QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: ruyam on November 13, 2014, 13:40

Title: Contour plot of the C component of the linear response current density
Post by: ruyam on November 13, 2014, 13:40

Dear Anders Sir,

In your paper on Nickel Graphene interface, there is contour plot(Fig-4) with the caption as the subject. I was wondering how to get that contour plot in ATK?
PFA the fig for reference.
 

Title: Re: Contour plot of the C component of the linear response current density
Post by: ruyam on November 17, 2014, 09:01

Dear QW Staff,

Pls Reply !

Title: Re: Contour plot of the C component of the linear response current density
Post by: kstokbro on November 27, 2014, 16:26

Please find attached the script we used

Title: Re: Contour plot of the C component of the linear response current density
Post by: ruyam on March 2, 2015, 10:06

1). In the file "cd_side.py" attached above by Dr. Stokbro, I would like to understand what does the following line in the code means:

Code

cd_xz = numpy.apply_over_axes(numpy.sum, cd[2,:,:,:] ,[2]).flatten()

Actually, when I run the program for my structure, I get the following error:

NL.ComputerScienceUtilities.Exceptions.NLValueError: Only 3 indices can be specified.

Title: Re: Contour plot of the C component of the linear response current density
Post by: ruyam on March 3, 2015, 07:15

Dear QW staff,

Pls reply!

Title: Re: Contour plot of the C component of the linear response current density
Post by: Umberto Martinez on March 3, 2015, 10:24

Please use the example scripts reported in the reference manual for the Current Density analysis:
http://www.quantumwise.com/documents/manuals/latest/ReferenceManual/index.html/ref.currentdensity.html
In particular, see the second example usage.

the python line you indicate will sum over the Y axis your current density data points.
see http://docs.scipy.org/doc/numpy/reference/generated/numpy.apply_over_axes.html
as the error says cd has only three indices.

Title: Re: Contour plot of the C component of the linear response current density
Post by: ruyam on March 10, 2015, 18:49

Dear QW Staff,

In the current density contour plot, in the metal region if we see for the same metal atom, current density is red coloured(positive) as well as blue coloured(negative). What does this signify or what am I failing to see?

Title: Re: Contour plot of the C component of the linear response current density
Post by: sathyajit on July 15, 2015, 16:28

Hi,

I am using the scripts attached here and have obtained the png file.
Can you please help me interpret the reason why the red regions and blue regions are alternating ? (positive and negative current densities)

Regards
Sathyajit

Title: Re: Contour plot of the C component of the linear response current density
Post by: Jess Wellendorff on July 20, 2015, 13:50

Im very sorry, but detailed discussions of specific scientific results are usually out of the scope of this forum.

Title: Re: Contour plot of the C component of the linear response current density
Post by: sathyajit on August 28, 2015, 15:40

Hi,

I wanted to know what is the significance of spectral weight in the definition given at
http://www.quantumwise.com/documents/manuals/latest/ReferenceManual/index.html/ref.currentdensity.html

Also what do you mean by
"positive weight corresponds to calculating right to left going modes (positive current), while a negative weight corresponds to calculating left to right going modes (negative current)."

I am not able to understand what you mean by positive current and negative current w.r.t. the plot given in the post.

Title: Re: Contour plot of the C component of the linear response current density
Post by: sathyajit on August 31, 2015, 16:20

Pls reply...

Title: Re: Contour plot of the C component of the linear response current density
Post by: Jess Wellendorff on September 2, 2015, 12:52

The spectral weight is defined as the difference between the right and left electrode Fermi functions. At zero bias between the electrodes, there is no current. However, we would still like to be able to calculate electronic properties such as transmission spectra etc. Using the option "energy_weights = Right," allows you to do this. "Right" means that you only consider left-to-right going modes (positive current", while "Left" means the opposite.

Title: Re: Contour plot of the C component of the linear response current density
Post by: asanchez on October 6, 2015, 16:57

There seems to be a contradiction.

Jess: you're saying that "energy_weights = Right" considers states going "TO" the right. The documentation states:

"Left/Right gives an energy independent weight of 1, for current carrying states from Left/Right."

So that would mean energy_weights = Right would be states coming "FROM" the right?

Am I interpreting something the wrong way? Which one's the way it works in the code?

PS. Maybe "Right" means electrons coming FROM the right (i.e. current flow TO the right)?

Title: Re: Contour plot of the C component of the linear response current density
Post by: Jess Wellendorff on October 7, 2015, 12:12

Let me try to reformulate:
- positive current means electron modes propagating from right to left in the device.
- "energy_weights"=Right gives a positive weight (+1) --> only right to left going modes are considered (positive current).
- "energy_weights"=Left gives a negative weight (-1)  --> only left to right going modes are considered (negative current).