Author Topic: Simultation of Quantum Dots devices for Quantum Computing  (Read 1568 times)

0 Members and 1 Guest are viewing this topic.

Offline mariosimoni

  • New QuantumATK user
  • *
  • Posts: 2
  • Country: it
  • Reputation: 0
    • View Profile
Dear all,
In the last few years, my main research interests have concerned the development of noisy compact models for quantum computers based on different technologies. A compact model lives at a higher level of abstraction than a physical one and is thought for the simulation of quantum circuits, taking into account the characteristic physical parameters of the devices.
Recently, I have decided to expand the simulation platform and model semiconductor quantum-dot-based quantum computers. I am already developing a noisy compact model for a semiconductor gate-defined QD device.
I am now looking for a physical level simulation toolchain capable of providing the physical input parameters required by the compact models, such as tunnel coupling (or exchange interaction), on-site Coulomb interaction, electron g-factor and valley splitting.
State-of-the-art approaches to simulate solid-state quantum dots for quantum computing usually include an electrostatic simulation (Poisson-Schrodinger) and a quantum mechanical treatment (kp, effective mass or tight binding) to compute single-particle wavefunctions and a post-Hartree Fock (typically, configuration interaction) method to gather many-body wavefunctions starting from previously computed single-particle wavefunctions.
Given that simulating the whole structure on QuantumATK (tight binding) would probably be too computationally expensive, my preliminary idea would be to try to perform the first part (electrostatic simulation) on Sentaurus or COMSOL in order to extract the electrostatic potential profile. Then, I would like to feed this potential profile to QuantumATK and try to simulate (tight binding) only the quantum region of the device (e.g., the 28Si layer in a SiMOS DQD structure) to extract the single-particle wavefunctions.
First of all, I would like to ask you if this workflow makes sense. Then, I would like to know if a potential profile can be provided in input to QuantumATK. I have been looking at the QuantumATK user manual and tutorials, but I was not able to find any answer.

Thanks for your time.

Offline AsifShah

  • QuantumATK Guru
  • ****
  • Posts: 144
  • Country: in
  • Reputation: 2
    • View Profile
Re: Simultation of Quantum Dots devices for Quantum Computing
« Reply #1 on: May 26, 2022, 15:57 »
Once you are done with bringing potential from say sentaurus? What simulation you would carry out on QATK? Will you be using NEGF? If so then I guess you want to use U (potential) into the NEGF equations which I think is not a good method because normally you require self consistent solution of NEGF with poison equation to arrive at final solution. So you will require to extract charge from NEGF into sentuarus & back potential from sentuarus into NEGF until convergence is achieved which I think is not possible to do or is challenging to do unless both devices on sentaurus & QATK resemble to one another.

PS: THis is just a hypothesis , do let me know if there is a way around.

Offline mariosimoni

  • New QuantumATK user
  • *
  • Posts: 2
  • Country: it
  • Reputation: 0
    • View Profile
Re: Simultation of Quantum Dots devices for Quantum Computing
« Reply #2 on: May 28, 2022, 12:28 »
Dear AsifShah,
Thanks for your kind reply and suggestions. My preliminary idea would be to get from QuantumATK single-particle energy states of the quantum dots in steady-state (the potential profile only creates the 2D confinement along the x and y-axis, whereas the confinement along z is usually guaranteed by the heterostructure), so I do not believe I am going to use NEGF. Maybe, I could use a semi-empirical tight-binding method and a bandstructure calculator? I am not sure, and any kind of advice is very welcome.
However, I believe that the self-consistency issue you mentioned will anyway be a problem.NVC

Offline AsifShah

  • QuantumATK Guru
  • ****
  • Posts: 144
  • Country: in
  • Reputation: 2
    • View Profile
Re: Simultation of Quantum Dots devices for Quantum Computing
« Reply #3 on: June 7, 2022, 14:50 »
Dear mariosimoni,
Yes, I if you want to just calculate energies then it should be possible to do it. NEGF would be more comprehensive & detailed though time expensive I guess.