IBM Quantum Team Develops New Framework for Quantum Dynamics Simulations

Researchers from IBM Quantum, IBM Research Zurich, Institute for Theoretical Physics ETH Zurich, and QC Ware Palo Alto and Paris have developed a computational framework for excited-states molecular quantum dynamics simulations. The framework uses quantum computing-based electronic-structure calculations and the fewest-switches surface-hopping method for simulating nuclear dynamics. The team applied their method to simulate the collision reaction between a hydrogen atom and a hydrogen molecule. The results showed that only methods that can capture both weak and strong electron correlation effects can properly describe the nonadiabatic effects that tune the reactive event.

What is the New Computational Framework for Quantum Dynamics Simulations?

A team of researchers from IBM Quantum, IBM Research Zurich, Institute for Theoretical Physics ETH Zurich, and QC Ware Palo Alto and Paris have introduced a novel computational framework for excited-states molecular quantum dynamics simulations. This framework is driven by quantum computing-based electronic-structure calculations. It leverages the fewest-switches surface-hopping method for simulating the nuclear dynamics and calculates the required excited-state transition properties with different flavors of the quantum subspace expansion and quantum equation-of-motion algorithms.

The team applied their method to simulate the collision reaction between a hydrogen atom and a hydrogen molecule. They critically compared the accuracy and efficiency of different quantum subspace expansion and equation-of-motion algorithms. The results showed that only methods that can capture both weak and strong electron correlation effects can properly describe the nonadiabatic effects that tune the reactive event.

How Does Quantum Computing Impact This Framework?

Quantum computing is emerging as a new computational paradigm with the potential of transforming algorithms for electronic structure calculations performed on classical hardware. With currently available noisy quantum hardware, it is improbable that large and complex physics and chemical systems will be entirely simulated on quantum devices. Most likely, mixed-quantum classical algorithms, where only a part of the full problem is solved on the quantum device, will play an important role for near-term quantum simulations.

This class of approaches includes embedding schemes in which the electronic system is partitioned into an easier component that can be successfully described with approximate classically-efficient methods, and a harder subsystem that is solved more accurately on a quantum computer. Additionally, error mitigation techniques remain indispensable for achieving the required level of accuracy. For a successful implementation of such schemes, an efficient integration of quantum and classical algorithms is crucial.

What is the Role of the Variational Quantum Eigensolver (VQE)?

The Variational Quantum Eigensolver (VQE) has appeared in the last decade as a promising hybrid classical-quantum algorithm to run quantum chemical calculations on quantum computers before the thresholds for fault tolerance are reached. The VQE optimizes classically the energy of a given parametrized ansatz state with respect to a fixed Hamiltonian by evaluating the energy cost function on a quantum device. The variational principle guarantees that the energy landscape will be bounded from below by the exact ground state energy of the Hamiltonian independently on the ansatz.

Most of VQE applications focus on ground-state energy calculation, although recent works have investigated the estimation of excitation energies for general Hamiltonians with algorithms tailored to near-term devices. Several strategies have been proposed, inspired by techniques developed for classical quantum-chemical simulations.

What are the Different Approaches to Excited State Calculations?

The extension of variational approaches to excited state calculations represents a crucial step to broaden their application range. In fact, while the electronic ground state of many classes of molecular systems are heuristically known to display weak correlation, strong correlation effects are omnipresent in excited states, even for simple molecular systems. Simulating molecules with many strongly-correlated electrons remains a major challenge for classical computing.

In the present work, the researchers focused on subspace expansion approaches. More specifically, they considered chemically motivated fermionic subspaces for describing the low energy spectrum of chemical systems, where the intermediary states are built applying fermionic excitations on the variational VQE ground state approximation.

How Does This Framework Apply to Molecular Dynamics?

The researchers presented a complete framework for ab initio ground and excited state molecular dynamics (MD) calculations where all relevant observables for the dynamics – excitation energies, nuclear energy gradients, nonadiabatic couplings – are estimated on the quantum processor. They applied this framework to the simple yet nontrivial system of a dihydrogen molecule-hydrogen atom collision process.

The paper is organized as follows: it introduces the prototypical chemical system for molecular dynamics that will be investigated in this work. It then summarizes the main aspects of the fermionic subspace expansion theory and its application to the calculation of electronic excited-state properties. Finally, it discusses theoretical aspects of different formulations of fermionic expansion for estimating these properties.