XanaduAI / QHack2022

QHack—The one-of-a-kind quantum computing hackathon
https://qhack.ai
88 stars 125 forks source link

Exploring the Potential Energy Surface (PES) of a Molecule with a Problem-Inspired Ansatz #97

Open cheechonghian opened 2 years ago

cheechonghian commented 2 years ago

Team Name:

Qanything

Project Description:

The potential energy surface (PES) of a molecular system describes its energy in terms of geometric parameters such as internuclear distances. For diatomic molecules with only one degree of freedom, i.e. one bond length, the surface is reduced to a potential energy curve or Morse potential. Knowledge of the PES allows the exploration of molecular properties, such as atomization or binding energy, and equilibrium bond lengths for various molecules. For more complex molecules, the optimal equilibrium bond lengths and angles for molecules can be studied. Also, the PES along a particular path can be used to map the energy changes in chemical reactions, allowing study of reaction energies and activation barriers.

image

In this project, we explore the PES using a problem-inspired ansatze, Disentangled Unitary Coupled Cluster (UCC) with Singles and Doubles Excitations Rotations. Such UCC methods are notoriously difficult to implement in practice with current NISQ devices due to their need for deep circuits [1] The ansatze are built using Givens rotations [2] – simple particle-preserving variational circuits useful for approximating molecular ground states – as building blocks. We will use PennyLane's built in optimisers to classically optimise the ansatz to the ground state and map out the ground PES's of simple molecules such as the Hydrogen Molecule H2, Helium Dimer He2 and Lithium Hydride LiH on an exact quantum simulator. After optimizing the ansatz, we will apply various IBM noise models and Error Mitigation techniques from Mitig to study how the PESs are affected by the current NISQ devices.

Presentation:

Qanything_Chem_Project_Slides

Source code:

Qanything_Chem_Project_Final

Which challenges/prizes would you like to submit your project for?

  1. Quantum Chemistry Challenge
  2. IBM Qiskit Challenge
  3. Hybrid Algorithms Challenge
  4. Simulation Challenge
  5. Science Challenge

References:

[1] Evangelista, F. A., Chan, G. K.-L., & Scuseria, G. E. (2019). Exact parameterization of fermionic wave functions via unitary coupled cluster theory. The Journal of Chemical Physics, 151(24), 244112.

[2] Arrazola, J. M., Matteo, O. D., Quesada, N., Jahangiri, S., Delgado, A., & Killoran, N. (2021). Universal quantum circuits for quantum chemistry.

isaacdevlugt commented 2 years ago

Thank you for your submission! There's still time to populate your submission with code, presentation material, etc. Please make any final adjustments before the deadline tonight at 17h00 EST!

Good luck!