QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: ramkrishna on January 17, 2014, 22:01

Title: Central region doping of a device configuration
Post by: ramkrishna on January 17, 2014, 22:01
Dear Sir,
   Actually I want to clarify myself about the doping of the central region of a device configuration.

In the release note of 13.8, it is saying that,
"Doping can also be introduced in the central region (earlier it was only possible to do for the electrodes) via so-called compensation charges. This allows for simulations of a wide variety of semiconductor devices, like p-n junctions, p-i-n doped Si nanowires, etc, without the need to introduce explicit dopant atoms. Doping charge can now be set in Scripter; the compensation."

Now, when I open a Scripter of a device configuration, I can only see options to put charge in the  electrodes not in the central region. So, I am confused here about the doping of the central region.
 So, I was going through the manual to understand the compensation charge from "http://quantumwise.com/documents/manuals/latest/ReferenceManual/index.html/ref.atomiccompensationcharge.html".
It is written that "For a device configuration the atomic compensation charge of the equivalent electrode atoms should be similar to the electrode. For the rest of the atoms there are no requirements, and the atomic compensation charge can for instance be used to model a certain doping distribution of the device."

So, basically I want to know the exact procedure to put charge in the central region in ATK. Previously, for the electrodes the commands was like

Code
# -------------------------------------------------------------
# Calculator
# -------------------------------------------------------------
#----------------------------------------
# Electrode Calculators
#----------------------------------------
left_electrode_calculator = LCAOCalculator(
    charge=0.001,
    )

right_electrode_calculator = LCAOCalculator(
    charge=-0.001,
    )

#----------------------------------------
# Device Calculator
#----------------------------------------

So what will be the command to put charge in the central region also?

If it is done by just putting charge in the electrodes, then that charge will be copied only for the equivalent atoms of electrodes in the central region and not the whole central region, am I right? If so, can you please explain, how this charge is being added to the equivalent atoms in the new version of ATK with respect to the previous one where it will not take care of the charge on the equivalent atoms?
Is it not possible to put charge in whole central region?

Thanks
Ramkrishna



Title: Re: Central region doping of a device configuration
Post by: ramkrishna on January 22, 2014, 16:54
Dear Sir,
     If you discuss a little bit about these quarries, it will be very much helpful to understand the doping of the central region.

Thanks
Ramkrishna
Title: Re: Central region doping of a device configuration
Post by: Anders Blom on January 25, 2014, 23:59
You can't add free charge to the central region, since it will just disappear in the selfconsistent cycle. What you need to do is add compensation charge, which is localized (bound to the atom cores) of the opposite sign, then this will attract carriers and create an effective doping effect. The same compensation charge should be added to the electrodes to make them neutral.

A tutorial on this is coming soon - unfortunately it requires a bit of scripting so it's not easy to explain in a few words.
Title: Re: Central region doping of a device configuration
Post by: ramkrishna on January 30, 2014, 00:12
I am eagerly waiting for that new tutorial :)

Thanks
Ramkrishna
Title: Re: Central region doping of a device configuration
Post by: Anders Blom on March 15, 2014, 20:09
Yes, that tutorial is long overdue... For now, I have attached 2 examples, for a doped bulk structure and a doped device (a Si p-i-n junction). That's a start at least :)

For bulk it's quite straightforward: you introduce a certain amount of charge, corresponding to the desired doping level. Then you neutralize this with compensation charges bound to the atomic cores to have a neutral system.

For the device, you dope the electrodes just like bulk. Then, the same amount of compensation charge is applied to any atom in the central region that we want to consider part of the doping region. This gives us the same level of doping in this part of the central region as the corresponding electrode (provided the crystal structure is the same) but indirectly - the corresponding additional potential will attract/repel the same amount of electrodes as it does in the electrodes.

You can't set the compensation charge in the Scripter - it's often not a trivial process because you have to make several choices (like, don't compensate H atoms, etc), so in most cases I imagine people will use scripting for this. But in 14.8 we will probably design a new Script Generator that can take it into account.

Hope this helps you!
Title: Re: Central region doping of a device configuration
Post by: mantun on June 6, 2014, 09:30
"For a device configuration the atomic compensation charge of the equivalent electrode atoms should be similar to the electrode. For the rest of the atoms there are no requirements, and the atomic compensation charge can for instance be used to model a certain doping distribution of the device."
What does this sentence means？Does it means the electrode copies should has the same compensation charge as electrodes, ie. the left electrode copy should have the same compensation charge with left electrode and the right electrode copy should have the same compensatioan charge with right electrode? But i found the sum of charge and compensation charge in left  electrode is zero, and the sum in right electrode is zero too. But the the central region may have extra compensation charge, this will lead to the two-probe device charged.
Can you explain it? thanks in advance!