Novel 2D Materials-based Resistive Devices

Yanming Liu,He Tian
DOI: https://doi.org/10.1109/3m-nano56083.2022.9941647
2022-01-01
Abstract:Resistive Random Access Memory (RRAM) has been widely investigated for its great synaptic and in-memory computing characteristic. Here, four kinds of novel RRAM or RRAM-based devices' structure have been developed, including graphene inserted RRAM, gate tunable RRAM, SnSe-RRAM, RRAM-based MoS <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> filament transistor. The graphene inserted RRAM improve the device-to-device and cycle-to-cycle stability. Moreover, its optimization makes devices more suitable for neuromorphic computing. The gate tunable RRAM allow devices to have more tunable dimensions, allowing finer tuning of resistive states. The SnSe-RRAM has a double-layer integrated RRAM structure, which leads to stochastic and flexible resistive converting. The RRAM-based MoS <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> filament transistor has quasi-0D contact characteristic, which shows record high On/Off ratio. These results demonstrated that structural optimization of RRAM still has great room for exploration, which leads to higher device integration and more applications.
What problem does this paper attempt to address?