electron-phonon interaction in two probe system

[rebeda]

In the tutorial given in the link below to calculate the IETS of  hydrogen molecule-
http://quantumwise.com/publications/tutorials/item/877-inelastic-electron-spectroscopy-of-h-molecule-between-1d-au-chains
1. why vibrational modes were not assign for the negative bias?
2. I calculated the same for my system[Au(111)-single molecular magnet-Au(111)] within the bias -0.1 to 0.1 volt and I got asymmetric IETS. How can I find the modes for the negative bias?
     In my vibrational analyser, mode zero lie at -53.99 meV which is -0.05399 volt, what about the IETS peaks within -0.1 to -0.05399 volt? How can we identify them?

Please help me. I will be waiting for your kind reply.
Regards,
Rebeda

[rebeda]

Dear Sir,

According to my knowledge transmission, being a probability, can have minimum value of 0 and maximum of 1 (though it depends on the no. of channels available for conduction) and hence,  is always positive, but refering to one of your quantumwise tutorial-
link: http://quantumwise.com/support/tutorials/item/877-inelastic-electron-spectroscopy-of-h-molecule-between-1d-au-chains
I observe negative inelastic transmission at mode 0 and mode 5 in the IETS study of H molecule between 1D Au chains. I am not able to interpret this "negative inelastic transmission". In other words, how transmission can have negative values in this case, though for IETS, one can understand to have both positive and negative values.
Will you please help me?

Regards,
Rebeda

[rebeda]

Dear sir,
I have some confusion on the thermoelectric transport properties in vnl window. What does T really stand for???

1. does it stand for average temperature (an average of hot and cold electrodes temperature of the two probe device)?
2. does it stand for the environment temperature of the device?

looking forward to your reply.

regards,
Rebead

[Daniele Stradi]

Hello,

- the IETS should be symmetric, so the modes at negative bias are the same as those at positive bias. Notice that the IETS is the normalized second derivative of the transmission spectrum, hence it is not required to have values between 0 and 1.

- The fact that you get an asymmetric IETS could indicate that the system is not converged, however, I cannot say anything more, without looking at the actual scripts.

- I suggest you to read the papers about the underlying methodology to the IETS calculations, which will help you in understand how the spectra should be understood and analysed:
https://journals.aps.org/prb/abstract/10.1103/PhysRevB.75.205413

- Regarding the thermoelectric transport properties, if I understood your question correctly, T should be the temperature which enters the the formula of the thermoelectric figure of merit: http://docs.quantumwise.com/tutorials/phonon_calcs/phonon_calcs.html

Regards,
Daniele.

[rebeda]

Thank you  Daniele Stradi for your quick response.

The T in the formula is introduced to make it a dimensionless value figure of merit "ZT". Usually thermoelectric behaviour is achieved when we make a temperature gradient/difference. In our case (two probe junction) the two electrodes are maintained one at cold (T_C) and the other at hot (T_H) temperature.

So, when we enter the temperature values to calculate the thermoelectic coefficients (in the vnl window) is -
         
1. T = T_H-T_C  ?
or
2. T=(T_H-T_C)/2 ?

Lastly, in atk calculations  what are the values of T_C and T_H ?

Awaiting for your kind reply.

Regards,
Rebeda

[Petr Khomyakov]

The temperature in the formula for ZT is given as an arithmetical average, T=(T_H + T_C)/2.

[rebeda]

Thank you so much for your reply it cleared my doubt.

T=(T_H + T_C)/2. 
In the above formula either T_C or T_H are kept constant in the entire atk calculations . If so, what is the constant temperature value?
 I need these values to calculate the maximum effeciency.

Is T_C = 4.2K (liquid helium temperature) or T_C = 0K ??

Awaiting for your useful comments.

Regards,
Rebeda

[Petr Khomyakov]

I am not sure I understand your problem. The T temperature is for ZT-calculation only; ZT is calculated in a post-processing manner after all the physical quantities in the ZT formula have been computed. What are the actual ATK calculations you are referring to?

[rebeda]

As I have mentioned before seebeck effect is realized when there is a temperature difference (in our two probe system, the electrodes are maintained one at cold T_Cand other at hot T_H temperature ). And ZT is directly proportional to seebeck coefficient by formula.

So, My question is, when ATK perform thermoelectric coefficient calculations , what T_C /T_H value it takes ?

As an example, when we want to find ZT experimentally using STM, the STM tip and the sample are maintained at different temperatures, one at cold (T_C) and other at hot (T_H) temperature. In this case the tip is (T_C) and the sample temperature is (T_H).

Similarly, in atk theoretical calculations T_C  and T_H must be assigned with some temperature values in which we keep on changing one of the temperature (say T_H ) to calculate thermoelectric coefficients at variable temperature.

I hope you might understand my question. I have added some link of articles to help undestand the problem.

http://www.pnas.org/content/112/27/8205.full.pdf
http://dx.doi.org.sci-hub.ac/10.1063/1.4958999

regards,
Rebeda

[Petr Khomyakov]

You may see one example of how the thermoelectrics calculations can be done in ATK in the following tutorials at http://docs.quantumwise.com/tutorials/thermoelectrics_cnt_isotope/thermoelectrics_cnt_isotope.html and http://docs.quantumwise.com/tutorials/phonon_calcs/phonon_calcs.html.

The calculations are done in the linear response regime, and no T_c and T_h are explicitly introduced in the ATK calculations. To compute the thermoelectric coefficients and ZT in a post-processing manner, one can use a thermoelectrics plugin as discussed in the tutorials, and the temperature T=(T_c+T_h)/2 is to be manually set in the widget. 

In ATK, you may also calculate interfacial thermal conductance, using a molecular dynamics (MD) approach  as described in http://docs.quantumwise.com/tutorials/interfacial_thermal_conductance/interfacial_thermal_conductance.html. In this case, the temperature gradient across the interface is explicitly taken into account in the non-equlibrium MD simulations.   

[rebeda]

Dear Sir,

It is great that atk-vnl is upgraded everytime with the upcoming of every new versions. But one big problem is that, we are not able to access the new versions with the previous  xxx.nc files calculated from lower versions. Some of the device calculations are time consuming and we rather not repeat the same calculations in the new versions. Due to it we are not able to utilize some of the new upgraded features which may help in better analysis of our calculated datas. Would you suggest a way to solve this issue???

Waiting for your response.

Regards,
Rebeda

[Anders Blom]

we are not able to access the new versions with the previous  xxx.nc files calculated from lower versions

We make a lot of effort to ensure new versions can read old NC files. I you have any other experience, please contact support with the specifics, and we will help you out.