Utilizing symmetry operations for high-throughput ferroelectric materials discovery

Abstract

Ferroelectric materials are technically important for applications such as information storage devices and in-memory computing, due to their reversible spontaneous polarization. Computational approaches can help to identify potential ferroelectric materials among a large number of candidate materials, which significantly accelerates the discovery of new ferroelectrics and expands our understanding of potentially available ferroelectric materials. Spontaneous polarization and switching barrier are generally considered as two critical screening criteria. Traditional approaches generally involve identifying the high-symmetry nonpolar structures through pseudo-symmetry analysis as the first step. However, identifying the nonpolar structure is nontrivial and previous approaches have two limitations: the inability to identify structures with large atomic displacement and the potential misidentification of the nonpolar structure. In this work, we propose an approach that directly identifies the polar structure after polarization switching, using the symmetry operations of the framework structure. The spontaneous polarization and switching barrier can be determined from the initial and final polar structures. Utilizing this approach, we conduct a high-throughput screening workflow to discover promising ferroelectrics from structures in the Materials Project database. The workflow leads to the identification of tens of structures that have spontaneous polarization and switching barrier comparable to known ferroelectrics. This simple yet effective approach not only yields promising ferroelectric materials for future experimental validation, but also displays the potential to discover other types of ferroelectric materials, such as antiferroelectrics and two-dimensional ferroelectrics.