Author Topic: Optical Spectra / Optical Conductivity  (Read 4278 times)

0 Members and 1 Guest are viewing this topic.

Offline Dipankar Saha

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 516
  • Country: in
  • Reputation: 5
    • View Profile
Optical Spectra / Optical Conductivity
« on: March 11, 2019, 12:44 »
Hi,

@ ATK (using KG formula):

1) How correct is the Optical Conductivity calculation,  for a 2D monolayer TMD (semiconducting)?
2) What does the optical spectrum portray, when it does not capture the many-body effects?

Best_
Dipankar Saha

Offline Petr Khomyakov

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1290
  • Country: dk
  • Reputation: 25
    • View Profile
Re: Optical Spectra / Optical Conductivity
« Reply #1 on: March 13, 2019, 11:36 »
The Optical Spectrum analysis object of QuantumATK is not meant to be used for 1D and 2D materials - only for 3D bulk materials.

For the latter, it is as accurate as the DFT electron spectrum, especially, for unoccupied states that are known not to be exact excited states of a many-body system related to single- and two-particle excitation. For that, one could use GW and BSE approximations (not implemented in QuantumATK) for many-body Hamiltonian. 

One may also want to use a hybrid functional approach such as HSE to correct for single-particle excitation spectrum, but one do not account for two-particle excitation, e.g., excitonic effect.

Offline Dipankar Saha

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 516
  • Country: in
  • Reputation: 5
    • View Profile
Re: Optical Spectra / Optical Conductivity
« Reply #2 on: March 22, 2019, 15:29 »
The Optical Spectrum analysis object of QuantumATK is not meant to be used for 1D and 2D materials - only for 3D bulk materials.

Why so? Is it because of the volume (V) involved in KG ?

One may also want to use a hybrid functional approach such as HSE to correct for single-particle excitation spectrum, but one do not account for two-particle excitation, e.g., excitonic effect.
What did you mean by saying_ "such as HSE to correct for single-particle excitation spectrum, but one do not account for two-particle excitation" ?

Thanks and Regards_
Dipankar Saha


Offline Petr Khomyakov

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1290
  • Country: dk
  • Reputation: 25
    • View Profile
Re: Optical Spectra / Optical Conductivity
« Reply #3 on: March 23, 2019, 20:26 »
Why so? Is it because of the volume (V) involved in KG ?
Yes.

What did you mean by saying_ "such as HSE to correct for single-particle excitation spectrum, but one do not account for two-particle excitation" ?
I meant exactly what I wrote that neither DFT nor hybrid functional approach (e.g., HSE functional) are able to describe two-particle excitations, e.g., excitons, see https://en.wikipedia.org/wiki/Exciton or a textbook on Solid State Physics or Many-Body Theory for more information on various excitonic effects.

Offline Dipankar Saha

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 516
  • Country: in
  • Reputation: 5
    • View Profile
Re: Optical Spectra / Optical Conductivity
« Reply #4 on: March 24, 2019, 09:29 »
Why so? Is it because of the volume (V) involved in KG ?
Yes.

So you mean, if one finds a rational for dealing with the volume term (specifically, the out-of-the plane axis), then KG can also be used for calculating optical spectrum (reasonably accurate) of any 2D material.

I don't think it's correct. So out-of-the plane axis is not the only concern. 


Offline Dipankar Saha

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 516
  • Country: in
  • Reputation: 5
    • View Profile
Re: Optical Spectra / Optical Conductivity
« Reply #5 on: March 24, 2019, 09:49 »

What did you mean by saying_ "such as HSE to correct for single-particle excitation spectrum, but one do not account for two-particle excitation" ?
I meant exactly what I wrote that neither DFT nor hybrid functional approach (e.g., HSE functional) are able to describe two-particle excitations, e.g., excitons, see https://en.wikipedia.org/wiki/Exciton or a textbook on Solid State Physics or Many-Body Theory for more information on various excitonic effects.

You have got it all wrong! Focus was not on the word "two-particle excitation".

Previously, you mentioned_ "HSE to correct for single-particle excitation spectrum".  I wanted more details from you, about this 'correct for'!!


Offline Petr Khomyakov

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1290
  • Country: dk
  • Reputation: 25
    • View Profile
Re: Optical Spectra / Optical Conductivity
« Reply #6 on: March 24, 2019, 19:06 »
Previously, you mentioned_ "HSE to correct for single-particle excitation spectrum".  I wanted more details from you, about this 'correct for'!!
Strictly-speaking, ground state DFT does not describe single-particle excitations  of a many-body system, e.g., for periodic solids it means conduction band energies, because ground state means no excitations by definition. In practice, the DFT energy spectrum is used to approximate the single-particle excitation spectrum. Hybrid functionals, e.g., HSE, are not just density functional, going beyond DFT and allowing one to describe the energy spectrum more accurately, e.g., correcting for the Kohn-Sham band gap of ground state DFT.  This is exactly what I meant in my previous post. 

Offline Dipankar Saha

  • Supreme QuantumATK Wizard
  • *****
  • Posts: 516
  • Country: in
  • Reputation: 5
    • View Profile
Re: Optical Spectra / Optical Conductivity
« Reply #7 on: March 24, 2019, 19:49 »
Okay ..... Thank you Petr for the details!  :)

Best_
Dipankar