Investigation of Defect Formation in YMnO₃ under Far-From-Equilibrium Conditions

The investigation of topological defects in far-from-equilibrium conditions remains a significant challenge in modern condensed matter physics, exemplified by the defect formation during rapid phase transition. The Kibble-Zurek (KZ) mechanism explains the symmetry breaking in extreme conditions, and predicts a scaling law between defect density and the quenching rate. However, the mechanism is still not fully understood in multiferroic manganite systems such as YMnO₃. We have investigated the defect density by diffuse neutron scattering experiments on a sequence of rapidly quenched multiferroic manganite YMnO₃ samples, along with the defect characterization using Neural Network Quantum Molecular Dynamics (NNQMD) with Deep Learning techniques. Based on an equivariant machine learning interatomic potential, our simulation efficiently model defect formation during YMnO₃ quenching, with excellent scalability on parallel computing architectures.

In this talk, I will discuss the results of structural characterization and defect dynamics in YMnO₃ from diffuse neutron scattering experiments and large-scale NNQMD simulations.