Laboratory of local diagnostics of semiconductor materials

Photoelectronic memristive devices demonstrating dynamic behavior of photoelectronic states, similar to the passage of signals in biological neural networks, can solve one of the main problems of modern opto-microelectronics associated with a critical increase in energy consumption in the detection and processing of big data, which significantly limits the autonomy of visual recognition and decision-making in real time, for example in unmanned vehicles. A modern supercomputer with a digital signal processing architecture that simulates the operation of a neural network with two orders of magnitude fewer neurons than in the human brain (driver) consumes a million times more energy than a biological neural network that performs the same tasks (driving). At the same time, the autonomous operation of the human trained visual neural network recognizes information orders of magnitude faster than a computer. The relevance of studying the formation and control of photoelectronic states in intelligent devices based on van der Waals hetero-integrated low-dimensional structures under conditions of electrical, optical and phonon excitation is determined by the need to create fundamentally new energy-efficient photoelectronic devices for autonomous optical sensing, when data detection, processing and storage occurs in parallel without physical separation of sensor, memory and processor similar to processes in biological neural networks. Intelligent photoelectronic devices can increase the speed of access to extensive neural networks and improve the energy efficiency of data processing and storage in them. The presented project is aimed at solving the scientific problem of the formation and control of multilevel photoelectronic states in devices based on van der Waals heterointegrated mixed-dimensional structures under conditions of electrical, optical and acoustic-electrical excitation. The project is aimed at experimental research and computer modeling of charge transfer processes in heterostructures with two-dimensional crystals and zero-dimensional quantum dots, as well as structural and electronic transitions leading to changes in their photoelectronic states under different conditions of electrical, optical and phonon excitation. Layered photoelectronic memristive van der Waals heterostructures with a photocell made of graphene/graphene oxide, BN, chalcogenides MoS2, SnS2, etc. will be studied for controlling photoelectronic states in the infrared (IR) range under conditions of controlling recombination of charge carriers and acoustic phonon excitations to solve urgent problems of autonomous pattern recognition, machine vision, and artificial intelligence.. The establishment of mechanisms for controlling photoelectronic states in van der Waals hetero-integrated devices will make it possible to create a new generation of information devices with ultra-low power consumption, ultra-high recording density and with the possibility of autonomous photo-/photo-acousto-electronic detection and recognition of signals, similar to in a biological sensory neural network.