperature you broaden the range of occupied states, so you create "hot electrons" which can tunnel more easily across a potential barrier. This is called thermionic emission.
Title: Re: Electron Temperature Significance
Post by: ruyam on October 8, 2014, 07:53
So, does this mean that if I want to calculate transmission spectrum and current in my device under some bias, I have to keep the temperature to 300K only ?

Secondly, As increasing temperature was said to make convergence easier(faster), but has other effects in transmission spectrum calculation, then how should I work towards smearing fermi width and making my structure converge faster ?
 
Title: Re: Electron Temperature Significance
Post by: Anders Blom on October 8, 2014, 14:54
The temperature is actually specified in two separate places - one for the Fermi function broadening during the self-consistent calculation (this is where you might control the convergence), and separately for the current evaluation, which is where you get the temperature effect on the current.
Title: Re: Electron Temperature Significance
Post by: ruyam on October 8, 2014, 16:16
Dear Sir,

PFA my python file. I have increased electron temperature in SCF settings and have kept the temperature in electrode as 300K itself. But still I get different temperatures. How to go about it (I want to increase temperature to speed up convergence and not have its effect in transmission spectrum and current) ?  ???

Title: Re: Electron Temperature Significance
Post by: Anders Blom on October 8, 2014, 16:28
Cf. http://quantumwise.com/documents/tutorials/latest/GrapheneDevice/index.html/chap.conductance.html#chap.conductance.sect1.conductance