【Domestic News】Professor Qian Kai's Team of Shandong University has Made New Achievement in the Research of 2D Gallium Oxide Neuromorphic Memristors Chip

      Recently, Professor Qian Kai's team from the School of Microelectronics made new progress in the research of 2D gallium oxide neuromorphic memristors chip, related results of which was published on ACS Applied Materials & Interfaces, titled “Squeeze-Printing Ultrathin 2D Gallium Oxide out of Liquid Metal for Forming-Free Neuromorphic Memristors”(JCR Zone 1, impact factor: 10.383). Professor Qian Kai is the corresponding author of the paper, Xu Yimeng, a graduate student from the School of Microelectronics, is the co-first author of the paper, and Shandong University is the first unit of the paper.

      In the traditional von Neumann computing system, due to the physical separation of computing and storage (i. e., the von Neumann bottleneck), a large amount of data is frequently transmitted back and forth between the processor and the storage unit, generating a large energy consumption (i. e., power consumption wall). On the other hand, as the speed increases between the processor and the storage unit, this phenomenon will be further intensified (i. e., the storage wall). Neuromorphic computing refers to the human brain for efficient and parallel learning, computing, and reasoning, which is one of the most promising alternatives to von Neumann computing architecture. Similar to the weight modulation of biological synapses, the memristor can achieve "memory and computing integration" information processing by controlling external electrical stimulation, which provides a solution for achieving high-throughput and energy-efficient information processing in neuromorphic computing. Therefore, the precise and effective simulation of synaptic plasticity function using memristors is a crucial step in the realization of brain-like neuromorphic computing.

      Since the discovery of monolayer graphene through mechanical stripping, 2D materials have greatly promoted the development of optoelectronic devices due to their excellent characteristics, among which 2D metal oxides bring a new paradigm for the design of high-performance optoelectronic devices. With excellent chemical stability, high breakdown field strength (~8 MV / cm) and ultra-wide-band gap (~ 4.8 eV), Ga₂O₃ is widely used in photodetectors, field-effect transistors, diodes and other optoelectronic devices. However, there are still many difficulties in how to prepare high-quality ultra-thin, large-size 2D Ga₂O₃ films and apply them in neuromorphic memristors. In this study, continuous and uniform large-size 2D Ga₂O₃ films (~ 3nm) were successfully prepared by liquid metal extrusion printing, and a neuromorphic memristors chip with tunable synaptic behavior was constructed. The device does not require an electrical forming process (forming process), and multi-resistance high-density storage can be achieved by adjusting the limiting current. In addition, with adjustable conductance, 2D Ga₂O₃ memristor successfully simulated multiple functions of biological synaptic plasticity, including dual-pulse facilitation, pulse time-dependent plasticity, long-term memory enhancement and inhibition, which lays the foundation for future implementation of neuromorphic computations, demonstrating the great potential of 2D Ga₂O₃ to further constructing neuromorphic computational systems.

      Professor Qian Kai’s team（https://faculty.sdu.edu.cn/qiankai/zh_CN/index.htm）has focused on flexible "integration of memory and calculation" neuromorphic memristors chip for a long time, mainly including neuromorphic memristors chip, flexible electrodes, biological tissue repair and intelligent sensors for medical and health monitoring.

Original link: https://doi.org/10.1021/acsami.3c02998