Optoelectronic dynamic memristor systems based on two-dimensional crystals

Abstract

Abstract Optical modulation of resistive switching in an optoelectronic memristor allows it to be optically controlled at ultra-high speed and ultra-low power consumption. Optical memristive systems with memory elements switched by electromagnetic radiation can be used in optically reconfigurable and tunable neural networks for neuromorphic computing and brain-inspired artificial intelligence systems. Two-dimensional (2D) crystals with unique electrical and optical properties demonstrate tremendous potential in the creation of highly efficient information and sensor systems for real-time data monitoring and processing. In this paper, we consider optoelectronic structures based on graphene, graphene oxide, and molybdenum disulfide for dynamic memristive signal processing. 2D optoelectronic memristor structures exhibit multiple states, which can be monitored in a wide range of optical excitations and used as sensors for pattern recognition and image processing. An optoelectronic dynamic memristive structure with quantum dots (QDs), controlled by ultrafast photoinduced structural phase transitions, is promising for creating information systems with stochastic data processing. The phase transition, controlled by charge and temperature, results in tunable periodic oscillations due to two state variables that exhibit chaotic behavior similar to that which occurs in a biological neuron network.