Physical Review B 110, 054109 (2024)

The prototypical antiferroelectric perovskite PbZrO3 (PZO)has garnered considerable attention in recent years due to its significance intechnological applications and fundamental research. Many unresolved issues inPZO are associated with large length- and time-scales, as well as finitetemperatures, presenting significant challenges for first-principles densityfunctional theory studies. Here, we introduce a deep-learnining interatomicpotential of PZO, enabling investigation of finite-temperature propertiesthrough large-scale atomistic simulations. Trained using an elaboratelydesigned dataset, the model successfully reproduces a large number of phases,in particular, the recently discovered 80-atom antiferroelectric Pnam phase andferrielectric Ima2 phase, providing precise predictions for their structuraland dynamical properties. Using this model, we investigated phase transitions ofmultiple phases, including Pbam/Pnam, Ima2, and R3c, which show high similarityto the experimental observation. Our simulation results also highlight thecrucial role of free energy in determining the lowtemperature phase of PZO,reconciling the apparent contradiction: Pbam is the most commonly observedphase in experiments, while theoretical calculations predict other phasesexhibiting even lower energy. Furthermore, in the temperature range where thePbam phase is thermodynamically stable, typical double polarization hysteresis loopsfor antiferroelectrics were obtained, along with a detailed elucidation of thestructural evolution during the electric-field induced transitions between thenonpolar Pbam and polar R3c phases.

‍

‍

‍

‍