Controlling Distinguishable Ferroelectric States with Pulsed Electric Fields

Ferroelectric materials are of increasing interest for novel computing paradigms such as neuromorphic computing, which require a large number of distinguishable states that can be set and read out in a controllable way. Achieving this necessitates a deeper understanding of polarization dynamics and their interplay with material grains and boundaries, down to the nanoscale. Using large ensembles of phase field simulations, we investigate domain switching dynamics under different field conditions for realistic device and grain size distributions. With statistical analyses and machine-learning approaches on these simulation results, we identify relationships between electric field pulse patterns and polarization states as a function of grain distributions, and quantify the capacity for realistic ferroelectric materials to be poled into a multitude of distinguishable states.