Hangyu Xu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Hangyu Xu has a total of 98 co-authors. Here are the top co-authors by number of shared publications:
- Weida Hu
State Key Laboratory of Infrared Physics, Shanghai Institute of Technical Physics, Chinese Academy of Sciences, Shanghai, CHINA
5 shared publications - Fang Wang
State Key Laboratory of Infrared Physics, Shanghai Institute of Technical Physics, Chinese Academy of Sciences, Shanghai, CHINA
2 shared publications - Jianbin Xu
The Chinese University of Hong Kong
2 shared publications - Jinshui Miao
Hangzhou Institute for Advanced Study, University of Chinese Academy of Sciences, Hangzhou 310024, China
2 shared publications - Sang-Hoon Bae
Department of Mechanical Engineering and Materials Science, Washington University in St. Louis, St. Louis, USA
2 shared publications - Tangxin Li
State Key Laboratory of Infrared Physics, Shanghai Institute of Technical Physics, Chinese Academy of Sciences 1 , Shanghai 200083, China
2 shared publications
Latest Publications
- Open Access
- 27 October 2026
- Published by Springer Nature in Research Square
Efficient visual motion prediction using memristive spatiotemporal compression
Dynamic visual prediction is essential for applications ranging from aerial sensing to autonomous systems and surveillance, but conventional approaches retain multi-frame observations and process full spatial representations, incurring substantial storage, data-transfer and computational overhead. Here we report a memristive spatiotemporal compression architecture for efficient visual motion prediction. Temporally, successive visual changes are integrated into an evolving conductance state, consolidating trajectory information without retaining a multi-frame history. Spatially, analogue row and column current projection compresses the two-dimensional conductance states into one-dimensional currents before prediction. A lightweight memristive multi-layer perceptron directly processes these projection currents to infer future target positions. On a 180-nm memristive platform, the system achieves a mean coordinate error of 1.31 array cells, compared with 1.04 cells for the graphics processing unit (GPU) baseline, while reducing latency by 58.6% and energy consumption by 228×. Hardware experiments in dual-object scenes further demonstrate trajectory prediction and proximity-risk assessment with accuracy close to that of the GPU baseline, while reducing latency by 69.4% and energy consumption by 684×. This work advances neuromorphic visual prediction for fast and energy-efficient perception in dynamic environments.
- 27 October 2026
- Published by Springer Nature in Research Square
Open AccessEfficient visual motion prediction using memristive spatiotemporal compression
Dynamic visual prediction is essential for applications ranging from aerial sensing to autonomous systems and surveillance, but conventional approaches retain multi-frame observations and process full spatial representations, incurring substantial storage, data-transfer and computational overhead. Here we report a memristive spatiotemporal compression architecture for efficient visual motion prediction. Temporally, successive visual changes are integrated into an evolving conductance state, consolidating trajectory information without retaining a multi-frame history. Spatially, analogue row and column current projection compresses the two-dimensional conductance states into one-dimensional currents before prediction. A lightweight memristive multi-layer perceptron directly processes these projection currents to infer future target positions. On a 180-nm memristive platform, the system achieves a mean coordinate error of 1.31 array cells, compared with 1.04 cells for the graphics processing unit (GPU) baseline, while reducing latency by 58.6% and energy consumption by 228×. Hardware experiments in dual-object scenes further demonstrate trajectory prediction and proximity-risk assessment with accuracy close to that of the GPU baseline, while reducing latency by 69.4% and energy consumption by 684×. This work advances neuromorphic visual prediction for fast and energy-efficient perception in dynamic environments.
- 18 September 2026
- Published by Wiley in Advanced Materials Technologies
Capillary‐Driven Patterned Transfer of MoS 2 Arrays for Gate‐Tunable Optoelectronic Computing
2D materials have emerged as promising building blocks for next‐generation electronic and optoelectronic devices. The clean transfer of high‐quality, as‐grown 2D materials onto target substrates is a crucial step toward their integration into advanced device architectures. However, the etching‐assisted transfer processes always introduce interfacial contamination and trap states, leading to undesired carrier recombination and degraded optoelectronic performance. Here, we develop a PDMS‐stamp‐assisted large‐scale transfer strategy for the fabrication of high‐quality MoS 2 optoelectronic device arrays. Guided by the work of adhesion theory, a standardized interface engineering approach is established to optimize interfacial adhesion, enabling reliable, high‐yield, and low‐damage transfer. Comprehensive material characterization and optoelectronic measurements confirm the high transfer quality and preserved device performance. A uniform 16 × 16 MoS 2 photodetector with a yield of 98.4% is successfully demonstrated. Finally, we demonstrate a proof‐of‐concept parallel RGB convolutional neural network (CNN) implementation, underscoring its potential in optoelectronic computing.
- 18 September 2026
- Published by Wiley in Advanced Materials Technologies
Capillary‐Driven Patterned Transfer of MoS 2 Arrays for Gate‐Tunable Optoelectronic Computing
2D materials have emerged as promising building blocks for next‐generation electronic and optoelectronic devices. The clean transfer of high‐quality, as‐grown 2D materials onto target substrates is a crucial step toward their integration into advanced device architectures. However, the etching‐assisted transfer processes always introduce interfacial contamination and trap states, leading to undesired carrier recombination and degraded optoelectronic performance. Here, we develop a PDMS‐stamp‐assisted large‐scale transfer strategy for the fabrication of high‐quality MoS 2 optoelectronic device arrays. Guided by the work of adhesion theory, a standardized interface engineering approach is established to optimize interfacial adhesion, enabling reliable, high‐yield, and low‐damage transfer. Comprehensive material characterization and optoelectronic measurements confirm the high transfer quality and preserved device performance. A uniform 16 × 16 MoS 2 photodetector with a yield of 98.4% is successfully demonstrated. Finally, we demonstrate a proof‐of‐concept parallel RGB convolutional neural network (CNN) implementation, underscoring its potential in optoelectronic computing.
- Open Access
- 12 June 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
K+-free mica-assisted epitaxy of Bi-based chalcogenide and oxychalcogenide single-crystals
A material family that intrinsically hosts topological, high-mobility semiconducting and ferroelectric properties offers a platform for exploring new physical phenomena and advancing next-generation electronics and optoelectronics. Bismuth-based chalcogenides and oxychalcogenides are especially promising in this context due to their compositionally tunable functionalities. However, their development is hindered by the lack of a general, scalable synthesis method, as existing approaches are material-specific and rely on costly, lattice-matched substrates. Here, we report a synergistic epitaxy strategy that pairs elemental precursors with an atomically engineered K+-free mica surface, enabling universal growth of nine millimeter-sized bismuth-based chalcogenide and oxychalcogenide single crystals, including topological insulators, high-mobility semiconductors, and ferroelectrics/dielectrics. High-quality materials enable high-performance transistors and broadband photodetectors with high responsivities. Furthermore, we demonstrate their integration into large-scale, uniform arrays and showcase their imaging capabilities across visible to infrared bands. This work establishes a scalable and material-agnostic synthesis framework, unlocking the systematic exploration and device integration of this multifunctional material family.
- 12 June 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Open AccessK+-free mica-assisted epitaxy of Bi-based chalcogenide and oxychalcogenide single-crystals
A material family that intrinsically hosts topological, high-mobility semiconducting and ferroelectric properties offers a platform for exploring new physical phenomena and advancing next-generation electronics and optoelectronics. Bismuth-based chalcogenides and oxychalcogenides are especially promising in this context due to their compositionally tunable functionalities. However, their development is hindered by the lack of a general, scalable synthesis method, as existing approaches are material-specific and rely on costly, lattice-matched substrates. Here, we report a synergistic epitaxy strategy that pairs elemental precursors with an atomically engineered K+-free mica surface, enabling universal growth of nine millimeter-sized bismuth-based chalcogenide and oxychalcogenide single crystals, including topological insulators, high-mobility semiconductors, and ferroelectrics/dielectrics. High-quality materials enable high-performance transistors and broadband photodetectors with high responsivities. Furthermore, we demonstrate their integration into large-scale, uniform arrays and showcase their imaging capabilities across visible to infrared bands. This work establishes a scalable and material-agnostic synthesis framework, unlocking the systematic exploration and device integration of this multifunctional material family.
- 6 April 2026
- Published by Springer Nature in Nature Materials
Two-dimensional materials for integrated sensing
Recent advances in electrical multiply-accumulate (MAC) operations leveraging resistive-switching materials have catalysed significant progress in optoelectronic sensing and computing technologies through the exploration of emerging materials. These innovative approaches facilitate the encoding of optical amplitude information such as retina-like functionalities. However, a critical dimensional mismatch persists between electrical and optical information, resulting in a substantial portion of high-dimensional data channels remaining unexplored in conventional MAC operation schemes. Combined with advanced device architectures and data algorithms, two-dimensional materials are considered promising candidates to realize in situ encoding and optoelectronic sensing of multi-dimensional optical information under precise control owing to their tunable physical properties. In this Review, we outline the progress of emerging two-dimensional-materials-based 'integrated sensors', and benchmark electrical inputs with optical scenarios in a framework unifying information encoding. Exciting opportunities for integrated sensors are discussed as well, highlighting the requirements and differences in the encoding of different dimensions of information and exploring the potential for integrated sensors in other fields.
- 6 April 2026
- Published by Springer Nature in Nature Materials
Two-dimensional materials for integrated sensing
Recent advances in electrical multiply-accumulate (MAC) operations leveraging resistive-switching materials have catalysed significant progress in optoelectronic sensing and computing technologies through the exploration of emerging materials. These innovative approaches facilitate the encoding of optical amplitude information such as retina-like functionalities. However, a critical dimensional mismatch persists between electrical and optical information, resulting in a substantial portion of high-dimensional data channels remaining unexplored in conventional MAC operation schemes. Combined with advanced device architectures and data algorithms, two-dimensional materials are considered promising candidates to realize in situ encoding and optoelectronic sensing of multi-dimensional optical information under precise control owing to their tunable physical properties. In this Review, we outline the progress of emerging two-dimensional-materials-based 'integrated sensors', and benchmark electrical inputs with optical scenarios in a framework unifying information encoding. Exciting opportunities for integrated sensors are discussed as well, highlighting the requirements and differences in the encoding of different dimensions of information and exploring the potential for integrated sensors in other fields.
- Top Cited Papers
- 18 February 2026
- Published by Springer Nature in Nature Nanotechnology
- Vol. 21 (4) , 571-578
Twelve-inch electrically anisotropic boridene for optoelectronic computing
Optoelectronic computing devices capable of bipolar responses offer a route to simplified architectures for processing complex tasks. However, advancing such systems towards large-scale, in-sensor computing has been constrained by the difficulty of monolithically integrating neuromorphic optoelectronic arrays with peripheral circuits, largely due to high material growth temperatures and the non-uniform performance of complex device stacks. Here we report Mo4/3B2Tz (Tz = O, OH, F) boridene as a low-thermal-budget platform for neuromorphic optoelectronics, enabling twelve-inch deposition below 150 °C with excellent wafer-scale uniformity. Ordered metal vacancies and interlayer registry variations generate an unusual electrical anisotropy in which through-plane conduction dominates over in-plane transport. This anisotropy enables a simplified three-terminal device architecture that exhibits intrinsic bipolar and highly linear programmable photoresponses. Correlative conductive atomic force microscopy and first-principles simulations reveal that vacancy-mediated interlayer charge transfer governs the observed behaviour. We further fabricate a 54 × 54-pixel2 optoelectronic computing array with a 99.48% yield and 16 fully separable states. Using a 3k-pixel system prototype, we demonstrate the diagnosis of ophthalmic disorders. Our work establishes Mo4/3B2Tz boridene as a scalable nanomaterial platform that brings neuromorphic optoelectronic computing closer to practical implementations.
- 18 February 2026
- Published by Springer Nature in Nature Nanotechnology
- Vol. 21 (4) , 571-578
Top Cited PapersTwelve-inch electrically anisotropic boridene for optoelectronic computing
Optoelectronic computing devices capable of bipolar responses offer a route to simplified architectures for processing complex tasks. However, advancing such systems towards large-scale, in-sensor computing has been constrained by the difficulty of monolithically integrating neuromorphic optoelectronic arrays with peripheral circuits, largely due to high material growth temperatures and the non-uniform performance of complex device stacks. Here we report Mo4/3B2Tz (Tz = O, OH, F) boridene as a low-thermal-budget platform for neuromorphic optoelectronics, enabling twelve-inch deposition below 150 °C with excellent wafer-scale uniformity. Ordered metal vacancies and interlayer registry variations generate an unusual electrical anisotropy in which through-plane conduction dominates over in-plane transport. This anisotropy enables a simplified three-terminal device architecture that exhibits intrinsic bipolar and highly linear programmable photoresponses. Correlative conductive atomic force microscopy and first-principles simulations reveal that vacancy-mediated interlayer charge transfer governs the observed behaviour. We further fabricate a 54 × 54-pixel2 optoelectronic computing array with a 99.48% yield and 16 fully separable states. Using a 3k-pixel system prototype, we demonstrate the diagnosis of ophthalmic disorders. Our work establishes Mo4/3B2Tz boridene as a scalable nanomaterial platform that brings neuromorphic optoelectronic computing closer to practical implementations.
- Open Access
- 7 February 2025
- Published by Springer Nature in Light: Science & Applications
- Vol. 14 (1) , 1-10
Critical band-to-band-tunnelling based optoelectronic memory
Neuromorphic vision hardware, embedded with multiple functions, has recently emerged as a potent platform for machine vision. To realize memory in sensor functions, reconfigurable and non-volatile manipulation of photocarriers is highly desirable. However, previous technologies bear mechanism challenges, such as the ambiguous optoelectronic memory mechanism and high potential barrier, resulting in a limited response speed and a high operating voltage. Here, for the first time, we propose a critical band-to-band tunnelling (BTBT) based device that combines sensing, integration and memory functions. The nearly infinitesimal barrier facilitates the tunnelling process, resulting in a broadband application range (940 nm). Furthermore, the observation of dual negative differential resistance (NDR) points confirms that the critical BTBT of photocarriers contributes to the sub-microsecond photomemory speed. Since the photomemory speed, with no motion blur, is important for motion detection, the critical BTBT memory is expected to enable moving target tracking and recognition, underscoring its superiority in intelligent perception.
- 7 February 2025
- Published by Springer Nature in Light: Science & Applications
- Vol. 14 (1) , 1-10
Open AccessCritical band-to-band-tunnelling based optoelectronic memory
Neuromorphic vision hardware, embedded with multiple functions, has recently emerged as a potent platform for machine vision. To realize memory in sensor functions, reconfigurable and non-volatile manipulation of photocarriers is highly desirable. However, previous technologies bear mechanism challenges, such as the ambiguous optoelectronic memory mechanism and high potential barrier, resulting in a limited response speed and a high operating voltage. Here, for the first time, we propose a critical band-to-band tunnelling (BTBT) based device that combines sensing, integration and memory functions. The nearly infinitesimal barrier facilitates the tunnelling process, resulting in a broadband application range (940 nm). Furthermore, the observation of dual negative differential resistance (NDR) points confirms that the critical BTBT of photocarriers contributes to the sub-microsecond photomemory speed. Since the photomemory speed, with no motion blur, is important for motion detection, the critical BTBT memory is expected to enable moving target tracking and recognition, underscoring its superiority in intelligent perception.
- 18 March 2024
- Published by AIP Publishing in Applied Physics Letters
- Vol. 124 (12)
In situ edge extraction enabled by reconfigurable van der Waals infrared photodetectors
At present, the widely used artificial intelligence image perception technology is composed of discrete detection and processing components, which is mismatching the demand for increasingly sizeable redundant image information processing. The transmission of non-critical information between components limits the efficiency of image perception systems. Inspired by the human vision system, which can extract the key features and reduce the transmission burden at the low-level detection end, we propose a symmetric structured mixed-dimensional n/p/n [n-molybdenum disulfide (MoS2)/p-germanium (Ge)/n-MoS2] interdigital van der Waals heterojunction infrared photodetector that can be tuned by bias voltage in response polarity and magnitude. The MoS2/Ge/MoS2 infrared photodetector has a bias-symmetric optoelectronic response and covers the detection band from visible to short-wave infrared. Through the joint detection of a pair of MoS2/Ge/MoS2 infrared photodetectors, we demonstrate the image in situ edge extraction at the detection end, which provides key features for high-level processing. This work has discovered the potential of retina-inspired infrared photodetectors on a 2D/3D integration platform, providing distinct opportunities for a neuromorphic visual perception hardware.
- 18 March 2024
- Published by AIP Publishing in Applied Physics Letters
- Vol. 124 (12)
In situ edge extraction enabled by reconfigurable van der Waals infrared photodetectors
At present, the widely used artificial intelligence image perception technology is composed of discrete detection and processing components, which is mismatching the demand for increasingly sizeable redundant image information processing. The transmission of non-critical information between components limits the efficiency of image perception systems. Inspired by the human vision system, which can extract the key features and reduce the transmission burden at the low-level detection end, we propose a symmetric structured mixed-dimensional n/p/n [n-molybdenum disulfide (MoS2)/p-germanium (Ge)/n-MoS2] interdigital van der Waals heterojunction infrared photodetector that can be tuned by bias voltage in response polarity and magnitude. The MoS2/Ge/MoS2 infrared photodetector has a bias-symmetric optoelectronic response and covers the detection band from visible to short-wave infrared. Through the joint detection of a pair of MoS2/Ge/MoS2 infrared photodetectors, we demonstrate the image in situ edge extraction at the detection end, which provides key features for high-level processing. This work has discovered the potential of retina-inspired infrared photodetectors on a 2D/3D integration platform, providing distinct opportunities for a neuromorphic visual perception hardware.
- Top Cited Papers
- 17 September 2021
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 373 (6561) , 1353-1358
2D materials–based homogeneous transistor-memory architecture for neuromorphic hardware
Memory and logic in the same device: Future artificial intelligence applications and data-intensive computations require the development of neuromorphic systems beyond traditional heterogeneous device architectures. Physical separation between a peripheral signal-processing unit and a memory-operating unit is one of the main bottlenecks of heterogeneous architectures, blocking further improvements in efficient resistance matching, energy consumption, and integration compatibility. Tong et al . present a transistor-memory architecture based on a homogeneous tungsten selenide-on-lithium niobate device array (see the Perspective by Rao and Tao). Analog peripheral signal preprocessing and nonvolatile memory were possible within the same device structure, promising diverse neuromorphic functionalities and offering potential improvements in neuromorphic systems on-chip. —YS
- 17 September 2021
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 373 (6561) , 1353-1358
Top Cited Papers2D materials–based homogeneous transistor-memory architecture for neuromorphic hardware
Memory and logic in the same device: Future artificial intelligence applications and data-intensive computations require the development of neuromorphic systems beyond traditional heterogeneous device architectures. Physical separation between a peripheral signal-processing unit and a memory-operating unit is one of the main bottlenecks of heterogeneous architectures, blocking further improvements in efficient resistance matching, energy consumption, and integration compatibility. Tong et al . present a transistor-memory architecture based on a homogeneous tungsten selenide-on-lithium niobate device array (see the Perspective by Rao and Tao). Analog peripheral signal preprocessing and nonvolatile memory were possible within the same device structure, promising diverse neuromorphic functionalities and offering potential improvements in neuromorphic systems on-chip. —YS
- 1 December 2017
- Published by IOP Publishing in Chinese Physics B
- Vol. 26 (12) , 127502
Low temperature magnetic and magnetostrictive properties in $\Pr {({\mathrm{Fe}}_{1-x}{\mathrm{Co}}_{x})}_{1.9}$ cubic Laves alloysThe structures, spin reorientations, magnetic, and magnetostrictive properties of the polycrystalline Pr(Fe1−x Co x )1.9 (x = 0–1.0) cubic laves phase alloys between 5 K and 300 K are investigated. Large low-field magnetostrictions are observed at 5 K in the alloys with x = 0.2 and 0.4 due to the low magnetic anisotropies of these two alloys. A large negative magnetostriction of about −1130 ppm is found in PrCo1.9 alloy at 5 K. The magnetizations of the alloys with decrease abnormally at the spin reorientation temperature , and an abnormity is detected in the alloy with x = 1.0 at its Curie temperature (45 K). The substitution of Fe by Co increases the value of in the alloy with x value increasing from 0.0 to 0.4, and then reduces the value of with x value further increasing to 0.6.
- 1 December 2017
- Published by IOP Publishing in Chinese Physics B
- Vol. 26 (12) , 127502
Low temperature magnetic and magnetostrictive properties in $\Pr {({\mathrm{Fe}}_{1-x}{\mathrm{Co}}_{x})}_{1.9}$ cubic Laves alloysThe structures, spin reorientations, magnetic, and magnetostrictive properties of the polycrystalline Pr(Fe1−x Co x )1.9 (x = 0–1.0) cubic laves phase alloys between 5 K and 300 K are investigated. Large low-field magnetostrictions are observed at 5 K in the alloys with x = 0.2 and 0.4 due to the low magnetic anisotropies of these two alloys. A large negative magnetostriction of about −1130 ppm is found in PrCo1.9 alloy at 5 K. The magnetizations of the alloys with decrease abnormally at the spin reorientation temperature , and an abnormity is detected in the alloy with x = 1.0 at its Curie temperature (45 K). The substitution of Fe by Co increases the value of in the alloy with x value increasing from 0.0 to 0.4, and then reduces the value of with x value further increasing to 0.6.