Polymorphic functionalization driven by ion displacement-induced antiferroelectric ordering in CuBiP₂Se₆.

Antiferroelectric two-dimensional materials, with their unique physical mechanisms, exhibit tunable polarization dynamics and layered structural characteristics, enabling the synergistic implementation of synaptic plasticity, sensory-mimetic functionality, and in-memory computing within a unified device architecture. These capabilities meet the growing polymorphic requirements of neuromorphic systems and position such materials as strong candidates for next-generation neuromorphic computing platforms. Among them, CuBiP₂Se₆ stands out among 2D antiferroelectric materials due to its intrinsic antiferroelectric properties, featuring a stable interlayer antiparallel Cu⁺ dipole configuration. This structure, combined with its relaxor-like behavior, enables a reversible transition between antiferroelectric and ferroelectric states under an applied electric field, along with gradual polarization tuning. This transition mechanism enables continuously tunable conductance states, providing essential physical support for the gradual modulation of synaptic weights and the hardware implementation of complex neural functions, making it particularly suited for high-precision emulation of multilevel synaptic plasticity in neuromorphic applications. In this work, memristor based on two-dimensional antiferroelectric CuBiP₂Se₆ exhibit stable multilevel conductance states, high endurance, and excellent device uniformity, thus supporting diverse neurosynaptic functions and advanced learning rules. These attributes highlight the immense potential of antiferroelectric 2D materials as a foundation for compact, energy-efficient, and highly integrated neuromorphic hardware.