Sweat glucose detected by wearable sensors can serve as a potential reference factor for diabetes management. However, inefficient sweat collection and fluctuating sweat pH hinder the precise detection of glucose. Herein, we designed a wearable sensing system comprising an ant-nest-structured sensor and a pH-responsive microchannel with multiple sweat inlets to continuously and accurately monitor the glucose level in sweat. Compared with the traditional planar electrode, this internally interconnected biomimetic structure provided a large number of accessible pores for contact with glucose molecules and electrolytes, remarkably decreasing the diffusion resistance. Furthermore, due to the formation of adjacent distributed cascade sensing units, the electron transmission distance from the catalytic center to the electrode surface was shortened, and the inoperative diffusion of H2O2 in the sensor was also weakened. The microchannel could convey the perspiration from the epidermis to the sensing zone within 250 s, while the coated pH indicator in the microchannel monitors the liquid's pH in a colorimetric zone. As a result, our wearable sensor integrated with a pH-responsive microchannel accurately tracked the change of glucose levels in sweat before and after a diet, displaying reliable results after calibration similar to those of commercial colorimetric kits. In addition, a longitudinal investigation of the relationship between blood and sweat glucose from healthy people and diabetic patients revealed a good correlation in glucose between the two. Wearable biosensors for real-time noninvasive analysis of glucose in sweat may significantly facilitate diabetes management at home.

The advancement of decentralized and real-time monitoring necessitates robust electrochemical sensors that can operate continuously in complex environments. However, transitioning these sensors from laboratory prototypes to reliable field devices remains challenging due to issues like signal drift, fouling, integration constraints, and insufficient selectivity under real-world conditions. This review explores emerging strategies designed to overcome the core challenges limiting the performance and longevity of electrochemical sensors in continuous monitoring scenarios. We first examine advanced materials and surface modification techniques, including permselective membranes, biorecognition elements, including molecularly imprinted polymers (MIPs), nanostructures, and ratiometric sensing, that enhance selectivity and minimize interference. Antifouling coatings, self-healing electrodes, and dynamic polymer systems are discussed as key developments to counteract degradation and improve structural durability. Machine learning and chemometric approaches are presented as powerful tools for signal deconvolution, drift compensation, and adaptive calibration, with a particular focus on their integration into sensing workflows. Finally, we address system-level challenges related to sample handling, highlighting innovations in microfluidics, passive and active sampling, and hybrid platforms incorporating biosensing and separation modules. Together, these multidisciplinary strategies form a roadmap to support the translation of electrochemical sensors into reliable tools for continuous, autonomous monitoring across environmental, biomedical, and industrial applications.

Inter- and intrasubject physiological variability makes it essential to track the real-time interplay between therapeutic medications and the human body for precision medicine in both daily life and special contexts such as spacefaring. This creates an urgent need for noninvasive wearable sensors capable of real-time, simultaneous monitoring of individualized pharmacokinetic (PK) and pharmacodynamic (PD) responses. Yet, their development remains largely unexplored and technically challenging. Here, we present an integrated multimodal wearable NanoTracker (P2NanoTrek) that enables skin-on, noninvasive, real-time monitoring of multiple PK and PD biomarkers following the administration of metoprolol, a widely prescribed arrhythmia medication. In human participants, P2NanoTrek demonstrated real-time measurement of sweat metoprolol (PK biomarker) and heart rate (PD biomarker) under controlled conditions relevant to drug absorption and elimination. Furthermore, we established its potential for personalized pharmacotherapy by correlating sweat and plasma metoprolol levels and evaluating heart rate responses during a standard dosing challenge.

Highly sensitive nucleic acid testing-assisted early disease detection is crucial for effective disease prevention and management, particularly when targeting low-abundance genetic materials in molecular diagnostics. This study describes CRATE (CRISPR/Cas controlled antisense oligonucleotide (ASO)-mediated translational signal enhancement), a novel ultrasensitive approach for nucleic acid detection by integrating Cas12a trans-cleavage, ASO-controlled gene expression, and cell-free signal protein amplification. This assay leverages the target-induced trans-cleavage of ASO-controlled gene expression for the amplification of signal proteins, with luminescent signal allowing for attomolar-level target DNA detection, as well as antigenic protein application enabling visual detection by lateral flow assay. The CRATE assay improves sensitivity using ASO-modified locked nucleic acid, achieving a 10-aM-level DNA detection. The proof of concept demonstrates 0.1 copies/μL detection of HPV genomic DNA from HPV-positive cancer cells as well as colorimetric lateral flow tests with ∼10 copies/μL sensitivity. The CRATE assay can detect the HBV target in plasma from HBV-positive patients with 100% sensitivity and specificity. With high specificity and accuracy, the CRATE assay retains the potential for detecting any nucleic acid of interest. By integration of precise CRISPR-based cleavage, ASO regulation, and efficient protein signal amplification, this approach provides a promising solution for highly selective and sensitive nucleic acid detection and potential applications in clinical diagnostics and point-of-care testing.

Incorporating magnetically responsive nanostructures into flexible elastomers enables the rapid construction of dynamically tunable photonic skins. However, existing systems fail to achieve pure-color-bending recognition due to monochromatic limitations. We present a chameleon-inspired trichromatic photonic skin (TPS) through the magnetic-field-guided assembly of Fe3O4@SiO2 nanoclusters within polyacrylamide hydrogels to miniaturized sizes. The TPS uniquely decouples mechanical stimuli (stretching, pressure, and bending) into distinct structural color pixels via independent peak shifting in the reflection spectra, of which the precision surpasses conventional single-color systems. Moreover, mask-assisted ultraviolet light polymerization further enables programmable pixelation for underwater sensing and encryptions. This multiparameter responsiveness allows real-time optical tracking of underwater motions, including shark locomotion and finger articulation. Our strategy establishes a single-component paradigm for creating intelligent multichannel optical interfaces with fine color resolutions.

In the present work, we report an electrochemical sensorbasedon rhenium (Re) nanoparticles embedded within a double-shelled ZnMn2O4 hollow microsphere (Re@ZnMn2O4) for ultrasensitive detection of epinephrine (EP) in biofluids.The Re@ZnMn2O4 material was synthesized viaa coprecipitation and annealing route, followed by a green, caffeicacid (CA)-assisted chemical reduction. Structural and morphologicalanalyses, including spectrophotometry, confirmed the high purity,crystallinity, and integrity of the material. Electrochemical performancewas evaluated using voltammetry and impedance spectroscopy. The Re@ZnMn2O4-modified electrode exhibited superior electrochemicalactivity attributed to its high conductivity, large surface area,abundant active sites, and efficient charge transfer enabled by thehollow architecture. EP oxidation followed a diffusion-controlled2e–/2H+ transfer mechanism. The sensordemonstrated a broad linear detection range (0.5–1951.3 μM),a low detection limit (0.21 μM), and good sensitivity (0.282μA μM–1 cm–2). Furthermore,it showed remarkable reproducibility, long-term stability, and strongresistance to common interferents. Its practical potential was validatedby accurate EP quantification in human serum and urine, highlightingits applicability in clinical diagnostics and biomedical monitoring.

Plant diseases threaten global food security, necessitatingefficient,field-deployable diagnostics. Traditional approaches, including morphologicalassessments and conventional sampling methods such as swabs or multistepDNA extraction, are often unreliable in the early stages of infectionwhen visible symptoms are absent and subsurface pathogens remain inaccessible,underscoring the need for molecular tools. Here, we present a microneedle(MN)-based diagnostic platform integrated with loop-mediated isothermalamplification (LAMP) and lateral flow assay (LFA), supported by acustom-built portable heater (LAMPbox). The MNs enable the directextraction of plant fluids and associated pathogens, offering a rapid,simple, and efficient sampling approach. In this study, two MN typeswere evaluated: poly­(vinyl alcohol) (PVA) and stereolithography-printedMNs fabricated from a plant-derived resin. The 3D-printed MNs exhibitedsuperior mechanical robustness, with a displacement ratio of only3.3% compared to 24.5% for PVA MNs, and provided an ∼12% higherDNA yield than conventional swab-based methods. Integrated with theLAMPbox, this platform enabled reliable detection of Puccinia triticina with clear differentiation betweenhealthy and infected leaves. This work establishes 3D-printed MNsas mechanically robust and effective tools for pathogen sampling whiledemonstrating the feasibility of a low-cost, portable, and sustainablesystem for early plant disease diagnosis.

To understand the gas-sensing mechanism of COFs and explore an effective modulation way to regulate their sensing properties, the adsorption and sensing behaviors of NO2, NO, SO2, O2, H2O, CO2, H2S, CO, N2, and NH3 gas molecules on the surface of pristine and n-doped (Na-adsorption) Tr-Th COFs are explored theoretically with first-principles calculations in this work. Attributed to the lowest unoccupied molecular orbital (LUMO) energy level of NO2, the adsorption of NO2 on Tr-Th leads to a larger increase in carrier concentration increment (n = 1.79 × 1012 to 6.26 × 1010 cm-2) and a greater work function shift (ΔΦ = 0.178 eV) compared to other gases, which suggest that Tr-Th is a highly promising material for NO2 sensing applications. n-Type doping elevates the Fermi level of COFs, resulting in a greater carrier concentration increment (n = 1.83 × 1012 cm-2 ∼ 2.52 × 1012 cm-2) and a larger work function shift (ΔΦ = 0.08 eV ∼ 0.30 eV) upon exposure to NO2, NO, SO2, or O2 compared to other gases. It means that apart from NO2, NO, SO2, and O2 gases will also trap electrons in n-doped Tr-Th COFs, increase the electrical resistance dramatically, and then quench the source leakage current of the COFs-FET gas sensor. Our study provides more detailed information about the gas-sensing mechanism of COFs and highlights the key role that surface doping strategy plays in regulating the gas adsorption and selection behaviors for its practical gas sensor applications.

The added flexibility in architecture and multidimensional applications make metal-organic frameworks (MOFs) and covalent organic frameworks (COFs) compelling candidates for the development of thin-film sensors. This review examines the on-surface synthesis of thin films of MOFs and COFs and focuses on advanced deposition methods, including chemical vapor deposition (CVD) and interface-assisted synthesis (IAS), as well as multifunctional sensing with MOF and COF thin films. Due to the high porosity, tunable structures, and excellent electronic/photoelectric properties, COF and MOF thin films have shown promising applications in gas sensing, environmental monitoring, and in wearable sensors. Additionally, the review discusses challenges related to long-term stability, film uniformity, and integration under real-world conditions. Some of the uniformity, longitudinal stability, and real-world conditions span deep time durability issues in practical use as well as long-term severe environment exposure alongside other prospective solutions for stilled gaps, which the review analyses. This overview aims to provide a roadmap for the development of next-generation sensors based on structurally robust and functional MOF/COF thin films.

High-sensitivity viral diagnostics typically use PCRto detectand amplify viral nucleic acids which requires fluorescence reporters,enzymatic amplification, specialized equipment and can be time-consuming.In this work, we describe fuel-free (FF) Rolosense, a diagnostic approachthat leverages mechanical force sensing as a fundamental transductionmechanism. We use the Brownian motion of aptamer-coated microparticleson an aptamer-modified surface for viral detection. The microparticlesfunction as both the sensing and transduction elements, reportingspecific molecular interactions where the presence of viral particlesstalls their motion by cross-linking them to the surface. FF-Rolosenseharnesses biased motion and thermal fluctuations to achieve rapid,sensitive, and specific detection of intact virionsthe activeagents of infection. This approach represents a fundamental shiftfrom conventional diagnostic methods and demonstrates a limit of detectionas low as 103 copies/mL for SARS-CoV-2 variants, includingBA.1 and BA.5, and effectively differentiates SARS-CoV-2 from otherviral pathogens such as Influenza A, HCoV OC43, and 229E. We alsoshow that FF-Rolosense readout is amenable to deep learning analysisrevealing single particle viral binding events. Finally, we demonstratepotential for point-of-care and home-based applications by using a3D-printed brightfield microscope, Roloscope, forFF-Rolosense readout. Taken together, this work shows a complementarystrategy for viral diagnostics that employs a mechanical mechanismof transduction.