Qiwen Kong
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Qiwen Kong has a total of 55 co-authors. Here are the top co-authors by number of shared publications:
- Xiao Gong
National University of Singapore
21 shared publications - Xiaolin Wang
Department of Electrical and Computer Engineering, National University of Singapore, Queenstown, Singapore
17 shared publications - Chen Sun
Department of Electrical and Computer Engineering, National University of Singapore, Queenstown, Singapore
16 shared publications - Leming Jiao
Department of Electrical and Computer Engineering, National University of Singapore, Queenstown, Singapore
15 shared publications - Gan Liu
National University of Singapore (NUS)
14 shared publications - Dong Zhang
Department of Electrical and Computer Engineering, National University of Singapore, Queenstown, Singapore
13 shared publications
Latest Publications
- 1 August 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 73  (8) , 5099-5105
Fluorine Plasma-Treated ITO Transistors: Reliability and Temperature Dependence From 77 to 375 K
- 1 August 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 73  (8) , 5099-5105
Fluorine Plasma-Treated ITO Transistors: Reliability and Temperature Dependence From 77 to 375 K
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 73  (7) , 4117-4124
BEOL-Compatible NV-SRAM With Stacked ITO FETs and HfO$_{2}$-Based Ferroelectric Capacitors
- 1 July 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 73  (7) , 4117-4124
BEOL-Compatible NV-SRAM With Stacked ITO FETs and HfO$_{2}$-Based Ferroelectric Capacitors
- 24 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 73  (4) , 1882-1888
Area-Efficient BEOL-Compatible MFMIS FeFETs Enabled by a 3-D Multi-Fin Floating Gate Structure
In this work, we propose a 3-D multi-Fin floating gate (FG) structure to improve the area-efficiency of ferroelectric field-effect transistors (FeFETs) based on the metal–ferroelectric–metal–insulator–semiconductor (MFMIS) gate-stack. This architectural innovation exploits the additional channel surface provided by the vertical Fin sidewalls to minimize footprint area consumption while boosting the memory window (MW) of MFMIS FeFETs. Extensive experiments and analyses are conducted to demonstrate the feasibility and effectiveness of the proposed design. Atungsten (W) multi-Fin FG is successfully fabricated, featuring a Fin height of 93 nm and a width of 29 nm. Electrical characterization confirms that the 3-D FG structure enlarges the effective channel area without introducing significant tradeoffs in the key device characteristics of the baseline MIS FETs employing ALD-deposited ZnO channels. MFMIS FeFETs incorporating the multi-Fin FG structure achieve over 50% channel area saving. For devices with a footprint area ratio (AR) of 1:4 between the ferroelectric (FE) layer and the channel region, the proposed 3-D structure significantly enhances the MW from 1.3 to 2.9 V, surpassing that of planar devices with a footprint AR of 1:8. Furthermore, our 3-D multi-Fin FG FeFETs exhibit reliability on par with their planar counterparts, with endurance exceeding $2\times 10^{{9}}$ cycles and retention extrapolated beyond ten years. Our work presents a promising approach to addressing the area-inefficiency concern of MFMIS FeFETs.
- 24 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 73  (4) , 1882-1888
Area-Efficient BEOL-Compatible MFMIS FeFETs Enabled by a 3-D Multi-Fin Floating Gate Structure
In this work, we propose a 3-D multi-Fin floating gate (FG) structure to improve the area-efficiency of ferroelectric field-effect transistors (FeFETs) based on the metal–ferroelectric–metal–insulator–semiconductor (MFMIS) gate-stack. This architectural innovation exploits the additional channel surface provided by the vertical Fin sidewalls to minimize footprint area consumption while boosting the memory window (MW) of MFMIS FeFETs. Extensive experiments and analyses are conducted to demonstrate the feasibility and effectiveness of the proposed design. Atungsten (W) multi-Fin FG is successfully fabricated, featuring a Fin height of 93 nm and a width of 29 nm. Electrical characterization confirms that the 3-D FG structure enlarges the effective channel area without introducing significant tradeoffs in the key device characteristics of the baseline MIS FETs employing ALD-deposited ZnO channels. MFMIS FeFETs incorporating the multi-Fin FG structure achieve over 50% channel area saving. For devices with a footprint area ratio (AR) of 1:4 between the ferroelectric (FE) layer and the channel region, the proposed 3-D structure significantly enhances the MW from 1.3 to 2.9 V, surpassing that of planar devices with a footprint AR of 1:8. Furthermore, our 3-D multi-Fin FG FeFETs exhibit reliability on par with their planar counterparts, with endurance exceeding $2\times 10^{{9}}$ cycles and retention extrapolated beyond ten years. Our work presents a promising approach to addressing the area-inefficiency concern of MFMIS FeFETs.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Lightwave Technology
- Vol. 44  (20) , 1-10
In-situ Trainable Non-volatile Photonic Neural Network
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Journal of Lightwave Technology
- Vol. 44  (20) , 1-10
In-situ Trainable Non-volatile Photonic Neural Network
- 26 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72  (11) , 6034-6040
BTI Reliability of IGZTO FETs: Hydrogen Dynamics, AC/DC Stress Effects, and Advanced Modeling
This article presents an in-depth investigation into the role of hydrogen (H) in the reliability of indium–gallium–zinc–tin-oxide (IGZTO) field-effect transistors (FETs) under both direct current (dc) and alternating current (ac) stress conditions. By systematically examining the effects of H on negative bias temperature instability (NBTI) and positive bias temperature instability (PBTI), we provide critical insights into H-induced degradation mechanisms and their response to ac signals. Our findings reveal: 1) dc NBTI exhibits a saturation behavior with lower degradation severity compared to PBTI; 2) H+ ions in NBTI demonstrate a weak response to ac stress, leading to ultrastable ac reliability performance; and 3) electrochemical impedance spectroscopy (EIS) confirms the presence and migration dynamics of H+ ions in IGZTO FETs. A comprehensive model based on the collective behavior of H+ migration is proposed, accurately capturing the threshold voltage shift (ΔVth) under ac and dc BTI conditions. This study provides essential insights for enhancing the long-term reliability of oxide–semiconductor (OS) devices in advanced electronics.
- 26 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72  (11) , 6034-6040
BTI Reliability of IGZTO FETs: Hydrogen Dynamics, AC/DC Stress Effects, and Advanced Modeling
This article presents an in-depth investigation into the role of hydrogen (H) in the reliability of indium–gallium–zinc–tin-oxide (IGZTO) field-effect transistors (FETs) under both direct current (dc) and alternating current (ac) stress conditions. By systematically examining the effects of H on negative bias temperature instability (NBTI) and positive bias temperature instability (PBTI), we provide critical insights into H-induced degradation mechanisms and their response to ac signals. Our findings reveal: 1) dc NBTI exhibits a saturation behavior with lower degradation severity compared to PBTI; 2) H+ ions in NBTI demonstrate a weak response to ac stress, leading to ultrastable ac reliability performance; and 3) electrochemical impedance spectroscopy (EIS) confirms the presence and migration dynamics of H+ ions in IGZTO FETs. A comprehensive model based on the collective behavior of H+ migration is proposed, accurately capturing the threshold voltage shift (ΔVth) under ac and dc BTI conditions. This study provides essential insights for enhancing the long-term reliability of oxide–semiconductor (OS) devices in advanced electronics.
- 8 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72  (5) , 2334-2340
Comprehensive Investigation of Bias Stress-Induced Instabilities in Highly Scaled ZnO FeFETs: Impact of Channel Thickness, Channel Length, and Switching Cycles
We present a comprehensive study of instabilities induced by positive and negative bias stress (PBS/NBS) in zinc oxide (ZnO) ferroelectric field-effect transistors (FeFETs), focusing on the dependence of threshold voltage ( ${V} _{\text {TH}}$ ) and memory window (MW) dynamics on channel thickness ( ${T} _{\text {CH}}$ ), channel length ( ${L} _{\text {CH}}$ ), and switching cycles. Based on Zr-doped HfO2 (HZO) and by optimizing ${T} _{\text {CH}}$ and scaling ${L} _{\text {CH}}$ down to 45 nm, high-performance ZnO FeFETs with an HZO-Al2O3–HZO ferroelectric (Fe) stack and an atomic layer deposition (ALD)-deposited channel are realized, achieving a large MW of 3.0 V, robust retention, and high endurance exceeding 108 cycles. Bias stress investigations reveal several key findings. First, devices with thinner ${T} _{\text {CH}}$ exhibit higher ${V} _{\text {TH}}$ susceptibility under PBS and NBS, which attributed to stronger electron trapping effects and the generation of more disorder state (DS) O2- defects, respectively. Thanks to the strengthened effect of NBS-enhanced erasing, thinner ${T} _{\text {CH}}$ results in better MW tolerance during NBS, thus partially offsetting the MW degradation due to polarization pinning. Second, this NBS-enhanced erasing effect is particularly pronounced in short-channel devices ( ${L} _{\text {CH}} = 45$ nm), even leading to an increase in MW. In contrast, during PBS, ${L} _{\text {CH}}$ has little impact on ${V} _{\text {TH}}$ and MW instabilities. Finally, it is observed that the degraded MW in heavily cycled devices can be slightly recovered after NBS.
- 8 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72  (5) , 2334-2340
Comprehensive Investigation of Bias Stress-Induced Instabilities in Highly Scaled ZnO FeFETs: Impact of Channel Thickness, Channel Length, and Switching Cycles
We present a comprehensive study of instabilities induced by positive and negative bias stress (PBS/NBS) in zinc oxide (ZnO) ferroelectric field-effect transistors (FeFETs), focusing on the dependence of threshold voltage ( ${V} _{\text {TH}}$ ) and memory window (MW) dynamics on channel thickness ( ${T} _{\text {CH}}$ ), channel length ( ${L} _{\text {CH}}$ ), and switching cycles. Based on Zr-doped HfO2 (HZO) and by optimizing ${T} _{\text {CH}}$ and scaling ${L} _{\text {CH}}$ down to 45 nm, high-performance ZnO FeFETs with an HZO-Al2O3–HZO ferroelectric (Fe) stack and an atomic layer deposition (ALD)-deposited channel are realized, achieving a large MW of 3.0 V, robust retention, and high endurance exceeding 108 cycles. Bias stress investigations reveal several key findings. First, devices with thinner ${T} _{\text {CH}}$ exhibit higher ${V} _{\text {TH}}$ susceptibility under PBS and NBS, which attributed to stronger electron trapping effects and the generation of more disorder state (DS) O2- defects, respectively. Thanks to the strengthened effect of NBS-enhanced erasing, thinner ${T} _{\text {CH}}$ results in better MW tolerance during NBS, thus partially offsetting the MW degradation due to polarization pinning. Second, this NBS-enhanced erasing effect is particularly pronounced in short-channel devices ( ${L} _{\text {CH}} = 45$ nm), even leading to an increase in MW. In contrast, during PBS, ${L} _{\text {CH}}$ has little impact on ${V} _{\text {TH}}$ and MW instabilities. Finally, it is observed that the degraded MW in heavily cycled devices can be slightly recovered after NBS.
- 8 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72  (5) , 2319-2326
Efficient Large Scale Neural Network Acceleration With 3-D FeNOR-Based Computing-in-Memory Design
In this work, we introduce and experimentally demonstrate a 3-D stacked ferroelectric nor (FeNOR) memory, featuring a back-end-of-line (BEOL) zinc oxide (ZnO) channel, and a metal–ferroelectric–metal–insulator5 semiconductor (MFMIS) unit cell. The main contributions of this work are as follows: 1) enhanced memory window (MW) and high ON/OFF ratio: The MFMIS architecture in 3-D FeNOR enables a tunable and large MW (~4 V), as well as an ON/OFF ratio (Ion/Ioff) of six orders of magnitude; 2) low operation voltage and high endurance: The integration of ferroelectric materials allows for low operation voltages (~4 V) and excellent endurance (107 cycles); 3) efficient neural network implementation: Leveraging the 3-D FeNOR structure, we further develop VGG-16 and ResNet-50 convolutional neural networks that achieve high prediction accuracy, decent area efficiency, and low power consumption. The emergence of 3-D FeNOR technology positions ferroelectric devices as a highly promising candidate for computing-in-memory (CIM) applications.
- 8 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72  (5) , 2319-2326
Efficient Large Scale Neural Network Acceleration With 3-D FeNOR-Based Computing-in-Memory Design
In this work, we introduce and experimentally demonstrate a 3-D stacked ferroelectric nor (FeNOR) memory, featuring a back-end-of-line (BEOL) zinc oxide (ZnO) channel, and a metal–ferroelectric–metal–insulator5 semiconductor (MFMIS) unit cell. The main contributions of this work are as follows: 1) enhanced memory window (MW) and high ON/OFF ratio: The MFMIS architecture in 3-D FeNOR enables a tunable and large MW (~4 V), as well as an ON/OFF ratio (Ion/Ioff) of six orders of magnitude; 2) low operation voltage and high endurance: The integration of ferroelectric materials allows for low operation voltages (~4 V) and excellent endurance (107 cycles); 3) efficient neural network implementation: Leveraging the 3-D FeNOR structure, we further develop VGG-16 and ResNet-50 convolutional neural networks that achieve high prediction accuracy, decent area efficiency, and low power consumption. The emergence of 3-D FeNOR technology positions ferroelectric devices as a highly promising candidate for computing-in-memory (CIM) applications.
- 11 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72  (4) , 1788-1794
Unveiling Ferroelectric HZO Cryogenic Performance (4–300 K): Kinetic Barrier Engineering and Underlying Mechanism
In this work, we perform comprehensive and in-depth investigation of the cryogenic characteristics of ferroelectric (FE) hafnium zirconium oxide (HZO) thin films with varying thicknesses (3/5/7/10 nm) across a broad temperature range (4~300 K), assisted by the extensive material and electrical characterizations. We discover: 1) 3 and 5 nm HZO films exhibit distinct temperature dependence in remnant polarization (Pr) and coercive field (Ec) as compared with 7 and 10 nm ones owing to different phase transition mechanisms and 2) the concentration and location of oxygen vacancies act as pivotal factors influencing the pinning effect as well as the trap-assisted-tunneling process, thereby affecting the temperature-dependent behaviors of Pr and Ec. Building upon these insights, we propose and experimentally demonstrate an innovative cryogenic barrier engineering approach for Pr enhancement, particularly valuable for ultra-thin HZO films.
- 11 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72  (4) , 1788-1794
Unveiling Ferroelectric HZO Cryogenic Performance (4–300 K): Kinetic Barrier Engineering and Underlying Mechanism
In this work, we perform comprehensive and in-depth investigation of the cryogenic characteristics of ferroelectric (FE) hafnium zirconium oxide (HZO) thin films with varying thicknesses (3/5/7/10 nm) across a broad temperature range (4~300 K), assisted by the extensive material and electrical characterizations. We discover: 1) 3 and 5 nm HZO films exhibit distinct temperature dependence in remnant polarization (Pr) and coercive field (Ec) as compared with 7 and 10 nm ones owing to different phase transition mechanisms and 2) the concentration and location of oxygen vacancies act as pivotal factors influencing the pinning effect as well as the trap-assisted-tunneling process, thereby affecting the temperature-dependent behaviors of Pr and Ec. Building upon these insights, we propose and experimentally demonstrate an innovative cryogenic barrier engineering approach for Pr enhancement, particularly valuable for ultra-thin HZO films.
- 4 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72  (5) , 2725-2732
Efficient Responsivity Enhancement of Surface-Illuminated InGaAs Avalanche Photodiode by Photon Trapping on the SOI Substrate
The short-wave infrared avalanche photodiodes (SWIR-APDs) with high signal-to-noise ratio (SNR) and high bandwidth are essential components for optical interconnections and large-scale photonic integrated circuits (PICs). However, most surface-illuminated APDs rely on the use of a relatively thick absorber for high absorption efficiency, sacrificing the bandwidth and dark current. To address the challenge, we reported InGaAs APDs with a photon-trapping (PT) structure on the silicon-on-insulator (SOI) substrate using a ~250-nm absorber, featuring a record-high responsivity of 0.75 A/W at unit gain with a wavelength of 1626 nm, equivalent to an external quantum efficiency (EQE) of 57.2%, and a low dark current of 51.3 nA at 90% of the breakdown voltage. A broad operation window from -4 to -24.1 V is achieved, with high detectivities of 9.34 × 1010 Jones at -4 V and 3.43 × 1012 Jones at -24.1 V. The work exhibits the dramatic advantage of the PT structure in photon responsivity enhancement with 2.5 times and the feasibility of InGaAs-based APD to integrate with silicon PICs, paving the way for high-efficiency and high-speed detectors targeting a broad range of emerging applications.
- 4 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72  (5) , 2725-2732
Efficient Responsivity Enhancement of Surface-Illuminated InGaAs Avalanche Photodiode by Photon Trapping on the SOI Substrate
The short-wave infrared avalanche photodiodes (SWIR-APDs) with high signal-to-noise ratio (SNR) and high bandwidth are essential components for optical interconnections and large-scale photonic integrated circuits (PICs). However, most surface-illuminated APDs rely on the use of a relatively thick absorber for high absorption efficiency, sacrificing the bandwidth and dark current. To address the challenge, we reported InGaAs APDs with a photon-trapping (PT) structure on the silicon-on-insulator (SOI) substrate using a ~250-nm absorber, featuring a record-high responsivity of 0.75 A/W at unit gain with a wavelength of 1626 nm, equivalent to an external quantum efficiency (EQE) of 57.2%, and a low dark current of 51.3 nA at 90% of the breakdown voltage. A broad operation window from -4 to -24.1 V is achieved, with high detectivities of 9.34 × 1010 Jones at -4 V and 3.43 × 1012 Jones at -24.1 V. The work exhibits the dramatic advantage of the PT structure in photon responsivity enhancement with 2.5 times and the feasibility of InGaAs-based APD to integrate with silicon PICs, paving the way for high-efficiency and high-speed detectors targeting a broad range of emerging applications.
- 27 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72  (5) , 2716-2724
Temperature-Dependent Two-Phase Switching of the HfOâ‚‚-Based Ferroelectric Polarization
In this work, we accurately observed the polarization switching characteristics of the HfO2-based ferroelectric (FE) materials using well-engineered Zr-doped HfO2 (HZO) metal–FE–metal (MFM) capacitors. By mitigating RC delay effects, we were able to report the intrinsic two-phase switching dynamics of HfO2-based FE materials for the first time. Temperature-dependent measurements, ranging from 300 to 10 K, unveiled the underlying mechanisms of switching dynamics, including short-term direct switching and long-term accumulative switching. Leveraging these insights, we developed a novel physics-based switching model capable of replicating two-phase switching behavior across a wide temperature range. The model provides a deeper understanding of the switching process within the multigrains of the polycrystalline FE layer. A thorough understanding of switching dynamics enables more accurate predictions of the operational behavior of HfO2-based FE memories and clarifies the fundamental mechanisms governing their performance.
- 27 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Electron Devices
- Vol. 72  (5) , 2716-2724
Temperature-Dependent Two-Phase Switching of the HfOâ‚‚-Based Ferroelectric Polarization
In this work, we accurately observed the polarization switching characteristics of the HfO2-based ferroelectric (FE) materials using well-engineered Zr-doped HfO2 (HZO) metal–FE–metal (MFM) capacitors. By mitigating RC delay effects, we were able to report the intrinsic two-phase switching dynamics of HfO2-based FE materials for the first time. Temperature-dependent measurements, ranging from 300 to 10 K, unveiled the underlying mechanisms of switching dynamics, including short-term direct switching and long-term accumulative switching. Leveraging these insights, we developed a novel physics-based switching model capable of replicating two-phase switching behavior across a wide temperature range. The model provides a deeper understanding of the switching process within the multigrains of the polycrystalline FE layer. A thorough understanding of switching dynamics enables more accurate predictions of the operational behavior of HfO2-based FE memories and clarifies the fundamental mechanisms governing their performance.