electrostatic difference potential

[Dipankar Saha]

Hi,
What does it signify.... the is iso-surface/ or,  the cut plane plot of electrostatic difference potential...??  / Perhaps, it will be further helpful, if I get the link of the tutorial which describes this phenomenon in details....

Thanks & Regards__
Dipankar

[zh]

At first, you have to understand the concept of "electrostatic difference potential". A useful message from the use of "electrostatic difference potential" plot may be the the voltage drop:
 http://quantumwise.com/documents/tutorials/latest/MolecularDevice/index.html/chap.iv.html#sect1.iv.vdrop

[Dipankar Saha]

Thanks Zh for your reply....
What I got is that.... if I set a bias voltage for example 1 Volt (and, say there is no concept of gate voltage)....then the contour plot of electrostatic  potential difference will give the spatial distribution of (the corresponding voltage drop + the electrostatic potential at Zero bias) at different points along the channel....../ Is it the fact??

[zh]

The voltage drop is  calculated by the difference  in the "electrostatic difference potential" between  a given bias voltage and zero bias voltage. The electrostatic difference potential is  calculated for a corresponding electron difference density that is defined as the difference between the electron density of the whole system and the electron density with a superposition of atom based densities (http://www.quantumwise.com/documents/manuals/ATK-12.2/ReferenceManual/index.html/chap.atkdft.html#sect2.atkdft.density).

From the the contour plot of electrostatic difference potential for a given bias voltage, it is hard to say something meaningful.

[Dipankar Saha]

So you mean to say...the interpretation will be meaningful... only if we plot the 1D potential variation along the transport direction...( considering the electrostatic potential difference at any given bias voltage, and zero bias voltage) .../ right??

Thanks & Regards__

Dipankar Saha

[Anders Blom]

You can plot it in 3D, no problem with that, but the point is that the potential by itself is less informative - what you can draw conclusions from is the difference in the potential between finite bias and zero bias - that is, the voltage drop.

I guess it's a bit confusing that the potential itself is called a "difference potential", but this is because it's the potential corresponding to the electron density difference between the selfconsistent state and the initial state where all atoms are just put together with their respective atomic electronic densities (no bonding etc).

It should be clear that the difference between two difference potential is equal to the difference between the full potentials.

[Dipankar Saha]

Dr. A. Blom,
Thank you sir... It is very much understandable now.... / Besides, I would like to ask you another question that is, .... in case of the 3D view of the electron difference density plot.... the cut plane is showing the variation in the range 0 --0.048--  0.097 (per amstrong³); where as the iso-surface plot shows -1--0--1 (per armstrong³),.....
1)why this difference?
2)Can it be converted to per cm³ , or, something else, using simple algebra.... ??
3)not the difference density, rather if I want to know the initial carrier conc.  ...., then how can I get it directly from the analyzed results? Is there any analysis...??

Thanks & Regards__
Dipankar Saha

[Anders Blom]

1. The isosurface is shown for one particular value of the quantity you are plotting. However, currently we show both positive and negative values together (not ideal, will be fixed later), hence the scale goes from -x to x, when the isovalue is set to x.

2. Of course: 1 Å^-3 = 10^24 cm^-3.

3. You can get the total electron density from DFT (but not from ATK-SE) at each bias via the ElectronDensity() analysis.

[Dipankar Saha]

Dr. A. Blom,

Sir, that means... Once I have my *.nc file (where I have done my calculations using ATK-DFT).... ; I can invoke the method "ElectronDensity()" with the help of "transmission_spectrum" object to get the required carrier conc. ....... True??

[Anders Blom]

Almost - the carrier density is computed based on the configuration, not the transmission. The script will be simple:

Code: python

conf = nlread("my_nc_file.nc", DeviceConfiguration, object_id="gID000")[-1]
ed = ElectronDensity(conf)
nlsave("ed.nc", ed)

Note that your NC file might contain several DeviceConfigurations - you should keep track of which one is which with the help of the object_id which you can see in VNL. Also, here I save the density in a seprate NC file, but you can put it inside the same file too of course.

[Dipankar Saha]

Okay......  / Thanks a lot sir.....