Ion-migration memristive synaptic devices provide a physical route for hardware neuromorphic computing by using ionic redistribution, conductive-filament evolution, and interfacial barrier modulation to regulate synaptic weights. However, existing studies are often discussed according to specific material systems or individual device demonstrations, which makes it difficult to compare how different ion-migration mechanisms determine synaptic behavior, device stability, and integration potential. This mini review organizes recent progress from the perspective of operating mechanism and device type. Electrochemical metallization (ECM)-based synaptic devices are discussed with emphasis on metallic-filament formation and the challenge of achieving gradual and reproducible conductance modulation. Filamentary valence-change memory (VCM)-based devices are reviewed in terms of oxygen-vacancy channel evolution, while interfacial VCM devices are examined through interfacial ionic modulation and barrier-controlled analog switching. Hybrid ECM-VCM devices are further discussed as integrated designs that couple multiple ionic processes to balance switching window, stability, and multifunctionality. By linking mobile ionic species, switching pathways, and synaptic functions, this review provides a mechanism-based framework for understanding ion-migration memristive synaptic devices and for identifying the material, interface, and device-level issues that remain before scalable neuromorphic hardware can be realized.
Just Accepted manuscripts are peer-reviewed and accepted for publication. They are posted online prior to technical editing formatting for publication and author proofing.
Copolymer organic semiconductors (OSCs) are promising for flexible and low-cost electronics. Although doping is commonly used to improve their charge transport properties, its effect on breakdown behavior remains unclear. This work systematically investigates the breakdown characteristics of doped copolymer-based Schottky junctions, employing two representative copolymers and dopants with distinct mechanisms. The results reveal that doping leads to an enhancement in breakdown voltage (BV), which contradicts the behavior of conventional semiconductors. Concurrently, the reduced Schottky barrier height accounts for the observed increase in leakage current. Subsequently, through UV-VIS-NIR absorption spectroscopy, we uncover the doping-induced bandgap broadening (DBB) effect, which raises the energy threshold for impact ionization and thereby accounts for the enhanced BV. To bridge the device performance with DBB effect, SILVACO simulations are employed, and the simulation results confirm that the DBB effect is pivotal to the enhanced BV. This work provides new insights into the doping modulated breakdown mechanism in copolymer OSCs and highlights their potential for high performance organic power devices.
Nanowires of narrow bandgap III–V semiconductors are very promising for efficient light-emitting and photosensitive devices operating in the mid-infrared spectrum range. Low-temperature infrared photoluminescence spectra of InAs and InAsP nanowires, including CaF2-passivated InAsP nanowires grown on a silicon substrate, were studied. The features in the spectra of InAsP nanowires were explained by the formation of a random close-packed stacking crystal structure with a predominance of the hexagonal (wurtzite) phase in nanowires and non-uniform distribution of stacking faults along their growth axis, as well as by radiation from parasitic bulk islands with the sphalerite phase. In structures without surface passivation, a red shift of spectrum was observed with increasing lattice temperature, consistent with the temperature dependence of the band gap. Passivation of the surface of InAsP nanowires using CaF2 leads to a nontrivial behavior of the spectral position of the photoluminescence intensity maximum due to a decrease in the contribution of surface states to nonradiative recombination at temperatures below 77 K. Moreover, it is shown that passivation allows to observe a photoluminescence signal at higher lattice temperatures compared to nonpassivated structures. The pump power dependence demonstrated a blue shift of the spectrum with increasing optical power, attributed to the band-filling effect (shift in the quasi-Fermi level)
With the rapid development of the Internet of Things (IoT) and wearable electronics, tactile sensors play an indispensable role in intelligent sensing systems. However, traditional tactile sensing systems follow the von Neumann architecture, where sensors and processing units are physically separated. This leads to frequent data transfer of large raw data volumes, causing high latency and energy consumption. Such bottlenecks cannot meet the requirements of real-time closed-loop control and edge intelligence. Inspired by the highly integrated "perception-storage-computation" mechanism of biological sensory systems, memristor-based neuromorphic computing offers a groundbreaking solution beyond conventional approaches. Memristors combine non-volatile storage with tunable resistance. They enable in-situ emulation of synaptic plasticity, in-memory computing, and brain-inspired processing, thereby holding the potential to significantly improve the energy efficiency and response speed of tactile systems. This review systematically discusses the physical mechanisms of mainstream memristors, highlights recent progress in memristor-based neuromorphic computing for tactile sensing, and outlines key challenges and future directions for neuromorphic tactile perception systems.
Aluminum scandium nitride (AlScN) features large remanent polarization (Pr) and high Curie temperature (TC), making it highly promising for next-generation nonvolatile memory and neuromorphic devices. However, the imprint-induced electrical asymmetry severely limits their practical application. Here, we fabricate Mo/Al0.8Sc0.2N/Mo ferroelectric capacitors (FeCap) via continuous in-situ magnetron sputtering. Characterization and measurements reveal a pronounced imprint effect, which is closely related to the lattice mismatch near the bottom electrode (BE). A serial configuration with shared BE was proposed that naturally forms a differential pair, compensating for the built-in voltage (Vbi). Then devices can achieve symmetric consecutive cycles, with the retention time of 105 s, and the breakdown-to-coercive-voltage ratio (VBD/VC) of 1.20. Leveraging their dynamic response, we design an architecture of FeCap integrated with a shallow convolutional neural network (CNN). It can achieve background suppression and edge enhancement of the Fashion MNIST dataset at FeCap. CNN only undertakes feature fusion and classification tasks, thus significantly reducing the computational complexity. The recognition accuracy of 94.31% can be achieved after only 30 epochs. This work provides an effective approach for imprint suppression and demonstrates the potential of AlScN-based FeCap for image recognition acceleration.
Ga2O3 is a leading candidate for next-generation high-power electronics and deep-ultraviolet optoelectronics due to its ultrawide bandgap and high theoretical breakdown field. While the stable β-phase has been extensively studied, interest is shifting toward metastable polymorphs to broaden functionality. Among these, the ε/κ phase is distinguished by its non-centrosymmetric orthorhombic structure, offering unique ferroelectric and piezoelectric properties. Despite progress in epitaxial stabilization, a comprehensive understanding of its crystal structure, physical behavior, and device potential remains limited. This review critically examines the ε-κ phase relationship and heteroepitaxial growth mechanisms on Al2O3, GaN, and other substrates, highlighting the roles of supersaturation, doping, and buffer layer engineering in stabilizing metastable phases while suppressing secondary phases. Furthermore, the work explores the unique physical properties of ε/κ-Ga2O3, including the substantial spontaneous polarization, a large piezoelectric coefficient, and the direct-bandgap nature favorable for DUV detection. The practical efficacy of these attributes is showcased across a diverse technological portfolio, including ferroelectric memristors for neuromorphic computing, high-frequency acoustic wave resonators, and high-sensitivity solar-blind photodetectors. Finally, the review identifies remaining challenges-such as rotational domain mitigation and polarization saturation-providing a strategic roadmap for realizing multifunctional ε/κ-Ga2O3-based devices.
AlGaN/GaN-based Schottky barrier diodes grown on Si substrates with n-GaN/p-NiO/n-GaN lateral junctions are proposed and fabricated. The n-GaN/p-NiO/n-GaN lateral junctions effectively extend the depletion region in the GaN drift layer. The more homogenized electric field reduces the electric field magnitude near both the Schottky contact and the field plate (FP) edge. Consequently, the fabricated lateral power device achieves a breakdown voltage (BV) of ~1.6 kV and yields a high Baliga’s figure of merit (BFOM) of ~1.47 GW·cm−2. The n-GaN/p-NiO/n-GaN lateral junctions also contribute to suppressing the electron trapping effect and reducing the dynamic specific on-resistance (Ron,sp) by 1.5 times. With the developed physical models, we have also investigated the impact of different p-NiO designs on the blocking effect for the proposed device. We report that the n-GaN layer in the n-GaN/p-NiO/n-GaN lateral junctions shall be effectively depleted so that AlGaN/GaN-based Schottky barrier diodes with high BV can be achieved.
Optical detectors based on metal oxide semiconductor thin films have always been a research hotspot in the field of optoelectronic detection due to their advantages of simple structure and convenient preparation. However, due to the characteristics of the material itself and the inherent working mechanism of the device, these devices often have their own shortcomings. In this work, Eu doped BaTiO3 thin film was introduced as a buffer layer to enhance the performance of photoconductive detector based on ZnO thin film. Compared with the un-doped BaTiO3 buffer layer, the Eu:BTO buffer layer can further enhance the photodetection performance of detector based on ZnO film. The detector on Eu:BTO buffer layer exhibits larger photocurrent (450 nA) and smaller dark current (1.01 nA) compared with that on un-doped BTO buffer layer (100 nA and 10 nA), leading to larger on/off ratio. Meanwhile, the response speed of the device can reach 0.51 s and 1.04 s, respectively, which is much superior than the device based on original ZnO film (10.55s and 14.98 s). At the appropriate doping concentration (1% Eu:BTO), the increased polarization electric field is the main reason for achieving performance enhancement.
A yield-aware optimization strategy is proposed to address the Trapped Charge Mismatch Effect (TCME) in split-gate SONOS flash by bridging device physics with lithography constraints. The study demonstrates that scaling the memory gate length (LMD) induces a lateral field extension that physically minimizes the region of residual trapped electrons. However, a fundamental physics-lithography trade-off is identified: while aggressive scaling maximizes erase efficiency, Process Variation Band (PVB) analysis mandates a 50 nm manufacturable optimum to maintain Critical Dimension (CD) variation within the 10% safety threshold. Experimental verification on an 8 Mb array confirms the strategy's effectiveness, achieving 99.3% yield and suppressing bit-to-bit disturbance to 10.9% of the memory window (MW). Supported by endurance exceeding 105 cycles, these results demonstrate a viable high-yield pathway for scaling high-reliability memories.
This work demonstrates high-performance quasi-vertical GaN Schottky barrier diodes (SBDs) on sapphire substrates utilizing a regrown p-GaN sidewall structure. The regrown p-GaN sidewall alleviates electric-field crowding at the Schottky-contact edge and suppresses sidewall-associated reverse leakage, leading to a dramatic enhancement in breakdown voltage (BV) from 71 to 847 V. Furthermore, the reverse leakage current density is suppressed by five orders of magnitude, decreasing from 0.29 to 5.2 × 10−6 A/cm2 at a cathode bias of −50 V. Two-dimensional TCAD simulations were performed to verify the field-modulation effect of the p-GaN sidewall. The quasi-vertical SBDs with p-GaN sidewall demonstrate robust stability under both thermal and reverse-bias stress, highlighting the potential of regrown sidewall architectures for the next generation of GaN-on-sapphire power electronics.
The mechanism of the etching process that ensures high yield in spin-orbit torque magnetic random-access memory (SOT-MRAM) featuring a novel channel-less (CHL) structure is systematically investigated in this paper. A steep sidewall morphology is identified as particularly favorable, and the precise relationship between the pillar profile and yield is established and thoroughly analyzed. When the step angle exceeds 80°, a larger net etching rate along the sidewall is achieved, enabling effective removal of metal residues near the tunnel barrier layer while minimizing plasma-induced damage at the MgO/CoFeB interface. The optimized devices exhibit an excellent thermal stability factor (Δ) of over 95 at room temperature, demonstrating superior operational reliability. Thanks to process optimizations targeting re-deposition removal, the in-die functional yield of the fabricated 16 Kb CHL SOT-MRAM array reaches an impressive 60 ppm. These advancements pave the way for high-yield and reliable mass production of SOT-MRAM devices.
We demonstrate a significant light-triggered current enhancement in SiC bipolar PIN diodes through ultraviolet (UV) light injection, achieving a 2.74× current gain at 3.5 V under 365 nm illumination. By integrating an optical window for targeted photon injection into the intrinsic region, UV-generated electron-hole pairs enhance conductivity modulation and lower the forward voltage drop. Crucially, we reveal a pronounced structural dependence: vertical devices outperform lateral counterparts due to more effective utilization of photogenerated carriers. This advantage stems from shorter carrier transport path of the vertical architecture and reduced surface recombination losses, whereas lateral devices suffer from enhanced carrier recombination due to surface-parallel carrier drift. This study establishes UV illumination as a scalable and non-invasive strategy for dynamic performance tuning of SiC bipolar devices, bypassing complex lifetime-enhancement processes while enabling adaptive conduction for smart power systems.
InGaAs/InP single photon diodes (SPADs) are widely used in quantum communication systems. The dark count rate (DCR), which describes the noise level, is one of the most important parameters of SPAD performance. Here, we demonstrate the technology computer-aided design and experimental test of low DCR InGaAs/InP SPAD to be applicable to the fiber quantum key distribution system under high-frequency gating. In order to achieve a lower DCR at higher operating temperature, the device structure is optimized by increasing the doping concentration of the charge layer and expanding the width of the multiplier layer. At the same time, the charge persistence effect is limited by optimizing the double Zn diffusion process. The results show that our InGaAs/InP SPAD can achieve an extremely low DCR of 0.1 kcps, 30% photon detection efficiency and 4.7% afterpluse probability at an operating frequency of 1.25 GHz, an operation temperature of 213 K and an excess bias voltage of 4.6 V.
Germanium−on−Silicon (Ge−on−Si) Avalanche Photodiodes (APDs) have attracted significant interest for LIDAR sensor applications over the last decade. However, further improvements are still needed to reduce dark current and excess noise while increasing the gain to fit the application requirements. Moreover, designing a good charge layer in separate absorption, charge and multiplication (SACM) structure for an APD is essential to control the electric field intensity in both the multiplication layer and germanium absorption layer. In this study, three different silicon charge layer thicknesses (80, 100, and 120 nm) and two boron doping concentrations of 2.5$ \times $1017 at.cm−3 and 5$ \times $1017 at.cm−3 were fabricated and characterized in vertical Ge−on−Si APDs. Our investigation focuses on the boron concentration of 2.5$ \times $1017 at.cm−3 to identify the physical contributions to current and noise at room temperature and a wavelength of 1550 nm. The device with a charge layer thickness of 80 nm exhibits the lowest equivalent input noise (NEI) of 1.7 pA/√Hz at a gain of 4.1, which is compatible with LIDAR application requirements. The results indicate that there may be an optimal combination of charge layer thickness and doping concentration for minimizing APD noise, which should be considered in future APD designs.
Dual-band photodetectors represent one of the important development directions of the third-generation focal plane photodetectors. Type II superlattice (T2SL) has emerged as a promising candidate for fabricating long-wavelength dual-band or multi-band infrared photodetectors due to its merits including engineerable band gap, low manufacturing cost and excellent uniformity. In this work, a long/long-wavelength dual-band photodetector with an NMπP-B-PπMN structure based on InAs/GaSb T2SL is reported. The cutoff wavelengths of the two bands are 7.8 and 11.2 μm, respectively. At 77 K, the short long-wavelength channel exhibits a peak quantum efficiency of 20.19%, a peak detectivity of 1.26 × 1010 Jones and dark current density of 0.91 × 10–2 A/cm2; the long long-wavelength channel achieves a peak quantum efficiency of 28.94%, a peak detectivity of 7.36 × 109 Jones and dark current density of 6.03 × 10–2 A/cm2. The device demonstrates high performance, laying a solid foundation for the fabrication of long/long-wavelength dual-band focal plane photodetectors.
AlGaN-based ultraviolet (UV) laser diodes (LDs), with emission wavelength in the 280–365 nm range, are promising for applications in medical diagnostics and biological sensing, making them a prominent research focus in both academia and industry in recent years. A key challenge in their development is the large stress induced during the epitaxial growth of LD structures, which arises from the lack of lattice-matched substrates, and severely degrades the quantum efficiency and overall LD performance. This study presents an in-depth investigation into the growth mode and stress evolution of thick Al0.2Ga0.8N template. Firstly, we used the compressive stress between the Al0.2Ga0.8N layer and AlN/Sapphire substrate to form spontaneously three-dimensional growth to annihilate dislocations. Secondly, based on the Nakajima's theory of elasticity, we refined the conventional theoretical models for AlGaN strain relaxation of the S-K growth mode and critical thickness by considering the crucial role of threading dislocations (TDs) in releasing compressive stress. The experimentally measured critical thickness for three-dimensional growth was consistent with the calculated results. Furthermore, a crack-free high-quality 5 μm-thick Al0.2Ga0.8N template was successfully grown on an AlN/sapphire substrate.
The thermal sensitivity of phase-change memory (PCM) poses a stringent thermal budget for back-end encapsulation, demanding high-performance diffusion barriers processable at low temperatures. Conventional low-temperature silicon nitride (SiNx) films, however, are typically porous and prone to oxidation due to abundant metastable Si–H/N–H bonds. Herein, we propose an in-situ plasma cycling strategy that reconstructs the bonding network of plasma-enhanced chemical vapor deposition (PECVD) SiNx at a record-low temperature of 200 °C. Through controlled Ar/N2 plasma exposure, we cleave metastable bonds and reorganize into a continuous Si–N network, achieving a near-theoretical density of 3.4 g/cm3 (a 61.9% increase) and a 143.8% enhancement in Si–N bonding proportion. The resulting 40-nm barrier effectively suppresses Te/O interdiffusion, reduces wet-etch rate by ~67%, and maintains thermal confinement within 1.6% deviation. Integrated into PCM devices, this barrier yields a 98.7% SET/RESET operation yield and a 1.4-fold wider resistance window. This work not only provides a reliable encapsulation solution for PCM but also establishes a generalizable plasma-mediated interfacial engineering approach for advanced electronic devices under thermal constraints.
