QuantumATK Forum
QuantumATK => General Questions and Answers => Topic started by: SAKINA on December 15, 2015, 04:19
-
Hai, Good morning :D :D
I have introduce strain in following structure. But when adding electrode the electrode are also in strain. Does it consider defect electrode? How to make sure that electrode are not strain? How can i run structure that are strain on channel only without affecting the electrode?
Thanks in advance :)
-
The DeviceConfiguration is composed of three distinct regions: left electrode, central region, and right electrode. However, they are not completely independent. The central region is in itself divided into three parts: left electrode extension, scattering region, and right electrode extension. See also http://www.quantumwise.com/documents/manuals/latest/ReferenceManual/index.html/chap.negf.html (http://www.quantumwise.com/documents/manuals/latest/ReferenceManual/index.html/chap.negf.html).
The electrode extensions MUST perfectly match the electrodes, so they must be copies of each other. Otherwise, the computational scheme required for NEGF falls apart. So if you want to consider strain effects, but do not want the electrodes to be strained themselves, you need to strain the scattering region only.
-
Thank you
However I did strain at the scattering region but since electrode extension are still consider scattering region the electrode are copied from the scattering region. what and how should i do to avoid the electrode to be strain?
-
Well, you need to come up with a central region that consists of unstrained electrode extensions and a strained scattering region. I suggest you first create two unstrained ribbons (to be used as electrodes) and then a strained ribbon for the scattering region. Then somehow interface these three building blocks, perhaps with some geometry optimization. Right now, you have a few hydrogen atoms in the middle of the structure that almost overlap - these will surely move away from each other during geometry optimization.
-
Thank you
Now i will try to use the method that you suggested :)
-
Hai,
I have try using the merging method that you have shown. But still it not working. Is there any way to make the electrode to be independent without affecting the scattering region. Because when electrode is defected this is what IV that I have obtained and i am not sure the reason of the IV is due to the constriction or the defected electrode or strain?
-
You seem to employ both strain and a significant defect in the middle of the ribbon. If you want the electrodes to stay fixed, you could use the Bulk Tools -> Stretch Cell tool to impose the strain on the central region while making sure that the electrode extension atoms are "moved" rather than "stretched". Then do geometry optimization of the central region with dixed constraints on the electrode extension atoms.