Efficient Excited-State Calculations for Molecules Based on Contextual Subspace Method and Symmetry Optimizations

TL;DR

This paper introduces a resource-efficient quantum computing framework combining the contextual subspace method with VQD and symmetry optimizations to improve excited-state calculations for molecules on NISQ hardware.

Contribution

It presents a novel integration of the CS method with VQD and a spin-conserving ansatz, reducing qubit and computational resources for excited-state molecular calculations.

Findings

The combined method effectively reduces qubit requirements.

The spin-conserving ansatz exploits symmetry for resource savings.

Optimization iterations are reduced by up to 3 times with similar circuit depth.

Abstract

Quantum computing methods for excited-state calculations remain underexplored in Noisy Intermediate-Scale Quantum (NISQ) hardware, despite their critical role in photochemistry and material science. Herein, we propose a resource-efficient framework that integrates the contextual subspace (CS) method with the Variational Quantum Deflation (VQD) algorithm to enable systematic excited-state calculations for molecules while reducing qubit requirements. On the basis of the numerical results, we find that it is unproblematic to utilize this combination in calculating the excited state to reduce qubits. Furthermore, we demonstrate that the implementation of a spin-conserving hardware-efficient ansatz, namely the $\mathcal{N}(\theta_x,\theta_y,\theta_z)$ block ansatz, allows exploitation of spin symmetry within the projected subspace, thereby achieving further reductions in computational resource demands. Compared to the commonly used $R_{y}R_{z}$ ansatz, using the $\mathcal{N}(\theta_x,\theta_y,\theta_z)$ ansatz can reduce the number of optimization iterations by up to 3 times at a similar circuit depth.