共找到 20 条结果
High-performance broadband photodetector arrays are pivotal for sensing, imaging and optical communication technologies, yet scalable materials that simultaneously offer broadband photoresponse coverage, fast response speed and uniform device performance remain elusive. Here, we present a 3×5 photodetector array based on few-layer MoTe2 grown via chemical vapor deposition (CVD) method. The MoTe2 photodetector exhibits a stable photoresponse ranging from ultraviolet to near-infrared regions. The device features a responsivity (R) of 3.82 mA/W, an external quantum efficiency (EQE) of 1.17% and a detectivity (D*) of 1.62×107 Jones under 405 nm illumination (5.04 mW/cm2) at Vds = 2 V and Vgs = 0 V. The response time and recovery time remain on the millisecond scale throughout the entire photoresponse coverage. This work provides a feasible route toward the construction of high-performance broadband photodetector arrays using single material alone.
To overcome limitations in the applications of existing SNP arrays in cotton genotyping and genomic selection (GS), we developed a liquid-phase SNP array, CottonSNP10K, for genomics-assisted breeding in cotton. Based on the high-quality reference genome of modern upland cotton cultivar NDM8, CottonSNP10K achieves precise probe design, and its marker system innovatively integrates the modern breeding genetic background, incorporating not only 13 agronomic traits associated loci (including fiber quality and yield-traits and stress resistance) identified via genome-wide association studies (GWAS), but also six exogenous gene markers targeting traits such as high lint percentage, herbicide resistance and insect resistance. The chip incorporates genome-wide background SNPs to ensure comprehensive genetic coverage, resulting in a final design comprising 11,159 SNPs, including 3,981 functionally trait-associated markers with 1,743 annotated genes and 7,178 genome-wide background markers. Through rigorous applications across diverse cotton accessions, CottonSNP10K performed exceptionally technical robustness with call rates > 99% and genotype concordance rates > 99%. The array can effectively support precisely marker-assisted selection (MAS) for agronomic traits and high-resolution breeding population analysis, and markedly enhance GS predictive accuracy for agronomically important traits using prediction models that we established. This integrated approach provides a high-throughput precision tool for parent germplasm characterization and breeding line selection in cotton, enabling reliable identification of elite germplasm and advancement of genomics-assisted breeding.
Despite major improvements in molecular characterization of breast cancer, current biomarkers still fall short in accurate treatment prediction. Interrogating tumor tissue ex-vivo in its native conformation is a direct strategy for guiding treatment of individual patients but presents a challenge. In this study, we developed a microfluidic tissue array (μFTA) using small biopsy samples (< 1mm3) mimicking physiological flow for consistent exchange of nutrients and waste, retaining the tumor native stroma. Cell/patient-derived breast cancer xenograft tissues were maintained over 2 weeks in the array and their response to therapeutic agents, doxorubicin or neratinib, were interrogated. Drug response in the uFTA showed >2-fold reduction in tumor cell viability which corroborated tumor size shrinkage in mice bearing the same tumor load. EdU/Ki67 assays indicated selective retention of cells with higher proliferative capacity after drug treatment, underscoring in vivo clinical relevance. We have also developed a valved-μFTA to increase throughput and variety of treatment conditions on the same chip. Together, this μFTA can be staged as a powerful, low-cost benchtop theranostic tool for personalized cancer therapeutics compatible with FDA's New Approach Methods.
Edge artificial intelligence (AI) and embodied vision call for compact, fast, and energy-efficient hardware that integrates sensing, linear analog computation, and nonlinear activation, while flexibly reallocating these functions as workloads change. However, in-sensor computing (ISC) and in-memory computing (IMC) platforms still implement activation with external peripherals and use fixed functional partitions, which break the analog signal path and restrict system reconfigurability. Here, we report a reconfigurable ferroelectric transistor (Fe-FET) array in which polarization-programmed local fields enable junction-barrier engineering in ambipolar tungsten diselenide (WSe2) channel. This junction-barrier engineering mechanism co-programs photoresponsivity, multilevel conductance, and tunable nonlinear transport within the same device, allowing each Fe-FET cell to be reassigned among weighted sensing (ISC), linear accumulation (IMC), and hardware-native activation. The array therefore functions as a uniform pool of physical units whose roles and spatial partitions can be dynamically allocated to match task demands without changing the hardware platform. Using this role-reconfigurable platform, we implement an end-to-end analog neuromorphic vision system in which broadband sensing, linear computation, and nonlinear activation are executed natively on the same Fe-FET platform. These results establish a task-adaptive and energy-efficient route toward scalable neuromorphic vision hardware for edge intelligence.
This paper proposes the dual functional microwave sensor with high gain array antenna using loaded U-resonator and asymmetric T-junction. The microwave sensor is designed to operate at fr = 2.11 GHz while the array antenna operates at fr = 2.84 GHz which is integrated using asymmetric T-junction based on microstrip line. The performance of the sensor is observed based on the frequency shift of S21 for solid material characterization with a permittivity range of 1-9.8 while for the antenna is observed based on the parameters S11, bandwidth, gain and radiation pattern. From the measurement results, the antenna and sensor have high performance and have independent characteristics with high isolation of ≤ -40 dB to work concurrently. Moreover, the antenna has performance of S11 ≤ -10 dB, Fractional Bandwidth (FBW) 2.11% and gain of 6.03 dBi at fr = 2.84 GHz. The sensors operating at fr = 2.11 GHz has a performance of ∆F of 0.4 GHz, FDR of 0.061 GHz / ∆εr, normalized sensitivity (NS) of 2.21% and accuracy of 99.38%. Therefore, this work can be recommended for application in industries such as material quality control, pharmaceuticals and biomedical as a solution for real-time measurement processes integrated with wireless communications.
Ensuring food safety necessitates rapid, non-destructive, and reliable monitoring strategies, among which colorimetric sensor arrays (CSAs), based on analyte-induced color changes, have emerged as cost-effective tools for real-time food quality monitoring. Recent advances in sensing materials, fabrication strategies, and colorimetric mechanisms have expanded the analytical capability of CSAs, particularly through integration with machine learning. Automated feature extraction, pattern recognition, and predictive modeling significantly improve sensitivity, specificity, and classification accuracy, enabling more reliable food freshness assessment. Both traditional algorithms and deep learning models have been applied to CSA data analysis, supporting their implementation in portable, smartphone-based, and embedded sensing platforms. However, practical deployment is still limited by sensor degradation, environmental variability, response inconsistency, and the lack of standardized imaging and evaluation protocols. Future efforts should prioritize adaptive and lightweight algorithms, standardized data workflows, multimodal sensing systems, and sustainable, high-stability materials to enhance robustness, reproducibility, and industrial applicability.
Biological vision systems tightly couple spectral sensing with temporal integration to extract task-relevant information with minimal data movement. In contrast, conventional optoelectronic vision hardware typically separates photodetection from electronic computation, incurring substantial latency and energy costs. Optical neural networks can alleviate this bottleneck, but many implementations are difficult to scale. Here, we propose a bio-inspired optoelectronic inference architecture based on a Fabry-Perot microcavity-integrated MoS2 photodetector array, in which sensing, weighting, and accumulation are unified within each pixel. Cavity-engineered wavelength selectivity encodes neural-network weights in the spectral domain, while the finite carrier lifetime of MoS2 enables analog temporal accumulation without external memory. The system achieves test accuracies of 99.6%, 94.8% and 94.0% on MNIST, CIFAR-10 and the Free Spoken Digit Dataset, respectively. Post-training optical Hessian pruning further reduces optical complexity while maintaining robustness. This architecture provides a compact route toward wavelength-aware in-sensor neuromorphic inference.
Conventional lens-based microscopes are constrained by a trade-off between resolution and field-of-view (FOV), which limits overall imaging throughput. Recent works have shown that on-chip imaging systems with LED-array-based illumination offer a cost-effective approach for large FOV phase imaging. However, this strategy faces two main challenges: (1) twin-image ambiguity can degrade phase reconstruction. While mask-based modulation can help, it adds system complexity due to fabrication and alignment requirements; and (2) the illumination angle from each LED varies across large FOVs and can degrade centimeter-scale phase reconstruction without calibration. Here, we present a computational framework to jointly achieve mask-free on-chip phase imaging and adaptive calibration of spatially varying illumination angles. The sensorFOVis divided into subregions, within each of whichLEDillumination is approximated as planar. LED illumination angles for each subregion are initialized geometrically. Phase retrieval is then performed within each subregion by constraining the reconstruction with a soft optical transparency prior while simultaneously refining angle estimates. Reconstructed phase maps are merged to produce a high-quality, large-FOV phase image. We demonstrate this approach by achieving centimeter-scale on-chip phase imaging (up to 2.7 × 1.7 cm2) with micron-scale resolution across various biological tissue sections. This approach provides a simple, low-cost, and scalable solution for large-FOV and label-free imaging.
To build a clinically translatable neonatal seizure detection algorithm using amplitude-integrated electroencephalography (aEEG) and compressed spectral array (CSA). Using a public dataset of annotated neonatal EEGs, features of the aEEG and CSA were extracted from the left and right centroparietal electrodes. These features were then used to train and test three machine learning classifiers, Random Forest (RF), Support Vector Machines (SVM), and Artificial Neural Networks (ANN). The trained RF, SVM, and ANN classifiers had areas under the curve (AUC) of 0.80, 0.69, and 0.79 for capturing seizure time periods and an average accuracy of 0.91, 0.90, and 0.92 respectively for capturing seizure and non-seizure time periods. Median accuracy scores were higher among patients without hypoxic-ischemic encephalopathy (HIE; median = 1 for all three classifiers) than HIE patients (median = 0.92, 0.93, 0.93). A clinically interpretable aEEG-CSA algorithm is feasible for neonatal seizure detection by extracting standard EEG features and coupling these features with a supervised ML classifier.
Excessive production of reactive oxygen species during photodynamic therapy (PDT) can exacerbate inflammation at the infection site and increase the risk of tumor metastasis, thereby posing significant challenges to treatment efficacy. In this study, a chessboard-structured microneedle patch (HPC@PCN CMN) was designed using a micromolding technique. This design spatially separates and encapsulates photosensitizers and antioxidants within distinct fast- and sustained-release microneedle systems, respectively, thereby achieving spatiotemporally ordered control of PDT and anti-inflammatory treatment. Additionally, by adjusting the crosslinking degree of hyaluronic acid, the release rate of antioxidant drugs can be regulated, which facilitates a prolonged anti-inflammatory effect. The results indicate that the anti-inflammatory process of HPC@PCN CMN can alleviate inflammation by inhibiting the ATP-P2RX4 pathway, demonstrating promising efficacy in the treatment of acne and melanoma. Overall, the developed chessboard microneedle platform is programmable, enabling not only the sequential regulation of oxidation-antioxidation but also the potential for broader applications in drug-delivery scenarios.
Calix[4]pyrroles with unsubstituted meso-positions were synthesized by precipitating the products from reaction mixtures. The introduction of meso-spiroadamantyl subunits resulted in the scrambling of oligopyrrole intermediates during [2 + 2] condensation reactions. This process afforded a combination of cis- and trans-arranged isomers, as well as significant spectroscopic modifications arising from close contacts between the adjacent π-planes by the bulky adamantyl groups attached to the meso-positions in a spiro configuration.
Staphylococcus aureus is a major human commensal and pathogen, with the Panton-Valentine leukocidin (PVL) genes being associated with increased virulence. Rapid detection and molecular typing of such isolates are essential for effective epidemiological surveillance and infection control. Studies from Lithuania indicate that PVL is relatively common among S. aureus isolates, but detailed molecular typing data are still lacking. Clinical S. aureus isolates were collected from two hospitals in Vilnius in 2018-2019 and 2024, and from healthy volunteers between 2012 and 2020. Isolates were screened for PVL genes using real-time PCR. In addition, positive isolates harvested directly from the agar plate were tested for PVL production using an experimental lateral flow assay (LFA). Positive isolates were characterised using DNA-microarrays that facilitated the detection of resistance markers and virulence genes including PVL as well as an assignment to clonal complexes, strains and SCCmec types. Epidemiologically relevant isolates were subjected to whole-genome sequencing using Oxford nanopore technology. Out of 1296 S. aureus isolates, 124 yielded PVL-positive PCR results. 100 isolates were available for genotyping. Two PCR-positive isolates were negative by array and LFA, but PVL detection by DNA-microarray and lateral flow test (LF) showed complete concordance. Among PVL-positive isolates, 61.2% were methicillin-resistant S. aureus (MRSA). The most common PVL-MRSA strain (n = 43) was a Clonal Complex (CC) 8 MRSA that resembled the North American "USA300" strain but that lacked the arginine catabolic mobile element (ACME). This suggested an outbreak in one of the participating hospitals. Other common strains were PVL-positive CC30-MSSA, CC121-MSSA and CC8-MRSA-[IV+ACME] "USA300", while other lineages were represented by single isolates only. Whole-genome sequencing of two ACME-negative CC8-MRSA-IV isolates and of one local "USA300" isolate, as well as a comparison to international reference sequences, showed a very high degree of similarity in core genome and prophage content. The dominant PVL-MRSA strains were "USA300", indicating a possible importation from North America, and a locally emerged variant of "USA300" that lost the ACME-associated genes of its SCCmec element. These findings suggest that real-time PVL detection is essential for outbreak prevention. Given its clinical relevance, routine PVL screening -via PCR or lateral flow assays-in routine diagnostics should be seriously considered and surveillance of PVL-MRSA is urgently recommended.
Meteorites are classified as either non-carbonaceous or carbonaceous, representing bodies that are likely to have formed in the inner or outer Solar System, respectively. Despite its location in the inner Solar System, the Earth is thought to contain either minor (~6%) or substantial amounts (~40%) of outer Solar System material. However, because neither interpretation leverages variations among multiple isotopic systems simultaneously, Earth's provenance remains equivocal. Here we examine variations in ten nucleosynthetic isotope anomalies among planets and meteorite parent bodies to show that the linear extension of an array defined by non-carbonaceous bodies in any two isotopic anomalies always intersects the observed isotopic composition of the bulk silicate Earth to within 1 standard deviation. The Earth therefore formed exclusively from inner Solar System material whose composition did not vary over the course of accretion and was, on average, unlike that of any chondrite. Extension of the non-carbonaceous array yields isotopic compositions for Mercury and Venus that are more extreme than for Earth, implying a spatial or temporal gradient during the formation of the terrestrial planets.
Accurately monitoring carcinogenic volatile aromatic hydrocarbons (BTXs) is crucial for assessing air-qualities and danger-classes in specific occasions, However, it remains challenging to conduct highly selective identification of them in complex environments. Here, we have developed a gas-shunting strategy by installing function-reversal ZnO materials into Ir-WO3 supports to diminish interference-gas responses and guide special aromatic hydrocarbons sensing. We find that ZnO materials can serve as reactively sacrificial sites for small-molecule H2S and CO and induce main aromatic hydrocarbons reactants into Ir-WO3 supports. This gas-shunting route guarantees highly-selective aromatic hydrocarbons sensing even in dual/ternary gas mixtures. Through integrating functional-opposite sensors into a system, the final sensing arrays achieve 100% classification accuracy for 10 single gases and 75 multi-compose gases with low training costs. In addition, we also show an autonomic "cruise-detection" system by equipping sensor arrays into robotic dog to accurately identify complex gases. Our findings emphasize sensors designs with selective features and may broaden integrated sensing-system analysis in complex environment.
Glycan-mediated interactions are vital to development, microbial colonisation, immune signalling, and cancer progression. Glycan microarrays have revolutionised glycobiology by enabling high-throughput analysis of these complex interactions, supported by techniques that reveal kinetics and dynamics in solution or at the cellular level. We introduce multifunctional glycan probes based on a tri-functional Fmoc-Amino-Azido (FAA) linker, enabling multi-platform investigation of glycan-mediated interactions. These FAA probes support glycan presentation on both covalent and non-covalent array platforms, allowing direct comparison of glycan recognition by diverse proteins. Notably, certain viral adhesins and immune lectins show a preference for the non-covalent platform. The azido group allows further functionalisation via 'click chemistry', enabling biotinylation for immobilisation on bio-layer interferometry biosensors for influenza virus binding, or fluorescent tagging for flow cytometry analysis of glycan-lectin interactions on cells. These versatile probes offer a unified platform for in-depth interrogation of glycan interactions using complementary approaches, with strong potential to advance glycan-based diagnostics and therapeutics.
Long-term stability of sensor arrays remains a major limitation for the practical implementation of electronic nose systems in non-invasive analysis of volatile organic compounds (VOCs) and differentiation of oncological status. Changes in the physicochemical properties of sensing layers during prolonged operation may alter diagnostic parameters and compromise classification consistency. This study evaluated the long-term stability of diagnostically significant parameters obtained from skin gas-profile analysis using a quantum dot-based electronic nose. An eight-sensor piezoelectric electronic nose array based on cadmium sulfide quantum dots with various modifiers was used for skin gas-profile analysis of volunteers and cancer patients. Pair sensitivity parameters (Ai/j) and kinetic sorption parameters derived from chronofrequency responses were assessed by comparing measurements obtained during the pilot study and after two years of sensor operation. Pair sensitivity parameters demonstrated substantial susceptibility to long-term surface degradation and environmental variability, particularly humidity, resulting in pronounced classification drift. In contrast, kinetic sorption parameters exhibited higher temporal stability and lower sensitivity to operating conditions. A correction strategy based on systematic changes in kinetic parameters was developed and applied to compensate for sensor aging effects. The proposed approach reduced classification drift. These findings indicate that correction of long-term sensor drift may help extend the operational lifetime of quantum dot-based electronic nose systems, while kinetic parameters may provide more stable analytical indicators for gas-profile analysis.
Acoustic beamforming is widely used for source localization and line-of-bearing determination. Although many different beamforming techniques have been formulated, atmospheric turbulence effects on acoustic arrays are usually ignored in both their theoretical formulation and practical implementation. As a result, the performance of conventional beamformers, formulated for a non-turbulent atmosphere, degrades in the presence of wind velocity and temperature fluctuations, which cause fluctuations in the received signal amplitude and phase. This article presents a mathematical framework in which the amplitude and phase fluctuations are effectively suppressed from the signals allowing application of any beamforming technique and mitigating its performance degradation in a turbulent atmosphere. The framework is constrained to a single source and by the monochromatic plane wave approximation. Application of such an approach to an experiment revealed that the phase and amplitude fluctuations with spatial scales smaller than the array aperture are successfully suppressed, but the larger fluctuations (resulting in wavefront random tilt) remain and cause errors in the line of bearing estimates.
This study investigates the effect of key manufacturing parameters and graphene nanoparticle additions on the tensile behavior of fiber metal laminates (FMLs) using a Taguchi-based experimental design. Several manufacturing parameters were considered: laminate configuration (glass fiber and hybrid glass-carbon fiber reinforcement), aluminum surface treatment (chemical treatment and laser surface texturing with scanning spacings of 1 mm and 2 mm), aluminum thickness (0.5, 0.7, and 1.0 mm), graphene nanoparticle content (0, 0.1, and 0.25 wt%), and curing pressure (2, 5, and 7 bar). Eighteen FMLs specimens were fabricated according to the Taguchi orthogonal array and tested under tensile loading. Ultimate tensile strength ([Formula: see text]ult), tensile modulus (E), toughness modulus (UT), and failure strain ([Formula: see text]f) were evaluated as performance responses. The findings show laminate configuration significantly affects [Formula: see text]ult, UT, and [Formula: see text]f, with the all-glass fiber configuration exhibiting superior performance responses. Graphene content and curing pressure had a minimal effect on tensile properties. The optimal parameter combination for [Formula: see text]ult and UT involved a glass fiber laminate configuration with chemically treated aluminum, an aluminum thickness of 0.5 mm, 0% graphene, and a curing pressure of 2 bar. Optimal parameters for E include a laser 1 mm scanning texture, glass fiber laminate configuration, aluminum thickness of 0.5 mm, 0% graphene nanoparticles, and a curing pressure of 5 bar. Additionally, optimal parameters for [Formula: see text]f are glass fiber configuration, chemical surface treatment, aluminum thickness of 1 mm, 0% graphene, and curing pressure of 2 bar. Validation tests indicated the model's predictions were accurate, with prediction errors under 5%, highlighting its statistical reliability.
Optimizing water and fertilizer management is critical for improving fruit yield and quality in extreme arid regions, yet the synergistic effects of irrigation and NPK fertilization on fragrant pear (Pyrus sinkiangensis) remain inadequately characterized. This study aimed to identify the optimal water‑fertilizer strategy for fragrant pear cultivation in China's extreme arid regions through orthogonal experimental design. An orthogonal field experiment was conducted using an L9(33×21) mixed-level orthogonal array design. The experiment evaluated two irrigation levels (2400 and 3600 m3·ha-1) and three levels each of nitrogen (N: 150, 300, and 450 kg·ha-1), phosphorus (P₂O₅: 112.5, 225, and 337.5 kg·ha-1), and potassium (K₂O: 55, 110, and 165 kg·ha-1), resulting in nine treatment combinations. Measurements included fruit yield, quality parameters (soluble sugar, titratable acid, soluble solids, stone cell content, and firmness), cell tissue structure (pulp cell area and cellular arrangement), and anatomical features (stratum corneum, epidermis, and subepidermal thickness). Multivariate analysis was performed using principal component analysis (PCA). The yield of fragrant pear was highest under T4 treatment (N-P2O5-K2O: 450-337.5-165 kg·ha-1, irrigation volume 2400 m3·ha-1). T6 treatment (N-P2O5-K2O: 300-337.5-55 kg·ha-1, irrigation volume 2400 m3·ha-1) achieved the most consistent fruit shape development pattern, as evidenced by its strong correlation with the least square fitting curve (R2 = 0.97626), reflecting optimal growth synchronization. The soluble sugar content was significantly higher under treatment T4 than under other treatments (P < 0.05), while treatment T3 (N-P2O5-K2O: 150-225-165 kg·ha-1, irrigation volume 2400 m3·ha-1) had a relatively high titratable acid content (0.17%), soluble solids content (12.23%), stone cell content (1.16%), and fruit firmness (7.17 kgf). Histological analysis revealed significant differences in cellular morphology among treatments. T1 and T5 treatments showed the largest mean pulp cell section area (0.015 mm2), and microscopic examination revealed that T3 treatment had a more compact cellular arrangement and improved pulp texture characteristics. Anatomical analysis showed that T6 treatment had an optimal structural configuration, with the stratum corneum (9.5 µm), epidermis (45.7 µm), and subepidermal layer (50.4 µm) demonstrating a thickness ratio of 1:4.8:5.3. Based on multivariate analysis through principal component analysis, T3 treatment (N-P2O5-K2O: 150-225-165 kg·ha-1, irrigation volume 2400 m3·ha-1) demonstrated superior performance across multiple evaluation parameters, particularly cellular development attributes and integrated quality indices. T3 treatment (N-P2O5-K2O: 150-225-165 kg·ha-1, irrigation volume 2400 m3·ha-1) emerged as the optimal water‑fertilizer strategy for high‑quality fragrant pear production in China's extreme arid regions, balancing yield, fruit quality, and cellular structural integrity. These findings provide a practical cultivation guideline for arid‑zone pear orchards.
Yellow rust (YR) is a major threat to both bread and durum wheat production, often causing substantial yield losses. Conventional visual scoring of YR severity, while widely adopted, is labor-intensive, time-consuming, and prone to human error. In this study, we evaluated the predictability (PA), defined as the correlation between predicted and observed values, using genomic and phenomic data for YR severity under multiple prediction scenarios in two biparental wheat populations (bread and durum). YR scoring was conducted on two dates, with YR severity visually assessed while unmanned aerial vehicle (UAV)-based high-throughput phenotyping (HTP) data were collected using a multispectral camera. HTP data were processed to extract spectral wavelengths and vegetation indices (VIs), and all lines were also genotyped using SNP arrays. We tested a diverse set of models, including parametric, machine learning, and deep learning approaches. PA increased markedly when HTP-derived data were used compared with genomic markers alone. For example, support vector regression (SVR) improved from 0.35 (markers only) to 0.87 (wavelengths only). However, integrating genomic and phenomic data did not yield further improvements, as models often plateaued when using HTP-derived features alone. Cross-crop prediction demonstrated promising generalization across bread and durum wheat, achieving PA values up to 0.83. For this last task, best linear unbiased prediction (BLUP) and multilayer perception (MLP) consistently provided robust performance across scenarios. These findings highlight the strong potential of UAV-based HTP for rapid, scalable, and accurate prediction of YR severity in wheat. While genomics retains broad utility for breeding, the practical integration of phenomics and AI-driven prediction pipelines will ultimately depend on breeding program strategies, resources, and objectives.