Quantum annealing for materials science

Finding the global minimum of a potential energy surface is a fundamental challenge in materials science, with applications ranging from protein folding to cluster physics. In recent decades, quantum annealing (QA) has emerged as a promising global optimization algorithm [1,2], leveraging quantum fluctuations in contrast to the thermal fluctuations driving classical simulated annealing. However, the direct implementation of QA can be computationally demanding for multidimensional problems due to the difficulty of sampling the quantum nuclear density.

To tackle this challenge, we propose a novel QA implementation based on path-integral molecular dynamics [3]. While maintaining the flexibility and simplicity of molecular dynamics simulations, this quantum algorithm often outperforms its classical counterpart on benchmark problems such as Lennard-Jones clusters. Furthermore, when combined with machine-learning potentials, it enables the solution of relevant materials science problems, such as reconstructing experimental structures with missing hydrogen sites.

[1] T. Gregor et al., Chem. Phys. Lett. 412, 125 (2005).

[2] L. Stella et al., Phys. Rev. B 72, 014303 (2005).

[3] M. Ceriotti et al.,Comput. Phys. Commun., 185, 1019 (2014).