Traditional optical fiber communication encryption methods lack sufficient dynamic adaptability and hardware flexibility,while reconfigurable logic gates can overcome this limitation,thereby significantly improving th...Traditional optical fiber communication encryption methods lack sufficient dynamic adaptability and hardware flexibility,while reconfigurable logic gates can overcome this limitation,thereby significantly improving the flexibility of encryption systems.This study reports a reconfigurable optoelectronic logic gate(OELG)system based on hafnium-zirconium oxide(HZO)ferroelectric thin films.Through ultra-low temperature atomic layer deposition technique,the fabricated HZO thin films demonstrate an exceptional pyroelectric coefficient(1835.91μC m^(−2) K^(-1))and robust multi-level polarization stability,enabling efficient broadband photon-to-current conversion.By leveraging the pyroelectric effect and tunable polarization states,the OELG device achieves dynamic optical signal modulation and logic processing.The OELG device supports five fundamental logic operations(AND,OR,NAND,NOR,NOT)via electrical bias and polarization control,without requiring hardware modifications.The OELG device demonstrates stable performance over 109 cycles with no degradation,meeting practical application requirements.Furthermore,a convolutional neural network(CNN)-integrated image encryption-decryption framework was validated,achieving 95.01%recognition accuracy on decrypted data,while unauthorized decryption attempts resulted in significant feature loss.This study addresses security challenges in optical communication networks by proposing an innovative solution that integrates pyroelectric materials with reconfigurable logic gate technology,offering a new pathway to enhance physical-layer security.展开更多
Neuromorphic computing provides a remarkably efficient and adaptable alter-native to traditional computing architectures by embodying the impressive power efficiency and parallel processing capabilities of the human b...Neuromorphic computing provides a remarkably efficient and adaptable alter-native to traditional computing architectures by embodying the impressive power efficiency and parallel processing capabilities of the human brain.How-ever,the prevailing focus on integrate-and-fire mode in current artificial neu-rons fails to fully acknowledge the nuanced multifunctionality and adaptive characteristics,especially the temporally variable operating modes and spatial heterogeneity present in natural neurons.Here we report a spatiotemporal-specific artificial neuron implemented with a ferroelectric planar memristor,by engineering the inherent in-plane ferroelectricity ofα-In_(2)Se_(3) and the exten-sive regulation capability of the co-planar multi-electrodes.With enhanced information processing capabilities,the artificial neuron facilitates adjustable reservoir computing and reconfigurable 16 types of logic-gate operations,ulti-mately achieving precise speech recognition with an accuracy approaching 100%.Our work clearly demonstrates the benefits of spatiotemporal specificity in artificial neurons,and contributes to the advancement of more realistic neuromorphic computing systems.展开更多
基金financially supported by the National Key R&D Program of China(2021YFB4000800)the Provincial Natural Science Foundation of Hunan(2025JJ60351,2023JJ30599,2023JJ50009)the National Natural Science Foundation of China(U23A20322)。
文摘Traditional optical fiber communication encryption methods lack sufficient dynamic adaptability and hardware flexibility,while reconfigurable logic gates can overcome this limitation,thereby significantly improving the flexibility of encryption systems.This study reports a reconfigurable optoelectronic logic gate(OELG)system based on hafnium-zirconium oxide(HZO)ferroelectric thin films.Through ultra-low temperature atomic layer deposition technique,the fabricated HZO thin films demonstrate an exceptional pyroelectric coefficient(1835.91μC m^(−2) K^(-1))and robust multi-level polarization stability,enabling efficient broadband photon-to-current conversion.By leveraging the pyroelectric effect and tunable polarization states,the OELG device achieves dynamic optical signal modulation and logic processing.The OELG device supports five fundamental logic operations(AND,OR,NAND,NOR,NOT)via electrical bias and polarization control,without requiring hardware modifications.The OELG device demonstrates stable performance over 109 cycles with no degradation,meeting practical application requirements.Furthermore,a convolutional neural network(CNN)-integrated image encryption-decryption framework was validated,achieving 95.01%recognition accuracy on decrypted data,while unauthorized decryption attempts resulted in significant feature loss.This study addresses security challenges in optical communication networks by proposing an innovative solution that integrates pyroelectric materials with reconfigurable logic gate technology,offering a new pathway to enhance physical-layer security.
基金National Natural Science Foundation of China,Grant/Award Numbers:52021001,52102156,52222206,U20A20244National Key Research and Development Program of China,Grant/Award Number:2021YFA0718800。
文摘Neuromorphic computing provides a remarkably efficient and adaptable alter-native to traditional computing architectures by embodying the impressive power efficiency and parallel processing capabilities of the human brain.How-ever,the prevailing focus on integrate-and-fire mode in current artificial neu-rons fails to fully acknowledge the nuanced multifunctionality and adaptive characteristics,especially the temporally variable operating modes and spatial heterogeneity present in natural neurons.Here we report a spatiotemporal-specific artificial neuron implemented with a ferroelectric planar memristor,by engineering the inherent in-plane ferroelectricity ofα-In_(2)Se_(3) and the exten-sive regulation capability of the co-planar multi-electrodes.With enhanced information processing capabilities,the artificial neuron facilitates adjustable reservoir computing and reconfigurable 16 types of logic-gate operations,ulti-mately achieving precise speech recognition with an accuracy approaching 100%.Our work clearly demonstrates the benefits of spatiotemporal specificity in artificial neurons,and contributes to the advancement of more realistic neuromorphic computing systems.
