QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: qnsyhanjiu on March 25, 2011, 03:41

Title: transmission spectrum
Post by: qnsyhanjiu on March 25, 2011, 03:41

Now,I use the 2008.10 atk.I want to plot the transmission spectrum,when I add the spin,how can I plot the spin up and spin down, and when use the An electrode,how can I add spin? For example,the Initial Scaled Spin how  to set up?

Title: Re: transmission spectrum
Post by: zh on March 25, 2011, 03:55

In the manual of ATK2008.10, there is a good tutorial for the spin-polarization of a two-probe system:
http://quantumwise.com/documents/manuals/ATK-2008.10/chap.twoprobespin.html

Please refer to it for the spin setup of a two-probe system.

The spin setup of the electrodes in  a two-probe system is specified by calling:
ElectrodeParameters(
...
electron_density_parameters=electronDensityParameters(
...
initial_scaled_spin =[x,x,x,..,x]
...
)
...
)

For the spin-dependent transmission spectrum, there is a simple example given in the manual to illustrate how to print out the spin-up and spin-down spectra:
http://quantumwise.com/documents/manuals/ATK-2008.10/ref.calculatetransmissionspectrum.html

Title: transmission spectrum
Post by: qnsyhanjiu on March 26, 2011, 13:39

When I  plot the transmission spectrum, the ATK whether has been do the 0 point as the fermi energy? AND how to make the plot more  Smooth ?

Title: Re: transmission spectrum
Post by: zh on March 26, 2011, 13:52

For the first question, the answer is yes. 

You can use smaller energy separation for the energy axis (i.e., more dense energy points) in the calculation of transmission spectrum.

Title: Re: transmission spectrum
Post by: qnsyhanjiu on March 27, 2011, 10:35

when I plot the transmission spectrum,I add the spin,which other Paramaters need set up in order to make the result much more precise.

Title: Re: transmission spectrum
Post by: qnsyhanjiu on March 27, 2011, 10:49

Thank you very much!

Title: Re: transmission spectrum
Post by: zh on March 27, 2011, 10:57

There are several parameters, e.g.,:
1) the size of k-mesh in brillouinZoneIntegrationParameters();
2) the basis_set_parameters;
3) the mesh cut-off in electronDensityParameters.

Please refer to the tutorial:
http://quantumwise.com/documents/tutorials/latest/Fe-MgO-Fe/Fe-MgO-Fe.pdf

Title: Re: transmission spectrum
Post by: qnsyhanjiu on March 28, 2011, 12:17

When I plot the transmission spetrum,the peak value is very tip,how can I make it smooth? Are there some parameter can be set up to make smooth? Thank you very much!

Title: Re: transmission spectrum
Post by: qnsyhanjiu on March 31, 2011, 12:56

I plot the transmission spectrum,when I raise the electrode temperature, if the transmission spectrum will be more smooth?

Title: Re: transmission spectrum
Post by: zh on April 1, 2011, 16:37

Quote from: qnsyhanjiu on March 31, 2011, 12:56

I plot the transmission spectrum,when I raise the electrode temperature, if the transmission spectrum will be more smooth?

The increasing of electrode temperature does not help. Try the use of more dense k-mesh in the calculation of transmission spectrum.

Title: Re: transmission spectrum
Post by: qnsyhanjiu on April 3, 2011, 04:18

Thank you very much! Except for using more dense k-mesh ,if I can use more history steps to make the transmission spectrum to smooth?

Title: Re: transmission spectrum
Post by: zh on April 3, 2011, 04:32

Quote from: qnsyhanjiu on April  3, 2011, 04:18

Except for using more dense k-mesh ,if I can use more history steps to make the transmission spectrum to smooth?

The use of more history steps also cannot make smoothing the transmission spectrum.

Title: Re: transmission spectrum
Post by: qnsyhanjiu on April 4, 2011, 14:15

When I computer the transmission spectrum,it need too much time ,how can make it quickly and self-consistent?