Thickness-dependent monochalcogenide GeSe-based CBRAM for memory and artificial electronic synapses

Abstract

Investigating the promising chalcogenide materials for the development of memory and advanced neuromorphic computing applications is a critical step in realizing electronic memory and synaptic devices that can efficiently emulate biological synaptic functions. However, the assessment of monochalcogenide materials for the fabrication of highly scalable memory and electronic synaptic devices that can accurately mimic synaptic functions remain limited. In the present study, we investigated the thickness-dependent resistive switching (RS) behavior of conductive bridge random access memory (CBRAM) based on a monochalcogenide GeSe switching medium for its possible application in high-performance memory and electronic synapses. GeSe thin films of different thicknesses (6, 13, 24, 35, 47, and 56 nm) were deposited via sputtering to fabricate CBRAM devices with a stacking sequence of Ag/GeSe/Pt/Ti/SiO2. The devices exhibited compliance current (CC)-free and electroforming-free RS with highly stable endurance and retention characteristics with no major degradation. All devices with a thickness of 6 nm had a low-resistance state (LRS), which required an initial reset to ensure reliable switching cycles. The devices with a thickness of 47 nm and above exhibited the co-existence of unipolar resistive switching (U-RS) and bipolar resistive switching (B-RS) with the CC-controlled transition between the two switching behaviors. Multilevel resistance states in the 24-nm device between a highresistance state (HRS) and an LRS were achieved by controlling the set-CC (from 5 mA to CC-free) and the reset stop voltage (from −0.5 to −1.0 V) during the set and reset processes, respectively. The analog RS behavior of the device was further investigated with appropriate pulse measurements to emulate vital synaptic functions, including long-term potentiation (LTP), long-term depression (LTD), spike-rate-dependent plasticity (SRDP), spike-timing-dependent plasticity (STDP), paired-pulse facilitation (PPF), paired-pulse depression (PPD) and post-tetanic potentiation (PTP). Overall, the detailed investigation of thickness-dependent GeSe monochalcogenide material indicates that it is a highly suitable candidate for use in highly scalable memory devices and electronic synapses for neuromorphic computing applications.