This study demonstrates a hybrid mode-locked fiber laser by incorporating a single-mode fiber-photonic crystal fiber-graded-index multimode fiber-single-mode fiber (SMF-PCF-GIMF-SMF) structure as a saturable absorber (SA) into a nonlinear polarization rotation cavity. The SA exhibits a modulation depth of 5.8% and a saturation intensity of 59.8 μJ/cm2. Stable fundamental mode-locking is achieved at a center wavelength of 1555 nm with a 3 dB spectral width of 7.2 nm. Compared to the hybrid mode-locked laser employing a conventional SMF-GIMF-SMF SA, the proposed SA reduces the fundamental-frequency mode-locking threshold from 120 to 50 mW, improves the signal-to-noise ratio of the RF spectrum from 35 to 48 dB, and decreases the pulse width from 580 to 493 fs. Harmonic mode-locking up to the eighth order is achieved by adjusting the pump power and polarization controllers. The long-term stability of the laser is confirmed by the power fluctuation of the fundamental-frequency pulse train over an 8-h period. This compact, low-cost laser exhibits a high damage threshold and good stability, highlighting its significant potential for both fundamental ultrafast photonics research and industrial applications.
Fan-out wafer-level packaging (FOWLP) has emerged as a crucial technology for high-density chip integration, with redistribution layers (RDLs) playing an important role in interconnection. However, interfacial delamination within RDLs, driven by the mismatch in coefficients of thermal expansion, poses a critical threat to packaging reliability. In this study, the interfacial delamination behavior of RDLs in FOWLP was systematically investigated. Double cantilever beam (DCB) tests showed that the interfacial fracture toughness decreased significantly with increasing temperature. Specifically, the energy release rate of the Cu/LSF60 interface decreased by 49.6% as the ambient temperature increased from 25 to 260 °C. Thermal cycling tests identified interfacial delamination near the center of the RDL, where pronounced stress concentration was predicted by finite element simulations. To improve interfacial adhesion, several surface treatment technologies were evaluated using DCB tests. Hydrogen plasma surface modification increased the energy release rate of the Cu/LSF60 interface by 12%, whereas citric acid treatment increased that of the Si/LSF60 interface by 90%. Further analysis using X-ray photoelectron spectroscopy and atomic force microscopy indicated that these improvements could be attributed to increased surface roughness, which facilitates mechanical interlocking, and the introduction of polar functional groups, respectively.
DNA conformational analysis requires accurate molecule-level delineation, meaning that segmentation is a prerequisite for quantitative analysis of atomic force microscopy (AFM) height maps. Although AFM provides direct height imaging, in grayscale renderings, DNA boundaries are often obscured by limited contrast. Salt deposits and scanning artifacts can produce fragmented traces and cause DNA to be merged spuriously with nearby salt deposits in the segmentation results. AFM images acquired under different scanning conditions often vary in contrast and sharpness, and the resulting degraded appearance challenges the robustness of segmentation. We propose a framework that integrates (i) quantile-bounded RGB mapping, (ii) Sobel-guided attention in UNet skip connections, and (iii) random-mapping training for robustness. With all components enabled (quantile-bounded RGB mapping, Sobel-guided attention at Skip1, and random-mapping training), the final model achieves an intersection over union (IoU) of 0.74 and a Dice score of 0.85 on the AFM test set. To assess robustness to boundary smoothing, we further applied post hoc Gaussian blur to the test inputs. At a Gaussian blur standard deviation of 3 pixels, the model trained with random mapping retains an IoU of 0.46, whereas the counterpart trained with fixed mapping drops to 0.26, an absolute gain of 0.20. Additional Gaussian-noise experiments on raw AFM height maps confirm that random-mapping training improves robustness to stochastic height-domain perturbations. Together, these results indicate that combining imaging-aware height-to-color encoding with training-time randomized mapping and lightweight, gradient-guided reweighting of skip features offers a practical route to robust and stable AFM–DNA segmentation, with potential transferability to other AFM image-analysis tasks.
The 4H polytype of silicon carbide (4H-SiC) has excellent properties such as high electron mobility and high saturation electron velocity; it has therefore become an important material for high-power and high-frequency devices. In the slicing step of wafer fabrication, however, conventional wire-saw cutting is constrained by the inherent high hardness and brittleness of 4H-SiC, causing excessive material removal and surface and subsurface damage. To overcome these limitations, this study proposes a femtosecond laser slicing technology combined with side scribing. A simulation model of the boundary effect in internal laser processing was established to analyze the optical energy distribution under spherical aberration. The simulations revealed boundary regions that resist laser modification because of the coupling between asymmetric focusing and spherical aberration, which degrades the slicing quality of conventional laser processing. Systematic parametric experiments were then used to investigate the influence of the laser processing parameters on the formation of the modified layer. Targeted experiments performed using the optimized processing parameters provided a quantitative characterization of the hard-to-modify regions and guided the application of the side-scribing process. The peeling results demonstrate that the proposed side-scribing-assisted femtosecond laser slicing process effectively reduces the peeling stress and the cracking that occurs during slicing, achieving a 68% reduction in peeling stress and a 27% reduction in the arithmetic mean roughness Sa of the peeled surface, together with the formation of laser-induced periodic surface structures that are better suited to subsequent processing, compared with conventional laser slicing. This study improves the processability of 4H-SiC wafer slicing and provides a new approach to substrate preparation for wide-bandgap semiconductor materials.
The advancement of point-of-care testing (POCT) is critical for decentralized molecular diagnostics, yet current platforms are hindered by inefficient mixing, complex workflows, long turnaround times, and operator dependency. This study presents the acoustofluidic-enhanced integrated system (AEIS), a fully automated platform that integrates a GHz acoustic resonator with a micro-valve control system for nucleic acid testing. The resonator generates intense acoustic streaming vortices, enabling second-scale mixing and reducing nucleic acid extraction time to under 4 min. Coupled with hydroxyl-modified polystyrene microbeads for nucleic acid extraction (98% efficiency) and automated fluidic control via a rotary valve, the system detects EGFR L858R mutations in plasma with a limit of detection of 2.6 copies/μl and completes the sample-to-answer workflow within 30 min. This enclosed, low-cost, and portable platform minimizes contamination and operator intervention, offering a sensitive and efficient solution for on-site testing in infectious disease control and early cancer screening.
Multichannel neural recording requires not only microelectrode arrays (MEAs) with sufficient spatial coverage but also stable front-end acquisition and online signal processing hardware. In this work, a four-shank 32-channel MEA and a field-programmable gate array (FPGA)-based online processing module are developed for multichannel neural recording. The four-shank structure is designed to provide broader spatial coverage while maintaining a compact probe footprint. Platinum nanoparticles (PtNPs) are modified on the electrode surfaces to improve the electrical interface of the recording sites. Electrochemical measurements demonstrate that PtNPs modification significantly improves the interfacial properties of the electrodes. An FPGA-based module is employed to enable communication with the front-end chip and perform online signal processing. Post-implementation verification further demonstrates that the FPGA design satisfies the timing constraints while maintaining acceptable power consumption. The end-to-end performance of the electrophysiological recording chain is evaluated through gain/bandwidth testing and simulated neural signal detection. The system exhibits an average gain of 45.57 dB within the main passband, with a bandwidth ranging from 4.28 to 2.36 kHz. In addition, simulated spike tests demonstrate stable online detection and waveform extraction. In addition, in vivo recordings from the hippocampal CA1 region demonstrate that the proposed system simultaneously acquires clear spike and local field potential signals from multiple channels, with representative spike waveforms exhibiting good consistency. The combination of a four-shank MEA with PtNPs-modified recording sites and FPGA-based online processing supports stable neural signal acquisition both in vitro and in vivo and provides a useful foundation for future neural interface applications.
To enhance the output thrust of a macro–micro precision actuator while maintaining its precision positioning capability, an optimization design method for the macro-motion structure based on Halbach permanent magnet arrays is proposed. By establishing an equivalent magnetic network model of the macro-motion magnetic circuit, the analytical efficiency for the magnetic field distribution is improved. A single-factor analysis method is employed to reveal the influence of key parameters, including axial permanent magnet length, air-gap thickness, and winding thickness, on the maximum output thrust. The Grey Wolf Optimizer is applied to perform comprehensive optimization of the main parameters, yielding a structural dimension combination that maximizes thrust. On the basis of the optimization results, a prototype is fabricated and experimentally tested, and the results show that under a 6A drive current, the macro-motion stage achieves an output thrust of 190.2N, with a relative error of 2.4% compared with the optimized model prediction. Meanwhile, a stroke positioning of 35.32mm is achieved within 25ms, with a maximum nonlinear error of 0.469mm, demonstrating good linearity of displacement output. This method ensures high-thrust output while maintaining rapid precision positioning, providing an effective technical solution for high-load precision drive applications.
The silicon vacancy (VSi) in 4H-SiC represents a promising candidate for solid-state qubits. However, conventional high-temperature annealing typically suffers from undesired impurity diffusion and limited spatial localization during fabrication. Here, we present a femtosecond laser annealing approach for the in situ restoration and property tunability of VSi color centers in proton-implanted 4H-SiC. By delivering extremely high transient peak temperatures within micrometer-scale thermal interaction zones, lattice damage is effectively eliminated and excessive defect diffusion is suppressed. Specifically, the room-temperature photoluminescence intensity was enhanced to 1.5 times its initial value. The zero-phonon line in low-temperature spectra was markedly sharpened with increased intensity, and efficient relaxation of residual stress in the implanted area was verified via Raman spectroscopy. Regarding quantum properties, the optically detected magnetic resonance contrast was increased by a factor of 1.6. Furthermore, the linewidth was significantly narrowed, and both the transverse/longitudinal coherence times and Rabi oscillation contrast were substantially improved. This work demonstrates that femtosecond laser annealing provides a high-spatial-resolution, nonequilibrium processing strategy, offering substantial utility for the fabrication of high-quality VSi quantum light sources with extended coherence times.
Ensuring long-term, stable operation of self-powered systems considering the stochastic and fluctuating nature of ambient energy remains a pivotal challenge, as existing multi-source harvesting architectures are confined to static, non-responsive paradigms. This work presents a self-powered wireless sensing system that transitions the field toward automatic, priority-driven management through a hardware-enforced, quadruple-buffered power scheduling strategy. The core of this strategy is a scheduling protocol that orchestrates energy flow in the following order: real-time photovoltaic (PV) energy, PV-stored energy, real-time thermoelectric-vibration hybrid energy, and auxiliary-stored energy. This is implemented via a dedicated circuit that enables automatic stored energy release and seamless source switching (<20 ms). A critical enabling feature is the electrical isolation between the primary and auxiliary power paths, which prevents mutual interference and resolves the impedance mismatch inherent in conventional static topologies. Experimentally, the system achieves 93 min of operation from a 10-min charge under controlled conditions, surpassing standalone PV-only and thermoelectric-only systems by 47.6% and 116.3%, respectively. Its practical viability is demonstrated by 75 min of stable field operation and consistent performance over a 10-day continuous test, with an overall system efficiency of 75%. This work provides a robust, maintenance-free power solution, demonstrating significant potential for sustainable wireless sensing in intelligent transportation and broader Industrial Internet of Things applications.
This study characterizes the effects of dynamic pulsing parameters on hydrogen (H2) gas detection using a novel microheater-integrated metal oxide semiconductor (MOS) sensor. The proposed single-device system operates through duty cycling by utilizing repeated pulsed power inputs. The impact of pulse characteristics, specifically the pulse period and duty cycle, on the overall sensor performance is systematically evaluated. The results demonstrate that operating at an optimal pulse period of 2.4 s with a 67% duty cycle provides a 32% enhancement in the hydrogen response without compromising the response time. In addition, this designed duty-cycling strategy significantly reduces energy consumption, achieving a low average power consumption of ∼19.8 mW. Fabricated using wafer-level batch processing techniques, this compact and energy-efficient sensor platform represents a highly scalable and commercially viable solution for integration into power-constrained Internet of Things monitoring applications.
To meet the demand for rapid on-site detection of trace Hg2+ in aquatic environments, this paper presents an integrated microfluidic electrochemical sensor based on micro-electromechanical systems technology. The core component of the sensor is a Au-based micro-interdigitated array (MIDA) electrode, which utilizes the efficient pre-enrichment of Hg via underpotential deposition on Au surfaces to form Au–Hg amalgams, combined with the radial diffusion effect of microelectrodes. Finite element simulations were conducted to investigate the effects of electrode width and spacing-to-width ratio on the electrochemical characteristics and detection performance of the MIDA. The results indicate that smaller electrode widths and spacing-to-width ratios enhance radial diffusion and redox cycling, thereby enabling faster attainment of steady-state conditions and higher response currents. The preconcentration process of differential pulse anodic stripping voltammetry was optimized using an in situ flow enrichment strategy, thereby significantly enhancing detection sensitivity. The structurally optimized MIDA chip exhibited a linear detection range of 0.5 to 12 ppb, with a detection limit of 0.1 ppb (S/N=3), along with excellent repeatability (relative standard deviation<5.0%) and reproducibility. The sensor was applied to the analysis of tap water and lake water, achieving spike recoveries of 101.4%–107.8% and 94.0%–100.2%, respectively, thereby validating its reliability for monitoring Hg in real water samples.
A virtual platform for modeling tapping mode atomic force microscopy (TM-AFM) based on Simulink is proposed. The platform integrates detailed models of cantilever dynamics, tip–sample interactions, and closed-loop feedback control, enabling precise prediction of measurements and internal parameter fluctuations through multiphysics co-simulation. The simulations provide effective predictions of system variations when scanning on multi-material surfaces, which reveal that the adhesion force significantly increases the actual amplitude of the cantilever while reducing the system stability. These effects are related to phase lag variations. Experimental validation is performed on a quartz–Au sample at different regions. By comparing with the measurements in contact mode, the experiments confirm that higher adhesion on quartz leads to measurable height errors and distinct phase responses, which agrees closely with simulated predictions. The platform effectively forecasts optimal proportional–integral–derivative (PID) gains for different materials, demonstrating its utility for pre-experimental optimization and the development of adaptive control strategies in TM-AFM.
Aluminum alloy micro-grooved miniature heat pipes offer significant advantages for dissipation of high heat flux densities in confined spaces within thermal management systems of electronic equipment in aerospace and other fields, owing to their high thermal conductivity, miniaturization, and lightweight properties. However, the insufficient capillary performance of the micro-grooved wicks in aluminum alloy miniature heat pipes has become a key factor limiting their heat transfer efficiency. This study addresses the enhancement of capillary performance in aluminum alloy heat pipes by proposing a self-sealing hydrothermal process (SSHP) to construct micro/nanostructures on the surfaces of micro-grooved heat pipes, thereby improving capillary performance. The study investigates the ways in which the SSHP process affects aluminum alloy surface wettability and the underlying mechanisms, analyzes the effects of SSHP process parameters on the capillary performance of aluminum alloy micro-grooved miniature heat pipes, and optimizes these parameters. The results demonstrate that SSHP treatment induces a superhydrophilic transformation on the aluminum alloy surface, strongly correlated with the shape characteristics of the constructed surface nanostructures. A smaller average aspect ratio and longer average length of boehmite nanostructures correlate with increased hydrophilicity of the aluminum alloy surface. Following SSHP treatment, the capillary performance of aluminum alloy micro-grooved heat pipes increases by up to 2679%. The mechanism behind this enhancement is elucidated, and optimal SSHP process parameters for aluminum alloy micro-grooved heat pipes are proposed. This study provides theoretical and technical support for the fabrication of high-capillary-performance aluminum alloy miniature heat pipes.
Sleep is an essential physiological process for maintaining survival and normal circadian rhythms in living organisms. Research on sleep regulatory mechanisms suffers from a lack of real-time electrophysiological recording and neuromodulation techniques targeting deep brain regions. In this study, we design and fabricate a microelectrode array for detecting and modulating sleep–wake activity in the preoptic area. We record electrophysiological signals associated with sleep–wake regulation in the medial preoptic area. Our findings reveal distinct behavioral and electrophysiological characteristics between sleep and wake states, and we identify different subtypes of neurons in the medial preoptic area that exhibit fundamentally distinct response mechanisms to sleep and wake. In addition, we successfully use deep brain electrical stimulation technology to induce mice to enter a sleep-like state. These findings fill a critical gap in the understanding of neural mechanisms underlying sleep–wake regulation and provide a novel technological approach for sleep modulation via deep brain stimulation.
Rapid advances in compact electronic devices have led to severe thermal management challenges in inherently confined internal spaces, where pump characteristics are required for air coolers to maintain their cooling effectiveness. In the present work, for the first time, a silicon-based piezoelectric jet pump is developed as a MEMS cooler. This MEMS air cooler with backflow suppression is realized by incorporating a piezoelectric MEMS actuator array with a multi-orifice outlet configuration. With a compact footprint of 9 × 9 × 1.7 mm3, the cooler achieves a back pressure of 300 Pa and a volume-normalized flow rate of 2.3 L ⋅ min−1 ⋅ cm−3. With a power consumption level of only 11 mW, it removes ∼500 mW of heat flow rate at a heater–cooler distance as small as 1 mm. In addition, quiet operation can be achieved by extending the driving frequency to the ultrasonic range. This MEMS air cooler represents an ideal solution with a compact size, high flow rate, low power, and low noise, showing strong potential for revolutionizing the landscape of miniaturized active cooling.
Piezoelectric actuator-driven nanopositioning systems offer advantages such as fast response and high resolution. However, owing to inherent hysteresis nonlinearities and external disturbances, achieving both high-speed and high-precision trajectory tracking remains challenging for such systems. Therefore, this paper proposes a reaching-phase free recursive terminal sliding mode control (RPF-RTSMC) based on a modified adaptive Hammerstein disturbance observer-based (MAH-DOB) approach. First, the RPF-RTSMC is designed on the basis of a recursive combination of fast terminal and integral sliding modes. Then, through optimization of the initial value configuration, the convergence of the sliding mode is accelerated. This adjustment enables the system to maintain global robustness capability throughout the entire control process. Furthermore, the MAH-DOB approach is integrated into the RPF-RTSMC for the estimation and compensation of lumped disturbances (e.g., rate-dependent hysteresis and uncertainties), thus improving disturbance rejection and mitigating control chattering. A stability analysis of the composite closed-loop system is performed via Lyapunov theory. Finally, experimental results demonstrate that the proposed method outperforms conventional fast nonsingular terminal sliding-mode controller schemes based on a nonlinear disturbance observer in terms of convergence error suppression and disturbance rejection capability.
Artificial intelligence (AI) has been widely adopted in augmented training systems for instrumental music performance and athletic skill development. However, current systems primarily rely on computer vision-based motion tracking, which faces inherent limitations such as restricted accuracy, operational complexity, susceptibility to environmental interference, and privacy concerns. Additionally, these systems often fail to detect subtle performance deviations invisible to visual sensors. To address these challenges, we developed a smart wristband that leverages the triboelectric effect to monitor the visible and invisible motion during sports and musical instrument practice. Integrated with machine learning algorithms, the wristband accurately distinguishes correct from incorrect drumming motions and recognizes various badminton hand gestures, achieving an accuracy of 98.3%. Moreover, the device employs low-cost materials of silicone and fluorinated ethylene propylene, ensuring affordability and scalability. By providing reliable motion analysis without the constraints of camera-based systems, this technology offers a robust and practical solution for next-generation training assistance.
A high-quality n-type 4H–SiC epitaxial layer was grown on a 200-mm, 4° off-axis 4H–SiC substrate using a horizontal hot-wall single-wafer chemical-vapor deposition reactor. The uniformity and defects of the epitaxial wafer were characterized using Fourier-transform infrared spectroscopy, the mercury-probe capacitance–voltage method, and a KLA-Tencor Candela tool. Through a combination of simulations and experiments, the effects of the hydrogen flow rate and the critical carbon-to-silicon ratio on the growth rate and defect formation were studied and optimized. As a result, high-quality 200-mm 4H–SiC epitaxial wafers were obtained, with a thickness uniformity of 0.86% and a doping-concentration uniformity of 1.38%. The useable-area yield reached 98.72% with a site size of 5 × 5 mm2. To assess production quality, data from more than 75 batches were analyzed statistically. The results showed average uniformities of 1.00% for thickness and 1.92% for doping concentration, with an average projected yield of 97.10% for 5 × 5-mm2 dies. Furthermore, 1200 V MOSFETs fabricated using these wafers exhibited an average device yield of 95.90%. Together, these multidimensional data demonstrate that the 200-mm 4H–SiC epitaxial wafer process has robust batch-production capabilities.
Existing microfabrication techniques for producing featured friction stir processing (FSP) tools are restricted to certain alloys, and generating precise microfeatures on the shoulder of a tungsten carbide (WC) FSP tool is particularly challenging. Conventional microfabrication methods such as micromilling and microturning require costly tooling, which limits their use on hard-to-cut materials. In this work, the capabilities of an emerging nonconventional microfabrication technique, reverse micro-electrical discharge machining (reverse-μEDM), are exploited to produce precise triple-spiral microgrooves on a WC FSP tool. The highly precise featured tool plate that is integral to the reverse-μEDM setup was first prepared using two different technologies: laser micromachining of titanium sheets, and micromilling of copper and brass sheets (chosen in view of their higher reflectivity). Reverse-μEDM was then performed using each tool plate, and the material removal rate, tool wear rate, and surface roughness of the fabricated microgrooves were measured as the output responses. The titanium tool plate performed best in generating the FSP tool microfeatures because of its favorable physical properties. The copper and brass tool plates had higher wear ratios than the titanium tool plate but produced poorer surface characteristics and lower geometrical precision. The grooves produced with the titanium tool plate had a precise depth of 150 μm with negligible wall taper along their entire length. To validate the feasibility of the fabrication, a pilot FSP experiment was also conducted on a cold-sprayed Inconel 718 substrate.
Understanding the evolution mechanism of alloy organization during selective laser melting (SLM) is crucial for adjusting process parameters and determining structural and organizational properties. However, it is very difficult to capture the corresponding microstructural change. This study employs molecular dynamics simulations to explore the microstructural evolution and laser melting behavior of Ti6Al4V nanoparticles at the atomic scale, which deepens the basic understanding of SLM. The nanoparticle size effect, structural changes, and sintering mechanism are investigated by establishing nanoparticle systems. The results demonstrate that surface diffusion and volume diffusion dominate the melting process, and surface atoms are more prone to pre-melting and violent movement. The melting temperature of nanoparticles depends on their size, with smaller-sized nanoparticles having a lower melting temperature. Varying size ratios in the two-nanoparticle system result in inconsistent diffusion and wetting rates of smaller nanoparticles. As the size ratio increases, small nanoparticles more easily diffuse and wet large ones. Furthermore, the melting behavior of a multiple-nanoparticle system is investigated.
