Search papers, labs, and topics across Lattice.
The authors map the multidimensional domain wall landscape across {100}, {110}, and {001} planes in the hybrid improper ferroelectric Ca3Ti2O7 by coupling group theory with machine-learned interatomic potentials (MLIPs). Incorporating long-range electrostatics via a charge-aware training framework with Born effective charges reduces domain wall formation energy prediction errors by 70% relative to density functional theory. Leveraging this potential for atomic-scale minimum energy pathway calculations uncovers a previously obscured low-energy antipolar configuration governing polarization switching.
Standard machine-learned potentials fail to accurately resolve symmetry-broken interfaces until charge-aware electrostatics cut energy errors by 70%, uncovering a hidden low-energy pathway for ferroelectric switching.
Ruddlesden-Popper phases are highly tunable and naturally layered structures, in which polarization can arise via a hybrid improper ferroelectric mechanism. This enables a complex domain wall (DW) structure where multiple order parameters, like octahedral rotations, polar distortions and strain, interact. In this work, we explore the rich set of DW structures in prototypical Ca3Ti2O7, mapping out the DWs in the {100}, {110} and {001} pseudo-tetragonal planes using group theory and machine-learned interatomic potentials (MLIPs). The trained potential reproduces the density functional theory (DFT) order parameter and polarization profiles for all wall types and orientations considered. A charge-aware training framework combined with reference Born effective charges reduces the prediction errors for the DW formation energies by 70%, revealing the importance of including long-range electrostatics to model symmetry-broken interfaces. Finally, the MLIP is used to identify minimum energy pathways at the atomic scale with nearly the precision of DFT calculations, revealing a low-energy antipolar configuration for polarization switching.