High-Throughput Screening Search for Antiferroelectricity and Experimental Validations

Ferroelectric materials have been widely used for many applications such as multilayer ceramic capacitors (MLCC), piezoelectric actuator, and random-access memory, and much effort has been made to explore new ferroelectric materials and improve their properties. Recently, antiferroelectric materials have attracted much attention due to their unique and useful responses to an electric field. An antiferroelectric material has a phase transition from an antipolar state to a polar state in an external electric field. This phase transition accompanies a drastic change in dielectric polarization. Consequently, the large dielectric responses (e. g. large permittivity and discharge capacity) are observed around the critical field. However, the known antiferroelectric materials are almost limited to the perovskite materials containing Pb²⁺ or Bi³⁺ ions on A-site, which strongly limits the designer’s perspective.<br/>Unlike ferroelectric materials, which can be identified from a space group symmetry, searching for antiferroelectric materials is not so simple because an antipolar structure could not be distinguished only by looking at the space group. Antipolar structures have no net polarization and are classified into nonpolar space groups. The symmetrical definition of antipolar structure is not apparent and had lacked for decades. In 2016, Toledano <i>et al</i>. proposed a new symmetrical definition of antipolar structure based on a local symmetry on crystallographic sites [1]. When a crystallographic site undergoes a symmetry lowering and acquires a polar site symmetry on the phase transition, the low-symmetry structure can be antipolar.<br/>In this study, we performed high-throughput screening of ICSD database in search for antiferroelectric materials. We adopted the symmetry criteria for antipolar structures in the initial screening, and the formula-based screening of chemical compositions to narrow down the candidates. Then, we picked up some candidate materials and performed a stable structure search via first-principles phonon calculations. In some materials, we successfully find polar and antipolar structures which are energetically comparable, possibly indicating antiferroelectricity. Finally, we experimentally synthesized the obtained candidate materials and fabricated MLCCs to characterize their dielectric and antiferroelectric properties. Though most of them are paraelectric, Bi₂Ti₄O₁₁ and Na₂Nb₄O₁₁ showed clear antiferroelectricity, evidenced by double-hysteresis polarization-field loops and positive DC-field dependences in dielectric permittivity. These experimental results clearly demonstrates that our screening search is effective. Our work provides an effective computational design strategy to explore antiferroelectricity, and moreover, this strategy could be applicable to other functionalities originating from structural phase transitions.<br/><br/>[1] P. Tolédano and M. Guennou, Phys. Rev. B 94, 014107 (2016)