講演情報
[17p-S2_204-14]Machine-Learning Force Field Analysis of Polarization Response in Tetragonal Phase BaTiO3 under Uniaxial Compressive Stress
〇Poyen Chen1, Teruyasu Mizoguchi1,2 (1.Univ. of Tokyo, 2.IIS, UTokyo)
キーワード：
machine learning potential、ferroelectric and dielectric property、molecular dynamic simulation
Barium titanate (BaTiO3: BTO) exhibits strong dielectric and piezoelectric properties, enabling its widespread use in capacitors and sensors. Recent studies have shown that mechanical stress can strongly influence the polarization-switching process by modifying the local atomic structure, energy landscape, and domain configuration. These effects lead to polarization rotation, changes in the polarization–electric field (P–E) hysteresis behavior, and the formation of domain-wall structures. In particular, uniaxial stress applied along a single crystallographic direction induces anisotropic lattice deformation and direction-specific electromechanical responses, resulting in distinct dielectric and switching behaviors.
In this study, we investigate the effects of uniaxial compressive stress on tetragonal BTO with emphasis on loading along the c-axis to activate the electromechanical coupling. We employ machine-learning force-field molecular dynamics (MLFF-MD) integrated with a Born effective charge (BEC) prediction model to investigate stress-induced changes in the P–E hysteresis behavior, enabling the application of external electric fields with high fidelity.
Our results reveal clear polarization rotation, the emergence of a double hysteresis loop, and the formation of domain walls under uniaxial compressive stress. This work provides a comprehensive understanding of how uniaxial compressive stress affects polarization behavior and domain-wall evolution in BTO, offering valuable insights for the design of piezoelectric and ferroelectric devices and for the control of stress-engineered domain structures in perovskite oxides.
In this study, we investigate the effects of uniaxial compressive stress on tetragonal BTO with emphasis on loading along the c-axis to activate the electromechanical coupling. We employ machine-learning force-field molecular dynamics (MLFF-MD) integrated with a Born effective charge (BEC) prediction model to investigate stress-induced changes in the P–E hysteresis behavior, enabling the application of external electric fields with high fidelity.
Our results reveal clear polarization rotation, the emergence of a double hysteresis loop, and the formation of domain walls under uniaxial compressive stress. This work provides a comprehensive understanding of how uniaxial compressive stress affects polarization behavior and domain-wall evolution in BTO, offering valuable insights for the design of piezoelectric and ferroelectric devices and for the control of stress-engineered domain structures in perovskite oxides.