Answers are grounded in this paper and cite their sources.
Reports low-leakage threshold switching in a novel device, highlighting its potential for neuromorphic computing.
Volatile Threshold Switching (TS) devices are critical selector components for suppressing sneak path currents in high-density crossbar arrays. However, active-metal-based TS devices often face the challenge of stochastic conductive filament overgrowth. In this work, we report a high-performance TS device based on a Graphene/CuInP₂S₆/h-BN/Au van der Waals heterostructure. By utilizing hexagonal Boron Nitride (h-BN) as an atomic-scale ion sieve, we achieve atomic-scale geometric confinement of Cu filament growth. The device exhibits excellent unipolar switching characteristics, featuring a high on/off ratio (>10 5 ), ultra-low off-state leakage current (~pA), and a steep subthreshold swing (<15 mV/dec). Mechanistic investigations reveal that the h-BN layer confines the filaments to the atomic scale, causing the metastable filaments to spontaneously rupture upon bias removal due to Rayleigh instability, thereby enabling rapid self-resetting. This interface engineering strategy effectively addresses the filament overgrowth issue, providing a reliable solution for low-power neuromorphic computing hardware.
No takes yet. Share an insight, caveat, or question.
Wen et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: