Controllable dynamic magnetism in reconfigurable ferroelastic domain-wall networks engineered via nanocavities

  • Guangming Lu
  • , Shai Rabkin
  • , Wei Liu
  • , Suzhi Li
  • , Guohua Dong
  • , Hyeok Yoon
  • , Harold Y. Hwang
  • , Beena Kalisky
  • , Ekhard K.H. Salje

Research output: Contribution to journalArticlepeer-review

Abstract

Ferroelastic domain walls (DWs) offer a unique platform for engineering dynamic magnetism through controlled symmetry breaking. Here, we demonstrate deterministic magnetic field generation in reconfigurable DW networks using nanocavity-patterned ferroelastic matrices. Atomistic simulations reveal that propagating kinks along polar twin walls induce strain-gradient polarization and displacement current vortices, producing localized magnetic fields (~10-7-10-6 T) via flexoelectric coupling. These fields exhibit scale-free avalanche dynamics with universal power-law exponents, mirroring mechanical energy release events during kink nucleation and depinning transitions. To overcome intrinsic disorder limitations, we mimic lithographically defined nanocavities that guide orthogonal DW propagation with cycle-to-cycle reproducibility, achieving local field enhancement compared to stochastic networks. Scanning superconducting quantum interference device (SQUID) microscopy on SrTiO3 detects out-of-plane magnetic signatures (~ 10-7 T) persisting for milliseconds, quantitatively matching simulations and indicating extended lifetimes. Crucially, kink velocity-dependent dynamic magnetism is established through synchronized magnetic and energy jerk profiles, resolving the atomic-scale mechanism linking DW motion to emergent magnetism. This work establishes a multiscale framework for defect-engineered ferroelastic materials, bridging atomic-scale polarization dynamics, mesoscale DW circuit design, and macroscale nonvolatile memory functionality. By decoupling magnetic responses from intrinsic disorder, our approach advances ferroelastic DW networks toward practical applications in strain-programmable spintronics and ultrahigh-density racetrack memories.

Original languageEnglish
Article number245403
JournalPhysical Review B
Volume112
Issue number24
DOIs
StatePublished - Dec 2025

Bibliographical note

Publisher Copyright:
© 2025 American Physical Society

Fingerprint

Dive into the research topics of 'Controllable dynamic magnetism in reconfigurable ferroelastic domain-wall networks engineered via nanocavities'. Together they form a unique fingerprint.

Cite this