Emulation of multiple-functional synapses using V2C memristors with coexistence of resistive and threshold switching

• Published in: Materials science in semiconductor processing Vol. 135; p. 106123
• Main Authors: Wang, Yu, Shen, Daqi, Liang, Yilei, Zhao, Yize, Chen, Xintong, Zhou, Lvyang, Zhang, Miaocheng, Xu, Jianguang, Liu, Xiaoyan, Hu, Ertao, Wang, Lei, Xu, Rongqing, Tong, Yi
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 15.11.2021
• Subjects: Artificial synapse; MXene memristor; Threshold/resistive switching; Artificial synapse; MXene memristor; Threshold/resistive switching
• ISSN: 1369-8001; 1873-4081
• DOI: 10.1016/j.mssp.2021.106123

Mxenes, an emerging class of two-dimensional materials, have attracted extensive attention due to their unique electronic properties. One of the MXenes that has been successfully synthesized recently is vanadium carbide MXene. In this work, this new member of MXenes, V2C, was applied as the active layers of memristors involving Ag and W electrodes. The transition from threshold switching to resistive switching was found in the V2C-based memristors which could be achieved via tuning compliance current. The essence of the switching mechanisms was assumed to be the diffusion process of Ag ions. This process can functionally resemble Ca2+ dynamics in synaptic transmission. Together with a positive analog behavior observed, the coexistence of the threshold and resistive switching mechanisms eventually allowed the V2C-based memristors to emulate various synaptic functions. Furthermore, the study of intrinsic frequency dependence of the V2C memristor also indicated that both spike-timing-dependent plasticity (STDP) and spike-rate-dependent plasticity (SRDP) could be achieved on a single V2C memristor without complex pulse engineering or spike overlapping, which may dramatically reduce the complexity of both circuit and algorithm design. The V2C-based memristors thus show great potential to promote the efficiency and scalability of brain-inspired computing systems in the future by simplifying the artificial synapses design. •V2C memristors are fabricated.•V2C memristors realize volatile and non-volatile switching characteristics.•Single V2C memristor resembles the synaptic Ca2+ movement.•Single V2C memristor emulates STDP and SRDP learning rules without extra elements and circuits.