CSU team proposes multi-scale quantum chemical computing framework

Recently, Academician Jinlong Yang's team at the Chinese University of Science and Technology (CSU) has proposed a multiscale quantum algorithm for computational chemistry, which provides a new idea for solving realized chemical and material problems using noise-containing intermediate quantum (NISQ) computers.

The results are published in the journal Chemical Science of the Royal Society of Chemistry under the title Multiscale quantum algorithms for quantum chemistry.

After demonstrating quantum advantages using Gaussian boson sampling, exploring the potential applications of quantum computers in materials design and drug discovery is attracting increasing attention.

However, the quantum resource requirements in materials and (bio)molecular simulations far exceed the capabilities of recent quantum devices. In this work, multiscale quantum computing is proposed: quantum simulations of complex systems by integrating multiple computational methods at different scale resolutions.

Schematic diagram of the architecture of multiscale quantum computing. (1) the protein-ligand complex is divided into two parts, including the molecule and the environment, described by the principles of quantum mechanics and molecular mechanics, respectively; (2) the molecule is further divided into fragments. The electron-electron interactions within and between the fragments are described by wave function theory and Hartree-Fock theory, respectively; (3) the molecular orbital space of each fragment is divided into active space and frozen space, corresponding to static and dynamic correlations, respectively; (4) the eigenvalues and eigenstates of the Hamiltonian in the active space are calculated on a quantum computer; (5) the dynamic correlations are calculated using a posteriori corrections.

In this framework, most of the computational methods can be implemented in an efficient way on a classical computer, leaving the critical part of the computation to a quantum computer.

The scale of simulations for quantum computing depends heavily on the available quantum resources. As a recent solution, the experimental team integrated the adaptive variational quantum solver algorithm (VQE), the second-order Moller-Plesset method, and Hartree-Fock theory within the framework of the Many Body Expansion (MBE) algorithm.

Strategies that can be integrated in multiscale quantum algorithms are briefly summarized

This new algorithm has a fairly high accuracy on classical simulators. The multiscale quantum computing framework proposed in the work can take advantage of quantum algorithms that can provide quantum advantages when fault-tolerant quantum computers become a reality, so that it will outperform classical multiscale simulation methods in the future, the team said. Thus, the multiscale quantum computing framework provides a viable strategy for applying quantum computers to complex materials and biomolecular simulations.

Source:

[1] https://pubs.rsc.org/en/content/articlelanding/2023/sc/d2sc06875c

[2] https://mp.weixin.qq.com/s/6YxvM589ePsaeVr_GQ8YqQ