Lihong Zhu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Lihong Zhu has a total of 182 co-authors. Here are the top co-authors by number of shared publications:
- Yijun Lu
National Innovation Platform for the Fusion of Industry and Education in Integrated Circuits, Department of Electronic Science, School of Electronic Science and Engineering, Xiamen University, Xiamen 361005, China
42 shared publications - Zhong Chen
Xiamen University
41 shared publications - Weijie Guo
National Innovation Platform for the Fusion of Industry and Education in Integrated Circuits, Department of Electronic Science, School of Electronic Science and Engineering, Xiamen University, Xiamen 361005, China
23 shared publications - Ziquan Guo
Department of Electronic Science, Fujian Engineering Research Center for Solid-State Lighting, Xiamen University, Xiamen, China
15 shared publications - Tingzhu Wu
Xiamen University
14 shared publications - Guolong Chen
National Innovation Platform for the Fusion of Industry and Education in Integrated Circuits, Department of Electronic Science, School of Electronic Science and Engineering, Xiamen University, Xiamen 361005, China
10 shared publications
Latest Publications
- Open Access
- 30 September 2026
- Published by Elsevier in SSRN Electronic Journal
Machine-Learning-Driven Optimization of AlGaN based UVC LEDs for Maximum Light Output Power Density
AlGaN-based ultraviolet-C (UV-C) light-emitting diodes (LEDs) have attracted significant attention due to their direct, wide, and tunable bandgap, making them highly promising for applications such as sterilization and disinfection. However, their practical deployment is severely limited by extremely low external quantum efficiency (EQE), which primarily results from inefficient transverse magnetic (TM) polarized light extraction and rapid performance degradation during operation. To address these challenges, structural optimization, particularly of the AlGaN/AlGaN superlattice (SL) electron blocking layer (EBL), has been widely explored to improve carrier transport and device efficiency. Traditional trial-and-error experimental approaches are often time-consuming and insufficient for identifying optimal device configurations. Here, we propose a data-driven framework using four machine learning models trained on a comprehensive dataset of AlGaN-based LED structures reported over the past decades to predict and optimize light output power density (LOPD). A convolutional neural network (CNN) achieves the best performance (R2 = 0.9806, RMSE = 0.133 W/cm2) and identifies optimized structures with a potential LOPD enhancement of up to ~77%. Exploratory analysis further reveals a coupling between the EBL aluminum composition (XEBL) and p-type magnesium doping concentration (PMg-dop). LOPD is maximized when XEBL is approximately 0.67-0.77 and PMg-dop is approximately 2.1-3.9 × 1019 cm-3 (peaking near XEBL = 0.76, PMg-dop = 3.8 × 1019 cm-3), consistent with a trade-off between electron confinement and hole injection. These predictions agree with TCAD simulations and independent experimental data, demonstrating that interpretable machine learning offers an efficient, scalable strategy for accelerating high-performance UVC LED design.
- 30 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessMachine-Learning-Driven Optimization of AlGaN based UVC LEDs for Maximum Light Output Power Density
AlGaN-based ultraviolet-C (UV-C) light-emitting diodes (LEDs) have attracted significant attention due to their direct, wide, and tunable bandgap, making them highly promising for applications such as sterilization and disinfection. However, their practical deployment is severely limited by extremely low external quantum efficiency (EQE), which primarily results from inefficient transverse magnetic (TM) polarized light extraction and rapid performance degradation during operation. To address these challenges, structural optimization, particularly of the AlGaN/AlGaN superlattice (SL) electron blocking layer (EBL), has been widely explored to improve carrier transport and device efficiency. Traditional trial-and-error experimental approaches are often time-consuming and insufficient for identifying optimal device configurations. Here, we propose a data-driven framework using four machine learning models trained on a comprehensive dataset of AlGaN-based LED structures reported over the past decades to predict and optimize light output power density (LOPD). A convolutional neural network (CNN) achieves the best performance (R2 = 0.9806, RMSE = 0.133 W/cm2) and identifies optimized structures with a potential LOPD enhancement of up to ~77%. Exploratory analysis further reveals a coupling between the EBL aluminum composition (XEBL) and p-type magnesium doping concentration (PMg-dop). LOPD is maximized when XEBL is approximately 0.67-0.77 and PMg-dop is approximately 2.1-3.9 × 1019 cm-3 (peaking near XEBL = 0.76, PMg-dop = 3.8 × 1019 cm-3), consistent with a trade-off between electron confinement and hole injection. These predictions agree with TCAD simulations and independent experimental data, demonstrating that interpretable machine learning offers an efficient, scalable strategy for accelerating high-performance UVC LED design.
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 73 (9) , 5591-5597
Surface-Relief-Treated 940-nm VCSELs for High Beam Quality in Proximity Sensing
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 73 (9) , 5591-5597
Surface-Relief-Treated 940-nm VCSELs for High Beam Quality in Proximity Sensing
- 29 May 2026
- Published by American Chemical Society (ACS) in ACS Photonics
- Vol. 13 (12) , 3387-3394
Spatial-Spectral-Temporal Dynamic Investigation during Carrier Injection in Mini-Light-Emitting Diodes
- 29 May 2026
- Published by American Chemical Society (ACS) in ACS Photonics
- Vol. 13 (12) , 3387-3394
Spatial-Spectral-Temporal Dynamic Investigation during Carrier Injection in Mini-Light-Emitting Diodes
- 11 May 2026
- Published by American Chemical Society (ACS) in ACS Photonics
- Vol. 13 (10) , 2807-2817
Efficient Hyperspectrum and Deep Learning-Based Distributed Optical Fiber Sensing for Mini/Micro-LEDs 3D Spatially Resolved Light Measurement
- 11 May 2026
- Published by American Chemical Society (ACS) in ACS Photonics
- Vol. 13 (10) , 2807-2817
Efficient Hyperspectrum and Deep Learning-Based Distributed Optical Fiber Sensing for Mini/Micro-LEDs 3D Spatially Resolved Light Measurement
- 4 May 2026
- Published by AIP Publishing in Applied Physics Letters
- Vol. 128 (18)
Quantitative evaluation of sidewall impact in micro-LEDs by ultrafast-gated hyperspectral imaging microscopy
Sidewall damage is widely recognized as a critical factor limiting the efficiency and reliability of micro-light-emitting diodes (micro-LEDs). However, its spatial impact region and transient manifestation mechanisms remain insufficiently understood. This work employs ultrafast-gated hyperspectral imaging microscopy in conjunction with scanning electron microscopy (SEM) to perform spatial-spectral-temporal-resolved electroluminescence characterization of GaN-based micro-LEDs, achieving precise localization of the sidewall impact region (SIR) with a pixel-sampling resolution of 72 nm. Based on joint analysis of time-resolved spectral parameters, fluorescence lifetime distributions, and SEM morphology, two quantitative metrics, namely, the SIR and the sidewall impact index (SII), are proposed to characterize the spatial extent and the overall strength impacted by the sidewall effect. The results indicate that the influence of sidewall effects extends well beyond the geometric mesa boundary and becomes more pronounced in smaller devices. By combining the SIR and SII metrics, a quantitative framework is established to precisely evaluate sidewall etching-induced perturbations in carrier recombination, providing experimental guidance for sidewall passivation optimization and micro-LED device design.
- 4 May 2026
- Published by AIP Publishing in Applied Physics Letters
- Vol. 128 (18)
Quantitative evaluation of sidewall impact in micro-LEDs by ultrafast-gated hyperspectral imaging microscopy
Sidewall damage is widely recognized as a critical factor limiting the efficiency and reliability of micro-light-emitting diodes (micro-LEDs). However, its spatial impact region and transient manifestation mechanisms remain insufficiently understood. This work employs ultrafast-gated hyperspectral imaging microscopy in conjunction with scanning electron microscopy (SEM) to perform spatial-spectral-temporal-resolved electroluminescence characterization of GaN-based micro-LEDs, achieving precise localization of the sidewall impact region (SIR) with a pixel-sampling resolution of 72 nm. Based on joint analysis of time-resolved spectral parameters, fluorescence lifetime distributions, and SEM morphology, two quantitative metrics, namely, the SIR and the sidewall impact index (SII), are proposed to characterize the spatial extent and the overall strength impacted by the sidewall effect. The results indicate that the influence of sidewall effects extends well beyond the geometric mesa boundary and becomes more pronounced in smaller devices. By combining the SIR and SII metrics, a quantitative framework is established to precisely evaluate sidewall etching-induced perturbations in carrier recombination, providing experimental guidance for sidewall passivation optimization and micro-LED device design.
- Open Access
- 10 April 2026
- Published by Elsevier in SSRN Electronic Journal
Unraveling Si-doped DX center in AlGaN-Based UVC LEDs Through 2D Cathodoluminescence Mapping
AlGaN-based UVC LEDs are compact, mercury-free, and cover the full 200–280 nm range, making them ideal for medical, military, and industrial uses. However, their internal quantum efficiency remains low due to defects prevalent in Al-rich group III nitrides. In this study, we introduce a novel approach using 2D Cathodoluminescence (CL) to identify defect types, Si-doped DX centers as well as their distributions in UVC LEDs. Focused ion beam scanning electron microscopy (FIB/SEM) was employed to prepare cross-sectional samples, facilitating electron excitation from the side view to optimize CL signal collection. Energy-dispersive X-ray spectroscopy (EDX) was also utilized for direct visual analysis and detailed compositional examination of the layers. To further investigate structural and optical properties, we applied spherical aberration-corrected transmission electron microscopy (AC-TEM) and atomic-resolution electron energy-loss spectroscopy (EELS). This comprehensive analysis provided direct, visual evidence of Si-doped DX center formation, confirming the spatial correlation between Si dopants and nitrogen vacancies (VN) within the N-AlGaN lattice at atomic resolution. Our findings provide essential insights into the detrimental impact of Si-doped defects on UVC LED performance. This understanding serves as a basis for developing targeted defect engineering and material optimization strategies aimed at enhancing the reliability and efficiency of UVC LEDs.
- 10 April 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessUnraveling Si-doped DX center in AlGaN-Based UVC LEDs Through 2D Cathodoluminescence Mapping
AlGaN-based UVC LEDs are compact, mercury-free, and cover the full 200–280 nm range, making them ideal for medical, military, and industrial uses. However, their internal quantum efficiency remains low due to defects prevalent in Al-rich group III nitrides. In this study, we introduce a novel approach using 2D Cathodoluminescence (CL) to identify defect types, Si-doped DX centers as well as their distributions in UVC LEDs. Focused ion beam scanning electron microscopy (FIB/SEM) was employed to prepare cross-sectional samples, facilitating electron excitation from the side view to optimize CL signal collection. Energy-dispersive X-ray spectroscopy (EDX) was also utilized for direct visual analysis and detailed compositional examination of the layers. To further investigate structural and optical properties, we applied spherical aberration-corrected transmission electron microscopy (AC-TEM) and atomic-resolution electron energy-loss spectroscopy (EELS). This comprehensive analysis provided direct, visual evidence of Si-doped DX center formation, confirming the spatial correlation between Si dopants and nitrogen vacancies (VN) within the N-AlGaN lattice at atomic resolution. Our findings provide essential insights into the detrimental impact of Si-doped defects on UVC LED performance. This understanding serves as a basis for developing targeted defect engineering and material optimization strategies aimed at enhancing the reliability and efficiency of UVC LEDs.
- 19 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Electronics
- Vol. 73 (8) , 12516-12526
Nondestructive Batch Testing of Mini-LEDs Using Flexible Electrode Arrays
Miniaturized light-emitting diodes (LEDs) are commonly used in displays and lighting for their high brightness, low power consumption and long lifespan. However, their small scale and high-density integration create challenges for efficient, nondestructive inspection, particularly for simultaneous electrical and optical characterization. This article presents a batch detection method utilizing a flexible electrode array that enables stable and simultaneous contact with multiple mini-LED chips. The flexible electrodes ensure reliable electrical connectivity and precise alignment with low contact resistance and nondestructive probing capabilities. Experimental result demonstrates turn-on voltage deviations within 44.83 mV, mean relative errors below 3.28%, peak and dominant wavelength shifts of less than 0.3%, color coordinate variations of approximately 2%, correlated color temperature fluctuations around 0.8%, as well as average radiant power and luminous flux deviations under 2.8%. The proposed method enables accurate, repeatable, and high-throughput characterization, providing a robust solution for the inspection of mini-LED arrays.
- 19 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Electronics
- Vol. 73 (8) , 12516-12526
Nondestructive Batch Testing of Mini-LEDs Using Flexible Electrode Arrays
Miniaturized light-emitting diodes (LEDs) are commonly used in displays and lighting for their high brightness, low power consumption and long lifespan. However, their small scale and high-density integration create challenges for efficient, nondestructive inspection, particularly for simultaneous electrical and optical characterization. This article presents a batch detection method utilizing a flexible electrode array that enables stable and simultaneous contact with multiple mini-LED chips. The flexible electrodes ensure reliable electrical connectivity and precise alignment with low contact resistance and nondestructive probing capabilities. Experimental result demonstrates turn-on voltage deviations within 44.83 mV, mean relative errors below 3.28%, peak and dominant wavelength shifts of less than 0.3%, color coordinate variations of approximately 2%, correlated color temperature fluctuations around 0.8%, as well as average radiant power and luminous flux deviations under 2.8%. The proposed method enables accurate, repeatable, and high-throughput characterization, providing a robust solution for the inspection of mini-LED arrays.
- Open Access
- 3 March 2026
- Published by Elsevier in SSRN Electronic Journal
Unraveling Si-doped DX center in AlGaN-Based UVC LEDs Through 2D Cathodoluminescence Mapping
AlGaN-based UVC LEDs are compact, mercury-free, and cover the full 200–280 nm range, making them ideal for medical, military, and industrial uses. However, their internal quantum efficiency remains low due to defects prevalent in Al-rich group III nitrides. In this study, we introduce a novel approach using 2D Cathodoluminescence (CL) to identify defect types, Si-doped DX centers as well as their distributions in UVC LEDs. Focused ion beam scanning electron microscopy (FIB/SEM) was employed to prepare cross-sectional samples, facilitating electron excitation from the side view to optimize CL signal collection. Energy-dispersive X-ray spectroscopy (EDX) was also utilized for direct visual analysis and detailed compositional examination of the layers. To further investigate structural and optical properties, we applied spherical aberration-corrected transmission electron microscopy (AC-TEM) and atomic-resolution electron energy-loss spectroscopy (EELS). This comprehensive analysis provided direct, visual evidence of Si-doped DX center formation, confirming the spatial correlation between Si dopants and nitrogen vacancies (VN) within the N-AlGaN lattice at atomic resolution. Our findings provide essential insights into the detrimental impact of Si-doped defects on UVC LED performance. This understanding serves as a basis for developing targeted defect engineering and material optimization strategies aimed at enhancing the reliability and efficiency of UVC LEDs.
- 3 March 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessUnraveling Si-doped DX center in AlGaN-Based UVC LEDs Through 2D Cathodoluminescence Mapping
AlGaN-based UVC LEDs are compact, mercury-free, and cover the full 200–280 nm range, making them ideal for medical, military, and industrial uses. However, their internal quantum efficiency remains low due to defects prevalent in Al-rich group III nitrides. In this study, we introduce a novel approach using 2D Cathodoluminescence (CL) to identify defect types, Si-doped DX centers as well as their distributions in UVC LEDs. Focused ion beam scanning electron microscopy (FIB/SEM) was employed to prepare cross-sectional samples, facilitating electron excitation from the side view to optimize CL signal collection. Energy-dispersive X-ray spectroscopy (EDX) was also utilized for direct visual analysis and detailed compositional examination of the layers. To further investigate structural and optical properties, we applied spherical aberration-corrected transmission electron microscopy (AC-TEM) and atomic-resolution electron energy-loss spectroscopy (EELS). This comprehensive analysis provided direct, visual evidence of Si-doped DX center formation, confirming the spatial correlation between Si dopants and nitrogen vacancies (VN) within the N-AlGaN lattice at atomic resolution. Our findings provide essential insights into the detrimental impact of Si-doped defects on UVC LED performance. This understanding serves as a basis for developing targeted defect engineering and material optimization strategies aimed at enhancing the reliability and efficiency of UVC LEDs.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 75 (00189456) , 1-8
Ultrafast Gated Hyperspectral Imaging System for Spatial-Spectral-Temporal-Resolved Electroluminescence of Miniaturized Light-Emitting Diodes
Miniaturized light-emitting diodes (Mini-/Micro-LEDs) for next-generation displays exhibit complex optoelectronic behaviors due to the shrunk chip size. Comprehensive multi-dimensional characterization is critical yet remains challenging for conventional techniques. In this work, we proposed an ultrafast gated hyperspectral imaging (UGHSI) system that synchronously captures spatial, spectral, and temporal dynamics in miniaturized LEDs, addressing the growing demand for high-resolution, multi-dimensional analysis of their complex transient behaviors. High-resolution two-dimensional (2-D) time-resolved electroluminescence (TREL) measurements are applied to demonstrate the system’s sub-micron spatial resolution and 500 picosecond temporal resolution. To enhance measurement accuracy, an optimized fitting model is developed to compensate the impact of the falling time of the electrical pulse signal on fluorescence lifetime. The 2-D fluorescent lifetime mappings for green Mini/Micro-LEDs derived from the 2-D TREL reveal distinct carrier distribution patterns and luminescence behaviors. In particular, the lifetime mappings clarify the enhanced anomalous emission at the sidewalls of the Mini/Micro-LEDs. In general, the proposed system achieves multi-dimensional optoelectronic characteristics with improved resolution and flexibility, providing a cutting-edge tool for studying complex behaviors of the next-generation devices and materials at microscopic and picosecond scales.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 75 (00189456) , 1-8
Ultrafast Gated Hyperspectral Imaging System for Spatial-Spectral-Temporal-Resolved Electroluminescence of Miniaturized Light-Emitting Diodes
Miniaturized light-emitting diodes (Mini-/Micro-LEDs) for next-generation displays exhibit complex optoelectronic behaviors due to the shrunk chip size. Comprehensive multi-dimensional characterization is critical yet remains challenging for conventional techniques. In this work, we proposed an ultrafast gated hyperspectral imaging (UGHSI) system that synchronously captures spatial, spectral, and temporal dynamics in miniaturized LEDs, addressing the growing demand for high-resolution, multi-dimensional analysis of their complex transient behaviors. High-resolution two-dimensional (2-D) time-resolved electroluminescence (TREL) measurements are applied to demonstrate the system’s sub-micron spatial resolution and 500 picosecond temporal resolution. To enhance measurement accuracy, an optimized fitting model is developed to compensate the impact of the falling time of the electrical pulse signal on fluorescence lifetime. The 2-D fluorescent lifetime mappings for green Mini/Micro-LEDs derived from the 2-D TREL reveal distinct carrier distribution patterns and luminescence behaviors. In particular, the lifetime mappings clarify the enhanced anomalous emission at the sidewalls of the Mini/Micro-LEDs. In general, the proposed system achieves multi-dimensional optoelectronic characteristics with improved resolution and flexibility, providing a cutting-edge tool for studying complex behaviors of the next-generation devices and materials at microscopic and picosecond scales.
- 11 July 2025
- Published by American Chemical Society (ACS) in ACS Photonics
- Vol. 12 (8) , 4234-4242
Investigating the Recombination Mechanism of V-Pits in InGaN/GaN Multiple Quantum Wells LEDs Combining Hyperspectral Electroluminescence and Cathodoluminescence
- 11 July 2025
- Published by American Chemical Society (ACS) in ACS Photonics
- Vol. 12 (8) , 4234-4242
Investigating the Recombination Mechanism of V-Pits in InGaN/GaN Multiple Quantum Wells LEDs Combining Hyperspectral Electroluminescence and Cathodoluminescence