A systematic variational approach to band theory in a quantum computer

TL;DR

This paper introduces a hybrid quantum-classical algorithm for calculating band structures in periodic systems, demonstrating its effectiveness on simulators and IBM quantum computers, and analyzing its scalability and noise resilience.

Contribution

The paper presents a novel hybrid quantum-classical algorithm for band structure calculation in periodic systems, addressing limitations of current quantum hardware and providing scalability analysis.

Findings

Algorithm performs reliably on low-noise devices.

Functional with low precision on current noisy quantum computers.

Complexity scales as Ω(M^3), similar to classical methods.

Abstract

Quantum computers promise to revolutionize our ability to simulate molecules, and cloud-based hardware is becoming increasingly accessible to a wide body of researchers. Algorithms such as Quantum Phase Estimation and the Variational Quantum Eigensolver are being actively developed and demonstrated in small systems. However, extremely limited qubit count and low fidelity seriously limit useful applications, especially in the crystalline phase, where compact orbital bases are difficult to develop. To address this difficulty, we present a hybrid quantum-classical algorithm to solve the band structure of any periodic system described by an adequate tight-binding model. We showcase our algorithm by computing the band structure of a simple-cubic crystal with one $s$ and three $p$ orbitals per site (a simple model for Polonium) using simulators with increasingly realistic levels of noise and culminating with calculations on IBM quantum computers. Our results show that the algorithm is reliable in a low-noise device, functional with low precision on present-day noisy quantum computers, and displays a complexity that scales as $\Omega(M^3)$ with the number $M$ of tight-binding orbitals per unit-cell, similarly to its classical counterparts. Our simulations offer a new insight into the ``quantum'' mindset and demonstrate how the algorithms under active development today can be optimized in special cases, such as band structure calculations.