Author Topic: How is it possible to design two heterogeneous electrode in vnl2008.10  (Read 6595 times)

0 Members and 1 Guest are viewing this topic.

Offline shank

  • Heavy QuantumATK user
  • ***
  • Posts: 33
  • Reputation: 0
    • View Profile
Dear all
I want to make two electrode one is Si(111) (ntype highly doped) and other is Hg.How shall I make these electrodes for I-V.
with regards.
shank

Offline zh

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1141
  • Reputation: 24
    • View Profile
It can not be built in an intelligent manner. The two different electrodes are built separately using the "Atomic Manipulator" and they are combined together by hand. You could try to build it  by the following outlined steps:
1) build a highly doped (n-type) bulk Si
    1a) build a supercell of Si
    1b) substitute  one or more Si atoms with other element (i.e., the dopant). Of course, the impurity could be interstitial defects, in this situation, insert other element (dopant) into supercell of Si
2) build a two-probe system of n-type Si
   2a). import the built n-type bulk Si into "Atomic Manipulator" and then build a two-probe system without conductor.
  2b). export the configuration into a script file
  2c). extract the configuration for left electrode by hand (i.e., copy&paste), meanwhile the configuration for the left surface layers should be also extracted by hand.
3) build a bulk Hg
 3a) build a supercell of Hg. The area in the directions perpendicular to the transport direction should be same to the one of n-doped Si
4) build a two-probe system of Hg
 4a). import the built Hg sample into "Atomic Manipulator" and then build a two-probe system without conductor.
  4b). export the configuration into a script file
  4c). extract the configuration for right electrode by hand (i.e., copy&paste), meanwhile the configuration for the right surface layers should be also extracted by hand.
5). manually combine those extracted in 2c) and 4c) steps
  5a). put those obtained in 2c) and 4c) into a blank script file
  5b). edit them to construct a two-probe configuration by hand. Especially, the distance between the left surface layer and the right surface layer should be chosen properly. Of course, this parameter need to be optimized.

 

Offline shank

  • Heavy QuantumATK user
  • ***
  • Posts: 33
  • Reputation: 0
    • View Profile
Thanks for reply.But I am unable to cleave Hg as after right click ,cleave option is inactive .What is the reason?
with regards

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5576
  • Country: dk
  • Reputation: 96
    • View Profile
    • QuantumATK at Synopsys
The mercury crystal you are referring to is the rhombohedral structure in the Crystal Cupboard, I guess. Cleaving rhombohedral structures is currently not possible in VNL; the "irregular" structure makes it very difficult numerically to find a finite surface cell. If you manage to cleave it manually, you can create the corresponding electrode by hand and set up the two-probe system in a NanoLanguage script.

Do you have an experimental or theoretical reference for the system you wish to consider?

Offline shank

  • Heavy QuantumATK user
  • ***
  • Posts: 33
  • Reputation: 0
    • View Profile
Yes, structure of Hg  is rhombohedral.I have experimental I-V for the system.I want to find out what is the cause of the current flow.Is it HOMO /LUMO taking part in the current flow.In the system one electrode is highly doped ntype Si and other end Hg drop was usued for soft contact.One more query, how i decide the number of step for SCF.
Thanking you
with regards

Offline Anders Blom

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 5576
  • Country: dk
  • Reputation: 96
    • View Profile
    • QuantumATK at Synopsys
I guess in the experimental situation you refer to, the Hg electrode is not crystalline, or at least not a simple cleaved surface, but rather some kind of cluster. You would need to find a clever way to set this up in ATK based on physical and geometrical considerations rather than purely mathematical. Perhaps one can use a cone-shaped or pyramidal electrode.

The number of SCF steps is not pre-determined; the loop runs until it converges. There is a max limit, so that the calculation does not run forever in case it doesn't converge; the default max is 100 steps, but this can easily be changed in the iterationControlParameters().

Offline shank

  • Heavy QuantumATK user
  • ***
  • Posts: 33
  • Reputation: 0
    • View Profile
Thanks for reply.One thing i like to know when SCF does not converge,what parameters one should look into.
with regards

Offline ugglebot

  • Regular QuantumATK user
  • **
  • Posts: 22
  • Country: se
  • Reputation: 1
    • View Profile
Have a look at the tutorial regarding convergence under Publications>Tutorials on our website. Also, there are several discussions on the Forum.

Offline shank

  • Heavy QuantumATK user
  • ***
  • Posts: 33
  • Reputation: 0
    • View Profile
Thanks.