Autonomous In Situ Biointerfacing Platform for Real-Time Electrophysiological Monitoring and Advanced Wound Management

Abstract

Flexible bioelectronic interfaces hold great promise for advancing modern healthcare and human–machine interactions. However, current bioelectronic interface technologies remain constrained by the intricate surface conditions of injured tissues. Even with intimate tissue-electrode adhesion, achieving simultaneous sensing and therapeutic intervention poses a formidable challenge. Here, we employed the principle of liquid-to-solid conversion to develop a seamless in situ forming biointerface platform, TLMG hydrogel, with robust and stable adhesion to irregular skin wounds, enhanced mechanical properties, real-time high-fidelity signal monitoring, and on-demand therapeutic effect for wound healing. By incorporating tea polyphenols/lignin microspheres, the TLMG hydrogel effectively achieved the integration of bioelectronic and bioactive interfaces. The multiple features of this in situ biointerface encompassed robust in situ adhesion (200 kPa), high ionic conductivity (0.27 mS cm^–1), and exceptional mechanical stability. Furthermore, the findings from several complex animal models and human tests proved the intelligent wound management, real-time dynamic signal monitoring, and wound healing boosting via immunomodulatory mechanisms. These results convincingly indicate that the TLMG biointerfacing platform provides a promising solution for integrated bioelectronic medicine for wearable healthcare systems.

Introduction

With the ongoing advancement in miniaturization and the development of flexible electronics, biomedical devices, such as brain-machine interfaces, smart bandages, therapeutic patches, and pacemakers, are increasingly pivotal in facilitating rapid disease diagnosis and comprehensive healthcare management. − Annually, about 50 million of these devices are implanted across the globe, significantly improving the management of both acute and chronic health problems. − The evolution of biointerface technology has opened new avenues for ensuring stable attachment and functional interaction between electronic devices and biological tissues, heralding promising applications in fields such as wound healing and implantable technologies. Notably, under conditions of injured tissue, the interface microenvironment differs markedly from that of healthy tissue. During the acute phase of tissue injury, the wound is irregular in shape, accompanied by substantial bleeding. Subsequently, the tissue undergoes a chronic healing process, an orchestrated sequence of biological phases: oxidative stress, inflammation, and tissue regeneration. External perturbation risks healing dysregulation, inducing delayed or nonhealing outcomes. − In this context, the seamless adhesion of bioelectronic interface materials to injured biological tissues is hindered by dynamic or complex curvatures, mechanical stresses, and the moist microenvironment induced by bleeding, leading to inferior adhesive stability, poor conductivity, or even detachment. , Unfortunately, current bioelectronic interface materials lack effective strategies that simultaneously enable electrophysiological signal monitoring and deliver therapeutic intervention, particularly under complex conditions in wound management.

Liquids can conform precisely to diverse surfaces owing to their intrinsic fluidity and low viscosity, which has inspired the advancement of in situ forming materials that begin as liquids and solidify on-site to occupy defect sites. − Gelatin-based materials, which undergo reversible thermally induced transitions between liquid and gel phases, offer a promising option for achieving seamless skin contact. , Nevertheless, gelatin-based in situ forming materials are often susceptible to mechanical stress or tissue dynamic movement, resulting in poor adhesion. Consequently, inadequate tissue anchoring fails to guarantee robust device operation in the complicated wound environment. − Currently, most bioelectrical interface materials still cannot meet the dual requirement of robust signal transduction and comprehensive management across all phases of wound healing, which specifically encompasses a seamless adhesive interface, enhanced mechanical properties, long-term stability, real-time signal monitoring, and on-demand therapeutic intervention.

To tackle these challenges, we applied the concept of liquid-to-solid transformation to develop a seamless in situ forming biointerfacing platform (referred to as TLMG hydrogel) with robust and stable adhesion, high-fidelity electrophysiological signal monitoring, and on-demand therapeutic effect for wound healing. The in situ TLMG hydrogel, formulated from tea polyphenols/lignin microspheres (TL NPs), MXene nanosheets, and gelatin, features a unique conformal adherence to irregular skin wounds and delivers bioactive functions. In this design, the liquid-to-solid conversion is governed by the thermoresponsive sol–gel transition of gelatin, while TL NPs act as multifunctional cross-linking hubs that modulate and stabilize the formation of a robust, tissue-conformal network.

The success of our TLMG hydrogel lies in the integrated bioelectronic and bioactive interfaces with the following features. First, the interaction between TL NPs and gelatin enhanced the adhesion of the in situ forming hydrogel interface. The gelatin underwent rapid, reversible liquid-gel phase transitions via a thermally triggered sol–gel mechanism, enabling immediate in situ formation of tissue-interfacing matrices. TL NPs further significantly improved the interfacial adhesion by interacting with gelatin units and tissue surfaces via hydrogen bonding, π–π stacking, and coordination, achieving robust adhesion (200 kPa) to allow conformal contact with irregular skin surfaces and even wounds. As for wounds, the TLMG hydrogel provided instant skin defect filling and constructed a physical hemostatic barrier, which was essential for hemostasis during the bleeding phase and for wound care during the later stages of wound healing. Second, TL NPs enhanced the electrical conductivity and the electrical stability of the flexible sensing interface. The aromatic ring structures of TL NPs bonded with the MXene layers, forming an enhanced interfacial conductive pattern, which enabled reliable, efficient, and high-fidelity signal transmission and acquisition (ionic conductivity, 0.27 mS cm^–1). In addition, the strong mechanical properties of TL NPs endowed the TLMG hydrogel with excellent damage endurance (puncturing, bending, and twisting) and structural stability (current retention, 84.6% over 50,000 folding cycles), allowing it to maintain a robust/stable and durable contact interface under complex or dynamic conditions, such as during skin stretching in motion or on uneven, curved skin surfaces in anatomically complex regions, thereby securing the stability of electrical conductivity. More importantly, TL NPs themselves provide potent biological functions and therapeutic effects, which can significantly promote wound healing. The cell experiments and animal experiments demonstrated that the integrated TLMG hydrogel enabled the sustained release of bioactive TL NPs, which in turn facilitated skin regeneration and accelerated wound closure by alleviating oxidative stress, modulating macrophage polarization, and enhancing angiogenesis. Across multiple complex animal models, the TLMG hydrogel demonstrated rapid hemostasis within 16.7 s (10-fold faster than untreated open wounds), achieved high-fidelity biosensing of the wound surface (SNR: 15 dB), and brought superior wound healing quality (complete epithelialization and abundant hair follicle neogenesis) and accelerated wound closure. Furthermore, it holds great application potential in artifact-resistant, long-term, real-time electrophysiological signal monitoring for exercise-related environments. Benefiting from the innovative design and integration strategies, the in situ TLMG biointerface platform perfectly bridges the gap between the realm of digital sensing platforms and dynamic wound scenarios, orchestrates bioelectronics, bioactive interfaces, and biosensors, and solves the seemingly contradictory demands of wound dressing and real-time electrophysiological monitoring, opening up new possibilities for the next generation of intelligent theranostic wearable electronics (Note S1 and Figure S1).

Results and Discussion

System Overview, Design Principle, and Configuration

Establishing reliable tissue-device interfaces demands synergistic optimization of material properties, mechanically adaptive architectures, and scalable fabrication approaches. The interface must concurrently ensure conformal adhesion to a morphologically complex wound while maintaining biophysical safety and minimizing delamination-induced signal artifacts. Accordingly, we proposed the development of an in situ gel biointerface platform integrated with a wireless system for real-time electromyography monitoring of the wound. The core biointerface (termed TLMG) was formulated from tea polyphenols/lignin microspheres (TL NPs), MXene nanosheets, and gelatin (Gel) (see Methods for preparation details). Specifically, gelatin was selected for its biocompatibility, water solubility, and thermoreversible phase transition mechanism, rendering it an ideal matrix for in situ interface formation. The successfully prepared MXene was incorporated to augment electrical conductivity, while TL NPs synergistically conferred mechanical robustness, tissue adhesion, and wound healing ability (Figures a and S2–S6). −

1.

Conceptual illustrations of the in situ gel biointerface platform. (a) Structure illustration of TL NPs and the TLMG biointerface. (b) Scheme of the rapid in situ process of the gel biointerface platform. (c) Overview of the TLMG gel platform, the wireless electrophysiology measurement system, and the application of the in situ biointerface platform for real-time EMG monitoring. (d) Demonstration of the in situ biointerface platform for the wound treatment mechanism. (e) Comparative configurations of the in situ (top) and ex situ (bottom) gel biointerface platform, quantifying motion artifact immunity in ambulatory sEMG biosensing.

Leveraging this in situ biointerface platform, the resultant electrode enabled EMG signal detection across various muscle groups. We presented a representative electrode array platform (12 pixels, 10 × 10 cm²), which comprised three functional layers from bottom to top: (1) an in situ TLMG biointerface loaded with TL NPs for immune regulation and tissue regeneration, (2) an ultrathin and stretchable screen-printed silver circuit layout (patterned using a mask template), and (3) a protective polyurethane encapsulation layer to prevent external contamination and ensure biological safety. Notably, the assembled device incorporating the in situ TLMG biointerface platform achieved data collection features that can be applied to real-time EMG monitoring on the wound interface and transmitted real-time data mapping wirelessly via Bluetooth to an external recorder (Figures b,c and S7–S10). The assembly protocol facilitated rapid deployment: tissue-conformal lamination of the TLMG flexible biointerface, multiplexed electrode interconnection via embedded Ag microtraces (width: 200 μm, sheet resistance: < 0.5 Ω/□), and modular electromechanical coupling to a signal processing unit.

Notably, the TLMG biointerface platform demonstrated dual diagnostic-therapeutic functionality in complex wound management, enabling real-time physiological monitoring while accelerating healing via macrophage polarization-guided regeneration. Figure d illustrates the working principles, where the in situ biointerface facilitated comprehensive management throughout the wound healing process. The released TL NPs promoted immediate hemostasis, establishing a sealed microenvironment conducive to skin regeneration. Additionally, these immunoregulatory nanoparticles efficiently scavenged reactive oxygen species (ROS), mitigating oxidative damage, and promoted macrophage reprogramming, thereby synergistically enhancing the wound closure rate, neovascularization density, and collagen remodeling. Meanwhile, achieving conformal adhesion of the TLMG biointerface to irregular skin topography was pivotal for stable EMG signal output under dynamic conditions. As shown in Figure e, the TLMG biointerface underwent rapid (<1 min) in situ gelation upon topical application, forming a skin-compliant adhesive layer that maintained conformal coupling during dynamic deformations. Unlike ex situ counterparts requiring preformed interfaces, this in situ procedure, facilitated by skin-mimetic microstructures enabling seamless biointegration via fluidic adaptation and interfacial compatibility, ensured mechanical compliance with damaged epidermal tissues while effectively mitigating motion-induced artifacts.

Comprehensive Characterization and Electrical Output Feature

To investigate the composition-dependent mechanical optimization of the TLMG biointerface platform, we synthesized TLMG with varying amounts of tea polyphenol/lignin nanoparticles (TL NPs) within a gelatin/MXene matrix (0, 25, 50, 75, 100, 125 mg, respectively). A facile one-pot ultrasonication-assisted self-assembly method was employed for TL NP thesis (Figure a). The TL NP formation was attributed to the abundant phenolic hydroxyl groups acting as active sites, bonding to functional groups within the tea polyphenols (primarily epicatechin gallate) under alkaline conditions (Figure S11). , Leveraging the TL NP-reinforced multidynamic cross-linking networks within the TLMG biointerface, the TLMG exhibited exceptional tolerance to extreme deformations, including stretching, twisting, and bending (Figure S12). Furthermore, the rapid thermoresponsive sol–gel transition (<1 min) enabled facile scalability of the TLMG biointerface, achieving uniform thickness (variation < 8% across 100 × 30 cm²) and overcoming fabrication challenges associated with geometrically adaptive bioelectronics (Figure b,c).

2.

Overall configuration and characterization. (a) Schematic diagram of the preparation of TL NPs. (b) Images of the scalable preparation of the in situ TLMG biointerface (100 × 30 cm²). (c) Photos showing the fast phase transition of the biointerface (30 °C, 1 min). (d) FTIR analysis of lignin, tea polyphenol, and TL NPs, respectively. (e) TEM images of lignin (left) and TL NPs (right), and an inset of the corresponding particle size distribution statistical chart. (f) NMR spectra of lignin and TL NPs. (g) Infrared thermal images reveal the phase transition under different heat conditions (30.3, 35.6, 39.6, and 45.5 °C). Scale bar, 1 cm. (h) Rheology test of in situ biointerface with varying amounts of TL NPs (0, 25, 50, 75, 100, 125 mg, respectively). (i) Photos demonstrating the puncture process of the biointerface. Scale bar, 1 cm. (j) Toughness and puncture force analysis of the biointerface with different TL NPs concentrations. (n = 3, biologically independent samples). (k) Conductivity versus temperature of the TLMG biointerface with different MXene concentrations (1, 3, 5, 8, and 10 mg mL^–1). (l) Comparison of resistivity pulses between TLMG biointerfaces with/without TL NPs even after 50,000 folding fatigue cycles (radius ≈ 2 cm, angle ≈ 180°). Error bars are defined as SD (n = 3 independent samples).

To underscore the underlying mechanism for the successful preparation of TL NPs, Fourier transform infrared spectroscopy confirmed that the peaks at 1114 and 1617 cm^–1 corresponded to C–O–C and C=C vibrational peaks in TL NPs, respectively (Figure d). The absorption band at 3430 cm^–1 was assigned to the O–H stretching vibrations. Compared to pristine lignin, absorption peaks at 2940 cm^–1 in TL NPs were attributed to the C–H stretching vibration. Similarly, TEM images revealed the formation of brown-red TL NPs with an average size of 53 ± 6 nm, significantly larger than the original lignin particles (31 ± 3 nm) (Figure e). The ¹H NMR analysis (Figure f) confirmed retention of methoxy (−OCH₃, δ ≈ 3.7 ppm) and phenolic hydroxyl (δ ≈ 3.3 ppm) signatures in both native lignin and TL NPs. Crucially, TL NPs exhibited pronounced signal enhancement at δ ≈ 2.3 ppm, a region with only trace intensity in pristine lignin, verifying the formation of covalent cross-linking between lignin and tea polyphenols. Collectively, the above analyses confirmed that the TL nanoparticle formation was governed by a self-assembly process co-driven by covalent cross-linking (primarily through C–O–C ether bonds and secondarily via C–C linkages) and noncovalent interactions (hydrogen bonding and π–π stacking). Thermal oxidation of catechol groups yielded quinone intermediates that covalently coupled with lignin, while noncovalent interactions coordinated structural organization and integrity. Notably, the incorporation of TL NPs (40 mg) significantly modulated the phase transition behavior for rapid in situ biointerface formation. Infrared thermal imaging confirmed a solid–liquid transition temperature of approximately 48 °C (Figure g and Note S2). As a further confirmation, rheological measurements revealed temperature-dependent viscoelastic changes (Figure h). Increasing the TL NPs content from 0 to 50 mg elevated the transition temperature from 32.3 to 52.8 °C. This thermotropic behavior stemmed from interfacial interactions between gelatin’s carbonyl/hydroxyl groups and surface hydroxyls on the TL NPs. These molecular associations can be used to regulate the gelatin chain reorganization during assembly. Indeed, increased TL NP content enhanced functional group coordination, stabilizing the network and perturbing intratriple-helix hydrogen bonding, thereby elevating the energy barrier for the solid–liquid transition.

Tensile tests and puncture resistance were also conducted on the TLMG biointerface with varied content of TL NPs (Figures i,j and S13–S15). As the amount of TL NPs increased from 0 to 125 mg, the Young’s modulus and puncture force of the TLMG biointerface enhanced, with values ranging from 30 to 152 kPa and from 0.5 to 1.3 N, respectively. This highlighted the TL NPs-induced reinforcement through dynamic-bond-mediated energy dissipation. Notably, the temperature-dependent conductivity was examined by electrochemical impedance spectroscopy (Figures k and S16), and the hydrogels achieved a positive correlation between conductivity and temperature, maintaining a conductivity of 0.29 mS cm^–1 even at 40 °C. Comparisons of conductivity across independently prepared batches revealed minimal fluctuation (20–60 Ω), highlighting the exceptional batch-to-batch consistency of the materials, thus meeting the required performance standards for practical applications (Figures S17). Additionally, thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) profiles collectively confirmed the environmental stability of the TLMG biointerface platform, supporting its potential for long-term and real-time monitoring applications (Figures S18 and S19). Equally intriguingly, to evaluate the durability, a homemade bending/flattening setup (angle ∼180° and radius ∼3 mm) was employed to monitor the current variation during prolonged cyclic deformation (Figure l and Video S1). Compared with the control biointerface that exhibited substantial current fluctuations (0.009–0.001 A) due to conduction failure under localized bending stress, the TLMG biointerface platform maintained robust electrical performance (current retention ≈ 87%) and showed no visible surface damage after 50,000 folding cycles. These results collectively demonstrated that TL NPs functioned as fatigue-barrier reinforcements and multifunctional nanofillers, facilitating efficient stress distribution.

In Situ biointerface Platform for Geometrical Synchronization and Wound Hemostasis

The TLMG biointerface platform revolutionizes tissue fixation by resolving dual limitations of conventional methods: eliminating suture-induced mechanical trauma via autonomous catechol-mediated adhesion, and achieving full interfacial compliance through dynamic topological entanglement. , To achieve this, mussel-inspired catechol chemistry mechanisms involving tea polyphenol/lignin nanoparticles (TL NPs) create multivalent bonding with polar substrates (e.g., metals, polymers, glass, and skin), enabling unprecedented noninvasive tissue-electronics integration. As exemplified in Figure a,b, the adhesion between the in situ and ex situ TLMG hydrogel interfaces with varying TL NPs concentrations (10, 20, 30, 40, and 50 mg mL^–1) was quantitatively evaluated via the lap-shear test on the aforementioned substrates. The introduction of catechol groups from TL NPs significantly enhanced the adhesive strength of the in situ interface on skin, increasing from 3 ± 2.5 to 21 ± 1.6 kPa. A similar trend of enhanced adhesion was observed across all substrates, surpassing both ex situ interfaces and most reported catechol-based hydrogel electrodes. , We examined the interfacial energy on various polar substrates via a standard 90-degree peel test (Figures c,d and S20). The in situ TLMG biointerface exhibited superior peeling energy on multiple polar substrates, including plastic (38 J m^–2), wood (670 J m^–2), pigskin (430 J m^–2), cloth (3785 J m^–2), rubber (2159 J m^–2), glass (463 J m^–2), and iron (738 J m^–2), far exceeding that of the ex-situ biointerface. We speculated that the superior in situ adhesion was related to both the rough porous and hydrophilic surfaces that facilitated the penetration of the hydrogel matrix and generated extensive physical interlocks. The strong and stable adhesion of TLMG hydrogels can be ascribed to the synergistic effect of the carboxyl groups of gelatin and catechol groups of TL NPs via multiple dynamic interactions, including hydrogen bonding, dipole–dipole interactions, electrostatic attraction, metal coordination, and anion/cation−π interactions, as illustrated in Figures S21 and S22. Notably, the maximum adhesion strength of ∼200 kPa was measured on dry, textured artificial skin and should be regarded as an upper bound under idealized conditions rather than that measured on wet wound beds. On saline-moistened porcine skin, the in situ TLMG biointerface still forms intimate initial contact via multidynamic interactions, but its adhesion gradually relaxes as hydration weakens interfacial bonds, which is beneficial for avoiding secondary tissue damage during dressing changes, as illustrated by the wet-state photographs (Figures S23 and S24).

3.

In situ biointerface platform for interface adhesion and hemostasis. Shear adhesion strength of (a) in situ and (b) ex situ TLMG biointerface (TL NPs concentration: 1–5 mg mL^–1), and the corresponding adhesion energy of (c) in situ and (d) ex situ TLMG biointerface on diverse substrates (cloth, wood, rubber, pig skin, glass, metal, and plastic, respectively). (e) Mechanistic schematic of the in situ biointerface for EMG signal acquisition. (f) 3D laser microscope images of in situ biointerface conformability on topographically complex substrates. Scale bar: 500 μm. (g) Modulus (left) and adhesion (right) AFM mapping of the in situ and ex situ TLMG biointerfaces. Scale bar: 1 μm. (h) Progressive interfacial delamination dynamics of the in situ biointerface from the skin. (i) Shear adhesion strength of the in situ biointerface on different substrates up to 20 successive peeling-off cycle tests. (j) Schematic illustration of the mouse liver hemostasis model. (k) Images of the in situ biointerface adhered to representative tissues, possessing superior in situ interfacial coupling for wound hemostasis. (l) Quantitation of blood loss and hemostasis time in the liver injury model (mean ± SD; *P < 0.05, ^** P < 0.01, ^*** P < 0.001).

Such a dynamic bonding mechanism offered distinct advantages over chemical adhesives based on acrylic polymers or cyanoacrylate derivatives. , Although these synthetic adhesives initially exhibit strong bonding in controlled settings, their clinical applicability remains constrained by (1) gradual release of cytotoxic monomers that impair tissue viability and (2) rigid interfacial bonding that restricts natural tissue deformation. As depicted in the schematic (Figure e), the self-adhesive and compliant in situ interface was formed to achieve the compliance of curved skin, thereby facilitating the capture of physiological signals. For another visual observation, the in situ TLMG biointerface could remain mechanically compliant to the dynamic and wet irregular skin to withstand deformation (0–50% stretching), such as at the elbow, wrist, and interphalangeal joints, without retraction or delamination (Figures S25 and S26 and Video S2). In addition, wet-state tests on saline-moistened porcine skin revealed that the in situ gel maintained intimate contact without edge detachment. Similarly, for the moist human joints, the in situ gel maintained the interfacial adhesion continuously, which proved the excellent wet interface adhesion property of the gel (Figures S23 and S24).

Given the significant variation in tissue geometry across body locations, we employed a complex-shaped 3D substrate possessing nondevelopable Gaussian curvature to further validate conformability on surfaces with diverse curvatures. As depicted in the laser microscope images (Figures f and S27), the in situ TLMG biointerface was entirely adaptive to the protrusions and grooves, exhibiting almost the same patterns as the original nondevelopable periodic curvilinear geometries. Moreover, the in situ interfacial roughness (S _a, average roughness) exhibited a consistent increase from 21 to 54 with increasing substrate geometric complexity, inferring that the in situ biointerface could fully follow the microscale contours of the underlying 3D substrates for seamless interfacial connection. Besides, as proven by modulus and adhesion atomic force microscope (AFM) mapping (Figure g), a robust (35 MPa) and strongly adhesive (3 nN) interface of TLMG biointerface was achieved, in sharp contrast to the biointerface without TL NPs, which was reminiscent of a robust adhesion mediated through a balanced contribution of adhesion and cohesion.

To better demystify the underlying mechanisms for seamless and interlocked interfaces, Figure h delineated the energy-dissipation governing interfacial fracture via the sequential optical images during in situ peeling. Nanofibrillar bridging networks spontaneously formed at the initial interface separation, establishing a nonlinear crack propagation threshold to prevent catastrophic delamination. Controlled sequential rupture of individual fibrils replaced bulk fracture modes to avoid continuous, catastrophic, and brittle interface failure, achieving residue-free detachment while maintaining interfacial integrity for reusability. This fracture-resistant yet compliant adhesion paradigm established an optimal balance between interfacial integrity and biological conservation, which is particularly critical for dynamic or damaged skin requiring both mechanical stabilization and physiological responsiveness. In addition, 20 successive cycles (Figure i) and 5 days of storage (Figure S28) of the peeling-off tests were conducted to evaluate the durability and stability of the in situ TLMG hydrogel interfaces. The TLMG hydrogels maintained superior adhesion even after repeated peeling/readhesion cycles (above 15 kPa) and exhibited stable adhesion (10.5 kPa) even after 5 days of storage, implying the outstanding adhesion durability. Moreover, the in situ TLMG hydrogel biointerface platform could be recycled through repeated sol–gel phase transitions, and the adhesion exhibited only a subtle decline over 20 recycling cycles, providing competitive features over reported adhesive hydrogels that cannot be reactivated postcuring (Figure S29).

Notably, the adhesive in situ biointerface facilitates effective trauma care: its rapid hemostatic capacity mitigates life-threatening hemorrhage in acute visceral injuries, while concurrently establishing an infection-resistant regenerative microenvironment for accelerated wound closure. An SD rat liver perforation wound model was used to evaluate the hemostatic potential (Figure j). After being punched, the in situ gel was directly attached to the bleeding site and immediately formed a physical hemostatic barrier, while no intervention was applied in the control group. As shown in Figure k and Video S3, filter paper underneath the liver showed only a small area of bloodstain in the in situ group. Conversely, a clear bleeding pathway was observed in the control group. Statistical analysis of the hemostatic time further verified that the in situ gel group effectively stopped the liver bleeding within ∼16.67 s, which required considerably less time than that of the control group (∼158.33 s). Regarding total blood loss, the in situ gel group exhibited significantly lower blood loss (∼0.08 g) than the control group (∼1.06 g) (Figure l). The above results underscored that rapid in situ gelation and wet bioadhesion, facilitated by catechol groups and hydrogen bonding, enabled instant defect filling and constructed a durable physical hemostatic barrier, which remarkably reduced blood loss by 92.5% and shortened hemostasis time by 89.5% in open wounds. The establishment of an ultrananotopographic biointerface platform through defect-free conformal contact fundamentally addressed the chronic signal attenuation problem in biopotential real-time monitoring. By achieving submicron-level topological interlock (surface roughness <0.5 μm) with tissue microstructures, this precision contact mechanics surpasses conventional electrode–tissue interfaces where air gaps and surface irregularities typically cause 40–60% signal loss, thereby creating the essential electromechanical foundation for capturing millisecond-scale action potential propagation with clinical-grade temporal resolution.

Biocompatibility and Antioxidant

Excellent biocompatibility is a prerequisite for bioelectronic interfaces used in wound healing and biological signal sensing. − The in vitro cytocompatibility and hemocompatibility of the in situ TLMG hydrogel biointerface were evaluated using mouse fibroblast cells (L929 cells). The CCK-8 assay demonstrated that a TLMG gel maximum concentration of up to 5 mg mL^–1 was the safety-critical TLMG gel concentration, and the cell viability remained at ∼96.5% (Figure S30). Once the concentration exceeded this threshold, the cell viability decreased significantly. In addition, live/dead staining was used to observe the cell biocompatibility visually. Both the control group and the MG gel, TL NPs, and TLMG gel groups presented normal L929 cell morphology with no noticeable red fluorescent cells (dead cells) (Figures a and S31, and S32). Cells treated with 5 mg mL^–1 of TLMG gel showed a similar proliferation trend to the control group (Figure c), confirming favorable cytocompatibility in vitro. Furthermore, the TLMG gel did not induce hemolysis, making it suitable for application in both severe bleeding wounds and newly occurring wounds (Figure b). We further examined the in vivo biocompatibility and biodegradability by performing subcutaneous injections of TLMG in healthy mice. After 4 weeks postinjection, the injected TLMG gel specimen and adjacent tissues were collected for histological analysis. Compared with the control group, no obvious inflammatory infiltrate was detected in the TLMG group. Notably, the injected TLMG was barely observed after 4 weeks. The epidermis was entirely regenerated, and the sebaceous glands were preserved intact and uninjured, exhibiting a structure comparable to that of healthy skin tissue (Figure d). Moreover, histological assessment of major organs indicated that there was a negligible effect on cell morphology and overall structural integrity of the heart, liver, spleen, lungs, and kidneys (Figure S33). No significant alterations in the general health or behavior of the mice were noted during the treatment period and subsequent observations (Video S4). These results collectively confirmed that the proposed in situ biointerface platform demonstrated suitable biocompatibility for bioelectronic dressings at wound sites.

4.

Biocompatibility and antioxidant efficiency of the in situ TLMG biointerface platform. (a) Live/dead staining of L929 cells after being treated with the TLMG gel for 1, 3, and 5 days, respectively. Scale bar: 200 μm. (b) Hemolysis rate of TLMG gel and the supernatant imaging of RBCs in different treatment groups. (c) Cell proliferation rate of L929 cells assessed by CCK-8. (d) Histological analysis of skin and subcutaneous tissue 4 weeks after TLMG gel injection, stained with H&E. Scale bar, 100 μm. (e) Free radical elimination via TL NPs and the TLMG gel in different concentrations. (f) Establishment of the H₂O₂-induced oxidant damage model in L929 cells. (g) In vitro H₂O₂ scavenging ability of TL NPs and the TLMG gel. (h) Representative fluorescence images of intracellular ROS scavenging capability using the DCFH-DA probe. Scale bar, 100 μm. (i) Quantitative analysis of remaining intracellular ROS of Rosup-stimulated L929 cells using the DCFH-DA probe. Data are shown as mean ± SD *P < 0.05, ^** P < 0.01, ^*** P < 0.001.

Accumulating excessive ROS at wound sites can induce severe oxidative stress damage and impede wound healing. − Tea polyphenols are considered effective agents that neutralize free radicals and nonradicals, directly combating ROS. To investigate the ROS scavenging capacity of the TLMG bioelectronic interface, the DPPH free radicals scavenging test and intracellular ROS scavenging test were conducted. As shown in Figure e, TL NPs and TLMG efficiently eliminated DPPH and exhibited a concentration-dependent effect. With the concentration of TL NPs increased from 5 to 50 μg mL^–1, the average DPPH scavenging ratio increased from ∼27.41 to ∼74.18%, and then decreased to ∼69.02% with a further increase in the TL NPs concentration to 100 μg mL^–1. Likewise, as the concentration of TLMG increased from 0.5 to 10 mg mL^–1, the average DPPH scavenging ratio rose from ∼34.06 to ∼72.10%. Specifically, TLMG displayed a comparable DPPH scavenging ability (>70%) to the equivalent amount of tea polyphenols in TL NPs once the hydrogel concentration exceeded 5 mg mL^–1, indicating that hydrogelation did not diminish the ROS scavenging capacity. Considering the effects of concentration on biocompatibility and free radical scavenging properties, the optimum TLMG intervention in subsequent experiments was determined as 5 mg mL^–1. Furthermore, the scavenging ability of intracellular ROS by the TLMG biointerface was also investigated. H₂O₂ was used to establish the oxidant damage model in L929 cells. With an increase in H₂O₂ concentration from 1 to 1.4 mM, the cell survival rate gradually decreased from 90.82% to 36.56%. Notably, an H₂O₂ concentration of 1.3 mM led to a 55.3% survival rate of L929 cells (Figure f). At this point, cells were susceptible to oxidative damage but not to massive cell death. The CCK-8 assay was also performed to verify the protective effect on cells against H₂O₂. The cell survival rate (Figure g) decreased to nearly 50% after the addition of 1.3 mM H₂O₂. In contrast, the survival rate significantly increased to approximately 100% after coincubation with TLMG, indicating that TLMG could fundamentally protect cells from H₂O₂ damage. Furthermore, the intracellular ROS scavenging capacity of the TLMG biointerface was investigated by fluorescent probe DCFH-DA, which can be oxidized by intracellular ROS and result in DCF with green fluorescence. Rosup served as the positive control to produce excessive intracellular ROS. As shown in Figure h, intense green fluorescence signals could be observed in L929 cells after being treated with Rosup, indicating elevated levels of intracellular ROS. The green fluorescence signals in Rosup-treated cells were obviously quenched after incubating with TL NPs and TLMG, suggesting that the TLMG biointerface could effectively eliminate the intracellular ROS due to the TL NPs component in the hydrogel (Figure h,i). These results jointly indicated that the TLMG biointerface platform could scavenge intracellular and extracellular free radicals, protecting cells from damage caused by an oxidative environment, and could be an effective option for a wound dressing to address the issue of excessive ROS in chronic wounds.

Validation in Preclinical Wound Models and Wound Healing Mechanisms

Wound healing progresses through spatiotemporally programmed phases, including hemostasis, inflammation, and remodeling, demanding phase-specific bioelectronic interfaces. Our TLMG biointerface platform orchestrated rapid hermetic sealing for coagulation microenvironment control, ROS scavenging synergized with inflammation regulation, and angiogenesis-epithelialization-coupled skin remodeling. As part of the preclinical efficacy validation, a mouse full-thickness skin defect model was utilized to investigate the highly integrated functionality and timely therapeutic delivery of the in situ TLMG bioelectronic interface throughout all phases of wound healing.

Here, wound defects were locally managed with in situ TLMG gel, TL NPs, MG gel, and commercially available Tegaderm films, which were used as the control group (Figure a). After 3 days postsurgery, noticeable dry and formed scabs were observed in the TL NPs group and TLMG biointerface group, whereas the control group still showed moist open wounds (Figure b). Real-time tracking of the dynamic healing process indicated that the TLMG biointerface displayed superior wound healing progress, with minimal scarring (Figure c). Quantitative analysis further corroborated that, 10 days after management, the wound closure rate of 93.2% in the TLMG biointerface group surpassed that of the MG gel group (72.58%) and the control group (72.91%) (Figure d). In addition, the hematoxylin and eosin (H&E) staining in the repaired region enabled the evaluation of wound healing quality (Figures e and S34). On day 14, complete epithelialization and abundant hair follicle neogenesis were observed in the regenerated skin tissue of the TLMG-treated group, whereas the other groups, particularly the control group, still showed scabs and incomplete skin appendage regeneration. Furthermore, the TLMG biointerface group exhibited a narrower granulation tissue width (∼1.16 mm), further confirming its ability to expedite wound closure (Figure f). These findings substantiate that the in situ TLMG biointerface therapy promotes superior wound healing quality and accelerates the wound healing process compared to that of commercial Tegaderm films.

5.

In situ TLMG biointerface platform promoted rapid and scarless wound healing in all phases of dorsal wounds. (a) Schematic illustration of the animal study design (created in BioRender. Yu, W. (2026) https://BioRender.com/jojh7go). (b) Representative optical images of wounds taken on different days for the 3 M Tegaderm-only (control) and experimental groups treated with MG gel, TL NPs, and in situ TLMG gel, respectively. Before imaging and measurements, the adhered hydrogels were removed. Scale bar: 2 mm. (c) Wound area tracing analysis related to the optical images. (d) Quantification of relative wound area on days 0, 3, 7, 10, and 14, respectively. (e) Representative histologic images from day 14 wounds with H&E staining. Left: scale bar, 500 μm. Black dashed boxes denote the area for zoomed-in views in the right panels, highlighting healed tissue. Right: scale bar, 200 μm. (f) Quantification of relative granulation tissue width on days 7 and 14. (g) DHE staining images of wounds on day 3. Scale bar, 100 μm. (h) Immunofluorescence detection of F4/80 (red) and CD86 (green) in wounds on day 7. Scale bar: 50 μm. (i) Immunofluorescence detection of CD31 (red) and α-SMA (green) in wounds on day 14. Scale bar, 50 μm. (j–l) Statistical analysis of DHE staining (j), the percentage of M1 macrophages (k), and the relative expression of CD31 (l). Data are shown as mean ± SD *P < 0.05, ^** P < 0.01, ^*** P < 0.001.

To further elucidate the mechanism underlying the enhancement of wound healing by the in situ TLMG bioelectronic interface, immunofluorescent staining was conducted. Excessive ROS is associated with an inflammatory environment and reduced angiogenesis, consequently impeding wound healing. On day 3, the DHE staining results showed that the ROS level in the wound area was substantially decreased in both TL NPs-treated and TLMG-treated groups compared to the control group, indicating an enhancement of the antioxidant effect within the wound site due to the crucial component tea polyphenols. Notably, as a consequence of the sustained release of TL NPs, the TLMG gel interface exhibited the highest oxidative stress-attenuating efficiency, even higher than that of TL NPs alone (Figure g,j). Furthermore, to assess the immunomodulatory effect of the in situ TLMG bioelectronic interface, the polarization of macrophages in the regenerated wound area was examined. , F4/80 (a marker for total macrophages) was expressed in all groups. Specifically, the applied TLMG and TL NPs induced particularly low levels of CD86 (a marker for the pro-inflammatory M1 macrophage phenotype), 2.02% and 4.93%, respectively, whereas the MG gel group and the control group had levels of 25.79% and 28.96%, respectively (Figure h,k). This contrast indicated the inherent inflammation-suppressive capacity of the TLMG gel bioelectronic interface. In addition, the TLMG-treated wounds exhibited enhanced microvascularization, as evidenced by the elevated density of CD31⁺ vessels and a higher level of α-smooth muscle actin (α-SMA) on day 14 compared to the other groups (Figure i,l). Overall, a comprehensive analysis of the morphology, structure, and mechanism of the healing tissue substantiated the superior performance of the in situ TLMG biointerface platform in wound healing, skin remodeling, and the acceleration of skin regeneration. These findings jointly highlighted the promising clinical applications of the in situ TLMG gel interface within the domain of bioelectronic wound dressings and regenerative medicine.

Validation of Continuous and High-Fidelity Physiological Signal Detection

The high-quality electrophysiological signals obtained in the context of dynamic exercise can be used to prevent injuries and efficiently manage health by assessing muscle fatigue, thus holding promising applications in clinical diagnosis, rehabilitation, and sports training. − Herein, we engineered an adhesive, conductive MXene-based hydrogel biointerface with tea polyphenol/lignin nanoenhancers, which established dynamically adaptive tissue-sensor coupling for precision acquisition of bioelectrical signals during physiological motion (Figure a). To validate functional reliability, the in situ gel electrodes achieved seamless integration with topologically complex wounds in rat models, capturing synchronized electrophysiological signatures during sciatic nerve stimulation (2 ms biphasic pulses, 300–800 μA). This biointerface demonstrated millisecond-precision signal acquisition from dynamic wound surfaces while maintaining electromechanical synchronization with underlying neuromuscular activity (Figures b, S35, and Video S5). Furthermore, the characteristic waveforms were successfully captured with a high signal-to-noise ratio (SNR) of 13.8 ± 1.6 dB, following similar intensity variation trends in the SNR with changes in the stimulation current. Close inspection of the waveforms and the increased SNR tendency revealed a consistent pattern without obvious fluctuations, indicating efficient signal acquisition from the in situ biointerface platform on the damaged skin interface (Figure c). To demonstrate clinical-grade fidelity, the in situ electrodes captured high-definition EMG signals from thigh muscles during kinetic escalation protocols (running → jumping → sitting → hunkering), with concurrent skin tension peaking at 39.6% (Figures d and S36). Spectrotemporal analysis revealed severe motion artifacts and baseline drift in the ex situ counterparts, while the in situ biointerface maintained electrophysiological integrity under maximal biomechanical stress, validating the fatigue-resistant superiority of our in situ biointerface platform.

6.

On-body evaluation of the in situ customizable electrode patch under daily exercise movements. (a) Schematic illustrations of the in situ biointerface platform seamlessly adhering to the skin tissue. The robust skin-hydrogel interface enabled stable bioelectrical signal recording between hydrogel bioelectronics and biological tissues. (b) EMG signals were recorded using the in situ gel electrode on the gastrocnemius medialis (GM) muscle upon sciatic nerve stimulation. (c) The high signal-to-noise ratio (SNR) indicates that the in situ gel could stably record the surface EMG signal. (d) Time-frequency spectrograms and corresponding EMG signals from the in situ electrode (left) and ex situ gel electrode (right) recorded during running. (e) Photographic images of the electrode array attached to a large-area skin surface, in combination with the connection to an external circuit via a flexible flat cable. (f) The consistency of multichannel signal detection using the in situ gel, in contrast to the signal fluctuations on a nonconformal skin surface. (g) Photographs of the EMG measurement system attached to the vastus medialis muscle during a basketball game with strain interference (maximum strain, 39.6%). (h) Long-term EMG (top) and ECG signals (bottom) comparisons in the early, middle, and late stages of exercise (16 min), and the insets show the zoomed-in data segments, and (i) the corresponding respiration rate and pulse measurement in per minute.

Furthermore, based on the proposed TLMG in situ gel biointerface platform, we designed various strip- and sheet-patterned electrodes that were conformable to different anatomical sites for on-skin real-time physiological monitoring. These electrodes could be seamlessly applied to curved regions, such as the back and thigh, maintaining intimate contact with the skin for stable signal acquisition (Figure e). The instantaneous sEMG maps were generated to visualize channel-specific signal amplitudes, revealing distinct and reliable intensity distributions and propagation dynamics of the in situ electrodes (Figures f and S37). We further conducted in situ EMG signal measurements during continuous exercise (16 min) using a wireless measurement system consisting of in situ gel electrodes, a PCB connector, and a portable wireless module, as illustrated in Figure g. The electrodes were affixed to the vastus medialis of the quadriceps, and the specific hardware parameters for real-time EMG signal collection were presented in Figures S38–S40.

Notably, the in situ electrodes maintained robust and conformal contact with the skin throughout the entire movement process, thereby ensuring clear and stable acquisition of both EMG and ECG signals, without introducing motion artifacts into EMG recordings or compromising the accuracy of dynamic ECG evaluation (Figure h,i). Moreover, the signal baseline ranged from 30.71 μV at the beginning to 60.54 μV at the end of the exercise, and the EMG signals in different motion states were clearly distinguished by their intensity and amplitude. This result indicated that the electrode remained intact after the exercise and no motion artifacts were generated during the workout. We further analyzed the beats per minute (bpm) during different activities, such as jogging (73.7 bpm), running (64.6 bpm), and resting (59.8 bpm), along with a gradual increase in respiration rate (from 15 to 40 breaths per minute) and pulse (from 72 to 93 beats per minute), and the SNR of the EMG signal during the early stage (jogging), midstate (running), and late state (resting) remained consistent throughout the entire duration. The above signal acquisition revealed that the assembled electrode maintained its interface integrity even under pressure loading conditions, successfully recording signals for the entire duration. Although harsh mechanical stress was applied to the electrode, we confirmed that the in situ electrode maintained conformal contact in both high-load and high-repetition exercise scenarios, minimizing motion artifacts and preserving its performance without damage, thus enabling successful long-term continuous signal measurements.

Study Limitations

This work has several limitations that should be acknowledged.

First, the reported adhesion strength of about 200 kPa was obtained on a dry textured substrate, representing an upper bound under idealized conditions. In hydrated environments, the hydrogel still forms conformal contact, but the interfacial strength gradually decreases with prolonged exposure to fluid.

Second, the biological complexity of real wound microenvironments is partially captured in our in vitro assays. Factors such as local pH, ionic composition, and bacterial burden were simplified, and their combined impact on gelation, degradation, and signal stability remains to be quantified in more realistic settings.

Third, although the TLMG system provides stable electrophysiological recordings during motion compared to PEDOT:PSS-based hydrogels, comparison with commercial Ag/AgCl electrodes is constrained by differences in geometry, gel formulation, and contact area. A strictly matched benchmark was not feasible, and comparisons rely on literature ranges and internal controls.

Fourth, batch-to-batch reproducibility showed modest variation in conductivity, adhesion, and mechanical properties. However, variability in bioderived precursors may lead to performance fluctuations that require further evaluation in scaled manufacturing.

Finally, while TLMG was validated in rodent models, translating it to clinical practice requires systematic studies in larger animal models and clinical trials to establish safety, efficacy, and usability in real wound care workflows. Unlike MXene-based hydrogels, which focus on antimicrobial properties, TLMG integrates multimodal therapies and high-fidelity EMG/ECG sensing, offering a broader functional range and greater mechanical robustness, making it a strong candidate for clinical use (Table S1).

Conclusions

In summary, this work presents a fully integrated, wireless smart wound dressing system designed for high-fidelity electrophysiological signal monitoring, as well as a parallel comprehensive management of dynamic wound healing. While many previous studies have explored individual aspects, such as in situ formed electronic interfaces or hydrogels for wound healing, our key innovation lies in the successful integration of wireless sensors and wound management into a single biointerface platform. Major achievements include: (1) active and stable real-time monitoring of a dynamic wound and (2) acceleration and regulation of the entire orchestrated sequence of wound healing phases, ranging from hemostasis and oxidative stress to inflammation and tissue regeneration. Notably, the high-quality wound healing achieved is attributed to complete epithelialization, abundant regeneration of skin appendages, and immunomodulatory mechanisms that enhance cellular proliferation and recruit pro-regenerative phenotypes. This dual-mode integration propels the field of signal monitoring under complex wound physiological challenges, offering an optimized and accessible solution that not only broadens patient indications but also elevates the standard of care.

Despite its promising capabilities, the current biointerface platform can be further improved. Expanding studies to large-animal models, followed by human trials, will provide more compelling validation of our preliminary findings, particularly for predicting electrophysiological signals during wound healing progression and their role in guiding treatment strategies. Future investigations will also focus on the integration of machine learning and multimodal sensors, such as those monitoring blood oxygen levels, temperature, and humidity, which offer significant potential for advancing the next generation of personalized closed-loop bioelectronic medicine.

Materials and Methods

Materials and Chemicals

Ti₃AlC₂ (Max, 400 mesh) powder, lithium fluoride (LiF, 99.99 wt%), Silver (Ag) pastes, hydrochloric acid (HCl), glycerin (Gly), tea polyphenols (TP), lignin, and gelatin were purchased from Beijing Chemical Reagents Company. Tris­(hydroxymethyl)­aminomethane (Tris) was purchased from Sigma-Aldrich. The ultrapure water used was deionized and ultrafiltered through an ELGA 136 Lab Water system with a resistivity of 18.2 MΩ cm (VWS Ltd., High Wycombe, France). All reagents were of analytical grade and used without further purification.

Cells and Animals

A mouse fibroblast cell line (L929) was obtained from the Cell Bank of the China Academy of Sciences and cultured according to the manufacturer’s guidelines. All animal procedures were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of Peking University Health Science Center (Approval No. BCAJ0289).

Fabrication of MXene Nanosheet

MXene nanosheet aqueous dispersion was obtained by a minimally intensive layer delamination (MILD) method. First, LiF (2 g) and HCl (40 mL, 9 M) were magnetically stirred (300 rpm for 15 min). Then, Ti₃AlC₂ (2 g) was slowly added and continuously stirred (35 °C, 300 rpm) in the above solution in a high-density polyethylene bottle for 24 h. The acidic suspension was repeatedly centrifuged in deionized water at 6000 rpm for 10 min until pH > 6. The obtained sediment was collected and dispersed in deionized water, followed by sonication under a nitrogen atmosphere (1 h). Finally, the MXene suspension was centrifuged at 3500 rpm for 20 min to obtain a uniform MXene colloidal solution. The colloidal solution was further freeze-dried to obtain delaminated MXene powder and stored at 4 °C before use.

Fabrication of TL NPs Composite Microspheres

Three g of TPs and 1 g of theophylline were weighed and added to 10 g of deionized water, and the mixture was sealed and maintained at 80 °C for heat preservation (350 rpm, 30 h). The system was then allowed to stand undisturbed at room temperature for 24 h under sealed conditions to facilitate complex formation. The product was centrifuged at 10,000 rpm to isolate the precipitate, followed by freeze-drying to obtain the TL NPs complex as a stable powder. A concentration gradient was prepared by separately dissolving 25, 50, 75, 100, and 125 mg of the freeze-dried TL NPs complex in 3 g of deionized water, respectively (equivalent to 8.3, 16.7, 25.0, 33.3, and 41.7 mg mL^–1, respectively).

Fabrication of In Situ Bioelectronic Interfaces

Briefly, 10 g of gelatin and 3 g of glycerin were dispersed into 3 mL of deionized aqueous dispersion containing freeze-dried TL NPs complex and 25, 50, 75, 100, or 125 mg of MXene nanosheets, respectively, followed by heating and stirring for 30 min in a beaker. Subsequently, the mixed colloidal solution was transferred into a vacuum drying oven, where a vacuum degree of 0.5 atm was maintained for 2 min, and then poured into molds to form in situ TLMG hydrogel bioelectronic interfaces.

Fabrication of Electrode Array

The stretchable serpentine interconnects were used to establish the electrical connection between the biogel and the wireless circuit module. Ag-based interdigital electrodes were patterned using a mask-template screen-printing process to achieve a stable and precise layout. A 325-mesh screen (aperture ∼40 μm, emulsion thickness 10–15 μm) was employed to define the microtraces. Specifically, 5 g of commercial Ag paste with a viscosity of 8–20 Pa·s at 25 °C (100 s^–1) was uniformly scraped across the stencil at a speed of 50 mm min^–1. The Ag traces were then heated at 30 °C for 20 min to solidify and ensure conductivity, preventing signal crosstalk. To avoid signal crosstalk, each sensing unit was designed with an independent electrode, and the layout of the tactile array is shown in Figure S10.

Characterizations

Scanning electron microscopy (SEM) observation and energy dispersive spectroscopy (EDS) were performed on a field emission scanning electron microscope (SU8010, Hitachi). The roughness and 3D profilometry were measured via a laser microscope (KEYENCE VK-X150). The transmission electrorn microscopy (TEM) images were obtained by JEM-F200 at an accelerating voltage of 80 kV. Morphological observation was carried out using a polarizing light microscope (BX43, Olympus). The atomic force microscope (AFM) images were obtained on a Bruker Multimode 8, Germany. The average surface roughness and 3D profilometry of the composites were measured via confocal microscopy (CSM700, Zeiss). The Fourier transform infrared (FTIR) spectra were recorded in the range 400–4000 cm^–1 at room temperature (Tensor II, Bruker). Thermal stability was analyzed with a TGA thermal analyzer (TA-60WS, Shimadzu), where the specimens (5–8 mg) were in an aluminum pan and then heated from 30 to 800 °C (10 °C min^–1) under nitrogen. For differential scanning calorimetry (DSC 8500, PerkinElmer), the samples were heated in a ceramic pan at the rate of 20 °C min^–1 from – 50 to 30 °C under a nitrogen atmosphere.

Skin Contact Impedance Measurement

The electrode-skin contact impedance was obtained using an LCR meter (E4980AL, Keysight Technologies) with a frequency range from 20 Hz to 1 kHz. All the electrodes had the same size of 1 × 1 cm². Each pair of electrodes was placed on the subject’s forearm, with a 5 cm spacing. All measurements were conducted under the same conditions.

Conductivity Measurement

The electrical conductivity (σ) was measured using a four-probe square resistance tester (HPS2523, China) and was calculated according to the following equation:

$$\sigma =\frac{1}{R_{\mathrm{s}}\times d}$$ 1

where R _s (Ω sq^–1) was the measured sheet resistance, and d is the thickness of the samples.

Mechanical Performances

Tensile mechanical measurements were performed on a universal testing machine (Zwick/Roell). The tensile tests were conducted on long-strip samples (10 mm in length and 2 mm in width at the narrow parallel section) that were pulled by a 200 N load cell at a constant velocity of 50 mm min^–1 at room temperature. Toughness was determined by integrating the area under the engineering stress–strain curve for each sample. The Young’s modulus was calculated from the slope of the initial linear region (0–5%) of the engineering stress–strain curves. For tear testing, the samples were cut into trouser-shaped specimens (length, 50 mm; width, 15 mm; crack length, 20 mm). The two arms of each sample were clamped, with one arm fixed and the other stretched at a rate of 50 mm min^–1.

Long-Term Stretch and Bending Stability Testing

For high-frequency stretch measurement, the sample was horizontally stretched at periodic frequencies ranging from 0.1 to 1.5 Hz by controlling a stepper motor with a computer-controlled stage (MK-9635–1, Maike Mechanical Testing Co., Ltd.). The strain applied to each sample was 50%, and the test speed was 2 mm min^–1 for 10,000 cycles. To test the bending stability, the sample (3 × 10 mm²) was fixed on a cross-biaxial linkage horizontal tension machine (GD150, PDV), and bending for 50,000 cycles was controlled by the movement of the stepper motor at a frequency of 1 Hz.

Electrochemical Assessment

To characterize the sensing performances, the relative current variations under different pressures were recorded using an electrochemical workstation (Nova 2, Metrohm) at 1 V. A digital multimeter with scanning matrix switches (PXI-4072, PXI-2530, PXI-2630B, National Instruments) was employed to acquire the output signals of the array device, which used soft copper wire for a stable electrical connection. For the experiments involving wearable sensors, human participants provided informed written consent to wear the device and participate in the study. All experiments involving human participants were carried out with the approval of the Committee on the Use of Humans as Experimental Subjects at Beijing Forestry University (Approval No. 202400305).

In Vitro Biocompatibility Assessment

The TLMG gel-conditioned medium was used for in vitro biocompatibility tests. To prepare the TLMG gel-conditioned medium, hydrogels at the corresponding dilution ratios were incubated in Dulbecco’s Modified Eagle Medium (DMEM) at 37 °C for 24 h, and pristine DMEM was used as the control.

L929 cells were plated in 96-well plates at a density of 2.5 × 10⁴ cells per well and allowed to adhere for 24 h. The cells were then treated with the TLMG gel-conditioned medium and incubated at 37 °C for another 24 h in 5% CO₂. According to the manufacturer’s instructions, in vitro cytotoxicity was evaluated using a Cell Counting Kit-8 (CCK-8) assay. The experiment was performed in five replicates to determine the mean values and standard deviations (SDs). The cytocompatibility of the hydrogel was assessed using a live/dead viability/cytotoxicity staining kit (Thermo Fisher Scientific). After staining with 100 μL of Calcein AM/EthD-1 dye for 15 min, an inverted fluorescence microscope (IX73, Olympus) was used to image the live cells.

Hemolysis Assay

To verify the hemolytic activity of the TLMG gel, mouse peripheral blood was stirred with a glass rod to eliminate fibrinogen, resulting in defibrinated blood. The erythrocytes were washed 2–3 times with a 0.9% sodium chloride solution to create a 4% suspension. The TLMG gel, PBS, and ddH₂O were added, respectively. The mixture was kept in a thermostat for 1 h, with the conditions set at 37 °C and 60 rpm, leading to the observation of hemolysis.

In Vitro ROS Scavenging Efficiency

A 2,2-diphenyl-1-picrylhydrazyl (DPPH) free radical scavenging assay was conducted to evaluate the ROS scavenging efficiency. Briefly, different amounts of TLMG gel samples were dispersed in 3.0 mL of DPPH ethanol solution (100 μM, Sigma) and homogenized using a tissue grinder. Ethanol was used as the control. The ROS scavenging efficiency of a TL NPs suspension containing the same amount of TL NPs as in the hydrogel was also evaluated. After the mixture was stirred and incubated in the dark for 30 min, the absorbance of the supernatants was measured at 517 nm using a microplate spectrophotometer (SpectraMax Paradigm, Molecular Devices). The DPPH scavenging ratio was defined as

$$\mathrm{DPPH}\mathrm{scavenging}\mathrm{ratio}\%=\frac{A_{\mathrm{control}}-A_{\mathrm{sample}}}{A_{\mathrm{control}}}\times 100\%$$ 2

Remaining radical ratio was defined as

$$\mathrm{Remaining}\mathrm{radical}\mathrm{ratio}\%=100\%-\mathrm{DPPH}\mathrm{scavenging}\mathrm{ratio}\%$$ 3

Intracellular ROS Scavenging Efficiency and Antioxidation

The intracellular ROS scavenging efficiency was evaluated using two methods: the H₂O₂ scavenging assay and the ROS assay. To assess H₂O₂-induced oxidative stress injury, cells were exposed to H₂O₂ (Sigma-Aldrich) at various concentrations for 2 h, and a challenge with H₂O₂ was performed at a final concentration of 1.3 mM. L929 cells (1 × 10⁴ cells/well) were seeded in 96-well plates and pretreated with 50 μg mL^–1 TL NPs or 5 mg mL^–1 TLMG gel. The cells were then exposed to 1.3 mM H₂O₂ in serum-free DMEM for 2 h to mimic the oxidative stress microenvironment. Subsequently, the cell viability was assessed using a Cell Counting Kit-8 (CCK-8) assay.

Next, the intracellular antioxidant efficiency was evaluated by using a ROS assay kit (Solarbio Life Science). L929 cells (1 × 10⁴ cells/well) were seeded in 96-well plates and incubated for 24 h. The cells were treated with Rosup (300 μg mL^–1) in DMEM containing 10% FBS, 50 μg mL^–1 TL NPs, or 5 mg mL^–1 TLMG gel for 4 h. 2′,7′-Dichlorodihydrofluorescein diacetate (DCFH-DA) was employed to detect intracellular ROS levels, and fluorescence images were obtained using a fluorescent microscope (IX73, Olympus). Quantification was performed using ImageJ software.

In Vivo Biocompatibility Assessment

Male C57BL/6 mice (20–25 g) were used for the in vivo biocompatibility study. Before injection, all hydrogels were sterilized with UV light for 1 h. Anesthesia was induced with isoflurane (2–3% in oxygen) and was maintained using a nose cone. The dorsal hair was removed, and subcutaneous hydrogel injection was performed using a 25 G 1 mL syringe with a controlled hydrogel volume of approximately 50 μL.

The activity of the mice was continuously monitored after subcutaneous injection. Four weeks after injection, the animals were euthanized by CO₂ inhalation. The relevant skin tissues and major organs were excised and collected for biocompatibility analysis. The subcutaneous injection model was performed on three independent samples, all yielding similar results.

In Vivo Hemostasis Effect

Sprague–Dawley (SD) rats (male, 12–14 weeks) were selected to investigate the hemostatic effect of TLMG gel using a hemorrhage injury model. The rats were anesthetized with isoflurane (3–4% in oxygen). Bleeding without intervention was designated as the control group. The liver was exposed, and a preweighed filter paper was placed under the liver. A circular liver section with a 6 mm diameter was resected to establish the liver perforation wound model. Next, 200 μL of TLMG in situ hydrogel was directly applied to the bleeding site. A digital camera was used to record the hemostatic process for 3 min. Blood loss was calculated based on the weight difference of the filter paper before and after soaking with blood. The time to achieve hemostasis was also recorded following liver puncture and treatment.

In Vivo Wound Healing Study

Male C57BL/6 mice (20–25 g) were employed and acclimated to the animal facility for at least 1 week prior to surgery. They were routinely weighed and maintained in a specific pathogen-free (SPF) environment under a 12/12 h light–dark cycle at constant temperature and humidity. Chow and water were provided ad libitum. Before injection, all solutions were filtered using 0.22-μm organic filter membranes, and all hydrogels were sterilized with UV light for 1 h. The mice were anesthetized with isoflurane, and circular full-thickness wounds (8 mm diameter, created using a biopsy punch; KAI Medical) were made on the shaved and disinfected dorsal skin. The wounds were immediately treated with sterile saline, MG gel, TL NPs, or TLMG gel. Subsequently, the wounds were protected with an occlusive dressing (Tegaderm, 3M) and secured with medical bandages. To quantitatively assess the wound healing process, the mice were observed daily and photographed on days 0, 3, 7, 10, and 14 using a standard camera. The relative wound area was defined as

$$\mathrm{Relative}\mathrm{wound}\mathrm{area}(\%)=\frac{S_{\mathrm{n}}}{S_0}\times 100\%$$ 4

Where S ₀ and S _n indicate the wound areas quantified by the ImageJ software on days 0 and n, respectively. The mice were euthanized by CO₂ inhalation on days 3, 7, and 14 postwounding, and the wounded skin samples were harvested and processed for DHE fluorescence, histology, and immunofluorescence.

DHE Fluorescence, Histology, and Immunofluorescence

Mouse wounds and organ tissues, following completion of the study (skin tissues on days 3, 7, and 14; organ tissues after 4 weeks), were excised and collected. All tissue sections were scanned using a Panoramic SCAN II digital scanner (3DHISTECH, Hungary) and viewed with CaseViewer 2.4 (3DHISTECH, Hungary). Dihydroethidium (DHE) fluorescence was used to assess in vivo ROS scavenging. The skin tissues on day 3 were collected and embedded in a Tissue-Tek OCT compound (Thermo Fisher). Cryosections (10 μm) were incubated with a self-luminous fluorescence quencher for 5 min, followed by rinsing for 10 min in distilled water. The sections were then incubated with a superoxide-sensitive dye, DHE (10 μM in 0.01% DMSO, Invitrogen), at 37 °C in the dark for 30 min. Subsequently, 4′,6-diamidino-2-phenylindole (DAPI; Invitrogen) was applied to stain the nuclei.

The regenerated skin tissues on days 7 and 14 were collected for histological and immunofluorescence analysis. The tissues were fixed in 10% buffered formalin and processed into paraffin sections with a thickness of 4 μm. H&E staining was performed, and the width of granulation tissue was quantified from the H&E images. For immunofluorescence analysis, the paraffin-embedded skin sections were deparaffinized and subjected to antigen retrieval. The sections were blocked with goat serum for 2 h at room temperature and then incubated overnight at 4 °C in a humidified chamber with the following primary antibodies: F4/80 (1:200, CST), CD86 (1:200, Bioss), CD31 (1:400, Abcam), and α-SMA (1:500, CST). The corresponding secondary antibodies were incubated for 1 h at room temperature. DAPI was used for nuclear counterstaining, and the sections were sealed using an antifluorescence quenching solution. Quantification was performed using ImageJ software.

In Vivo Electromyography (EMG) Monitoring

For sciatic nerve stimulation, male SD rats (12–14 weeks) were deeply anesthetized with isoflurane (3–4%). The hindquarter fur was removed using an electric shaver and a depilation cream. The gluteus maximus and biceps femoris muscles were dissected to expose the sciatic nerve. Electrical stimulation (DS3, Digitimer) was performed using biphasic rectangular electrical current pulses ranging from 300 to 800 μA. Following the skin incision, the TLMG in situ electronic patch was attached to the gastrocnemius medialis muscle with gentle pressure for 20 s. For EMG recordings, a multichannel physiological signal acquisition system (RM6240) and an electrochemical workstation (Chi 760e) were employed. The signal-to-noise ratio (SNR) was defined as

$${\mathrm{SBR}}_{\mathrm{dB}}=10{\log }_{10}\left(\frac{P_{\mathrm{signal}}}{P_{\mathrm{noise}}}\right)$$ 5

Statistical Analyses

Statistical analyses were performed with SPSS 26.0 software (IBM, USA). All data were presented as the mean ± standard deviations (SD). In the statistical analysis for comparison between two or multiple groups, t tests or one-way analysis of variance (ANOVA) were conducted, respectively. The thresholds for statistical significance were set at P-value <0.05 (*), P-value <0.01 (**), and P-value <0.001 (***).

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsnano.5c17466.

• Supplementary Notes S1–S2: Importance of developing an in situ integrated biointerfacing platform, conversion mechanism of the liquid–solid phase. Supplementary Figures S1–S40: Recent progress in the realm of wearable intelligent wound dressings, Schematics, AFM, TEM, SEM, Digital images, Strain–stress curves, Puncture resistance test, Current variation, TGA, DSC, conformal contact, Adhesive durability and stability, Adhesive strength and Biocompatibility of TLMG biointerface. Multichannel EMG comparison of electrode array based on ex-situ and in situ gels, hardware parameters for wireless real-time recording. Supplementary Table S1: Key performance metrics benchmarking of TLMG biointerface with recent literature (PDF)

• Supplementary Video S1: Current variation during prolonged cyclic deformation (MP4)

• Supplementary Video S2: Adhesion performance (MP4)

• Supplementary Video S3: Hemostasis performance (MP4)

• Supplementary Video S4: Biocompatibility (MP4)

• Supplementary Video S5: Real-time EMG signal monitoring of in situ TLMG electrode patch (MP4)

#.

W.Y., S.H., and G.Z. contributed equally. The manuscript was written through the contributions of all authors. All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.

Due to a production error, in the version of this paper that was published ASAP February 12, 2026, authors Wenting Yu and Sanwei Hao were linked to the wrong affiliations. The corrected version was posted February 24, 2026.