Photonic neural network has been sought as an alternative solution to surpass the efficiency and speed bottlenecks of electronic neural network.Despite that the integrated Mach-Zehnder Interferometer(MZI)mesh can perf...Photonic neural network has been sought as an alternative solution to surpass the efficiency and speed bottlenecks of electronic neural network.Despite that the integrated Mach-Zehnder Interferometer(MZI)mesh can perform vector-matrix multiplication in photonic neural network,a programmable in-situ nonlinear activation function has not been proposed to date,suppressing further advancement of photonic neural network.Here,we demonstrate an efficient in-situ nonlinear accelerator comprising a unique solution-processed two-dimensional(2D)MoS2 Opto-Resistive RAM Switch(ORS),which exhibits tunable nonlinear resistance switching that allow us to introduce nonlinearity to the photonic neuron which overcomes the linear voltage-power relationship of typical photonic components.Our reconfigurable scheme enables implementation of a wide variety of nonlinear responses.Furthermore,we confirm its feasibility and capability for MNIST handwritten digit recognition,achieving a high accuracy of 91.6%.Our accelerator constitutes a major step towards the realization of in-situ photonic neural network and pave the way for the integration of photonic integrated circuits(PIC).展开更多
Electronic regulation of two-dimensional(2 D)transition metal dichalcogenides(TMDCs)is a crucial step towards next-generation optoelectronics and electronics.Here,we demonstrate controllable and selective-area defect ...Electronic regulation of two-dimensional(2 D)transition metal dichalcogenides(TMDCs)is a crucial step towards next-generation optoelectronics and electronics.Here,we demonstrate controllable and selective-area defect engineering in 2D molybdenum disulfide(MoS_(2))using a focused ion beam with a low-energy gallium ion(Ga^(+))source.We find that the surface defects of MoS_(2)can be tuned by the precise control of ion energy and dose.Furthermore,the fieldeffect transistors based on the monolayer MoS_(2)show a significant threshold voltage modulation over 70 V after Ga+irradiation.First-principles calculations reveal that the Ga impurities in the monolayer MoS_(2)introduce a defect state near the Fermi level,leading to a shallow acceptor level of 0.25 eV above the valence band maximum.This defect engineering strategy enables direct writing of complex pattern at the atomic length scale in a controlled and facile manner,tailoring the electronic properties of 2D TMDCs for novel devices.展开更多
基金This work is supported by Agency for Science,Technology and Research(A*STAR)Singapore National Research Foundation's Returning Singapore Scientist Scheme(NRF-RSS2015-003)+2 种基金Singapore under its AME Programmatic Funds(A1892b0026)National Research Foundation Grant RSS2015-003the Singapore Hybrid-Integrated Next-Generation u-Electronics(SHINE)Centre hosted at the National University of Singapore(NUS).
文摘Photonic neural network has been sought as an alternative solution to surpass the efficiency and speed bottlenecks of electronic neural network.Despite that the integrated Mach-Zehnder Interferometer(MZI)mesh can perform vector-matrix multiplication in photonic neural network,a programmable in-situ nonlinear activation function has not been proposed to date,suppressing further advancement of photonic neural network.Here,we demonstrate an efficient in-situ nonlinear accelerator comprising a unique solution-processed two-dimensional(2D)MoS2 Opto-Resistive RAM Switch(ORS),which exhibits tunable nonlinear resistance switching that allow us to introduce nonlinearity to the photonic neuron which overcomes the linear voltage-power relationship of typical photonic components.Our reconfigurable scheme enables implementation of a wide variety of nonlinear responses.Furthermore,we confirm its feasibility and capability for MNIST handwritten digit recognition,achieving a high accuracy of 91.6%.Our accelerator constitutes a major step towards the realization of in-situ photonic neural network and pave the way for the integration of photonic integrated circuits(PIC).
基金supported by Fujian Minjiang Distinguished Scholar Programthe Department of Science and Technology of Fujian Province(2020J01704 and 2019L3008)the Scientific Research Foundation from Jimei University(ZP2020066 and ZP2020065)。
文摘Electronic regulation of two-dimensional(2 D)transition metal dichalcogenides(TMDCs)is a crucial step towards next-generation optoelectronics and electronics.Here,we demonstrate controllable and selective-area defect engineering in 2D molybdenum disulfide(MoS_(2))using a focused ion beam with a low-energy gallium ion(Ga^(+))source.We find that the surface defects of MoS_(2)can be tuned by the precise control of ion energy and dose.Furthermore,the fieldeffect transistors based on the monolayer MoS_(2)show a significant threshold voltage modulation over 70 V after Ga+irradiation.First-principles calculations reveal that the Ga impurities in the monolayer MoS_(2)introduce a defect state near the Fermi level,leading to a shallow acceptor level of 0.25 eV above the valence band maximum.This defect engineering strategy enables direct writing of complex pattern at the atomic length scale in a controlled and facile manner,tailoring the electronic properties of 2D TMDCs for novel devices.