Resource-Efficient Quantum Circuits for Molecular Simulations: A Case Study of Umbrella Inversion in Ammonia

TL;DR

This paper introduces a resource-efficient quantum circuit design for molecular simulations, significantly reducing circuit depth and gate count while maintaining accuracy, demonstrated through the ammonia umbrella inversion case study.

Contribution

The authors present a novel quantum circuit that cuts circuit depth and two-qubit gates by about 60%, improving performance on noisy quantum devices for molecular energy calculations.

Findings

Reduced circuit depth and gate count by approximately 60%.

Achieved near-chemical accuracy in ground state energy predictions.

Demonstrated robustness against device noise in molecular simulations.

Abstract

We conducted a thorough evaluation of various state-of-the-art strategies to prepare the ground state wavefunction of a system on a quantum computer, specifically within the framework of variational quantum eigensolver (VQE). Despite the advantages of VQE and its variants, the current quantum computational chemistry calculations often provide inaccurate results for larger molecules, mainly due to the polynomial growth in the depth of quantum circuits and the number of two-qubit gates, such as CNOT gates. To alleviate this problem, we aim to design efficient quantum circuits that would outperform the existing ones on the current noisy quantum devices. In this study, we designed a novel quantum circuit that reduces the required circuit depth and number of two-qubit entangling gates by about 60%, while retaining the accuracy of the ground state energies close to the chemical accuracy. Moreover, even in the presence of device noise, these novel shallower circuits yielded substantially low error rates than the existing approaches for predicting the ground state energies of molecules. By considering the umbrella inversion process in ammonia molecule as an example, we demonstrated the advantages of this new approach and estimated the energy barrier for the inversion process.