A self-healing e-skin for quadruple-modal sensing

A self-healing e-skin for quadruple-modal sensing | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article A self-healing e-skin for quadruple-modal sensing Henghui Wang, Hao Xue, Yapeng Song, Junjie Yuan, Huijun Cao, Quan Yuan, and 7 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7503393/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract E-skins capable of multimodal perception are essential for intelligent human-machine interaction, yet integrating real-time responsiveness with structural self-healing across multiple sensing modalities remains a significant challenge. Here, we report a biologically inspired, self-healing e-skin that enables the perception of tactile, pressure, nociceptive, and thermal stimuli. The device adopts a vertically stacked compact architecture comprising double-network cross-linked polymer layers and eutectic liquid metal layers, enabling rapid and complete structural healing even after severe mechanical damage over a wide temperature range. We demonstrate that the e-skin retains its functional integrity after healing, with triboelectric and capacitive sensing units enabling high-sensitivity, spatiotemporally resolved tracking of tactile and pressure stimuli, respectively. Meanwhile, by quantitatively analyzing the mechanoluminescent and thermochromic spectra, the optical waveguide sensing units enable real-time optical encoding of nociceptive and thermal stimuli, thereby allowing effective classification of impact and burn injury risks. Our work lays a solid foundation for the development of intelligent robotic systems capable of adaptive perception and injury prevention in complex and dynamic human-machine environments. Physical sciences/Materials science Physical sciences/Optics and photonics Physical sciences/Physics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Skin, the largest peripheral sensory organ in biological systems, exhibits complex sensory functionalities mediated by an array of cutaneous receptors 1 . These include free nerve endings, Meissner's corpuscles, Ruffini endings, and Merkel disks, which collectively enable multimodal perception of thermal, nociceptive, tactile, and pressure stimuli. Inspired by these biological sensing modalities, the development of e-skin with integrated multimodal sensing capabilities has emerged as a critical frontier in the research of artificial sensory systems 2 – 6 . Rigid or semi-rigid materials are typically employed in sensor construction to ensure signal stability, combined with structural engineering to enhance sensitivity and expand detection range 7 – 9 . Nevertheless, mechanical incompatibility with biological tissues 10 , along with wear and accidental damage during operation 11 , remains a primary cause of device failure. These challenges underscore an urgent need for e-skin possessing both flexibility and intrinsic self-healing capabilities. To bridge the functional gap between e-skin and natural skin, flexible, self-healing polymers with varied electrical properties (insulators, semiconductors, and electronic/ionic conductors 11 ) have been employed in the construction of e-skins for applications in prosthetics 10 , 12 , robotics 13 – 15 , health monitoring 16 , 17 , and human-machine interfaces 18 , 19 . The self-healing mechanisms of these polymers primarily involve: (i) extrinsic repair, in which embedded microcapsules rupture to release healing agents 20 ; and (ii) intrinsic repair, which relies on the reconstruction of dynamic covalent bonds (e.g., disulfide 15 , imine 21 , and boronic ester 22 bonds), supramolecular interactions (e.g., hydrogen bonds 21 , 23 , metal-ligand coordination 24 , π-π stacking 25 , and van der Waals forces 26 ), and topological entanglement 27 within supramolecular networks. Ideally, e-skin should combine efficient self-healing with sufficient toughness to ensure robust sensing performance. However, a trade-off exists between healing efficiency and bond strength 28 . To address this limitation, recent research has focused on molecular design to develop polymers that integrate multiple self-healing mechanisms 16 , 29 . Despite these advances at the material level, the system-level integration of diverse sensing units, comprising distinct materials 14 , 28 and architectures 14 , 16 , remains a major obstacle, impeding the development of e-skins capable of reliable multisignal acquisition while retaining flexibility and self-healing functionality. Addressing system-level integration challenges requires not only innovations in materials but also advances in sensing architecture. Unimodal configurations often entail trade-offs among functionality, sensitivity, and detection range, while multimodal strategies may offer a more promising route toward comprehensive and high-fidelity perception. Thus far, progress in e-skin development has largely centred on bimodal sensing architectures, leading to configurations such as triboelectric-capacitive tactile sensors 30 , triboelectric-piezoresistive tactile sensors 31 , piezoelectric-piezoresistive tactile sensors 32 , and light-resistance-based tactile-temperature sensors 33 . Despite significant progress in compact integration and signal decoupling, these architectures remain limited in sensory diversity and largely overlook self-healing functionality. Notably, triboelectric sensing can mimic the function of Meissner's corpuscles, enabling the detection of surface contact and frictional stimuli 31 , as well as material discrimination 34 . Capacitive sensing may resemble the roles of Merkel discs or Ruffini endings in the dermis, allowing for sustained pressure detection under moderate strain 30 , and when configured as arrays, spatial resolution 28 . Optical waveguide sensing can emulate the function of free nerve endings by capturing light spectrum variations induced by external stimuli 33 . When doped with thermochromic or mechanoluminescent particles, it enables the perception of thermal 35 or nociceptive 14 stimuli under large strains. Given the functional complementarity among triboelectric, capacitive, and optical waveguide sensing modalities, system-level integration may offer a promising strategy for developing a quadruple-modal e-skins capable of simultaneously sensing tactile, pressure, nociceptive, and thermal stimuli. Here, we present a self-healing e-skin featuring a vertically stacked, five-layer architecture (Fig. 1 ). The 1st, 3rd, and 5th layers serve as insulated layers composed of a double-network cross-linked polymer (DCP) formed from polydimethylsiloxane (PDMS) and polyborosiloxane (PBS). Red or blue thermochromic particles (TPs) are embedded in the 1st and 3rd layers, while ZnS:Cu mechanoluminescent particles (MPs) are incorporated into the 5th layer. The 2nd and 4th layers consist of identical 3 × 3 electrode arrays, fabricated by screen-printing eutectic Ga-In-Sn liquid metal (ELM) doped with Ni particles (NPs) onto the underlying insulating layers. Functionally, the 1st layer and each electrode in the 2nd layer forms a single-electrode triboelectric unit for tactile sensing; each pair of parallel electrodes between the 2nd and 4th layers forms a parallel-plate capacitive unit for pressure sensing; and the 1st and 3rd layers, and the 5th layer, independently serve as optical waveguide units for temperature and nociceptive sensing, respectively. By integrating triboelectric, capacitive, and optical waveguide sensing modalities into a unified platform, our e-skin can achieve compact quadruple-modal perception (touch, pressure, nociception and temperature) alongside self-healing, high sensitivity, spatiotemporal resolution, and scalable fabrication. These features mark a critical step toward the practical implementation of e-skin technologies. Results Material design and characterization To enable a self-healing e-skin with quadruple-modal sensing functionality, the polymer matrix must rapidly self-repair under harsh conditions, maintain structural integrity and optical transparency, and exhibit mechanical responsiveness akin to natural skin. Both PDMS and PBS are silicone-based polymers. PDMS exhibits excellent toughness and optical transparency but lacks self-healing functionality, whereas PBS is viscoelastic, transparent, and self-repairable. To integrate these complementary properties, we synthesized the supramolecular DCP via one-pot strategy comprising both PDMS and PBS chains (Fig. 2 a). The PDMS chains were formed through hydrosilylation between polymethylhydrogensiloxane (PMHS) and dimethylvinyl-terminated dimethylsiloxane (VPDMS), while the PBS chains were generated by esterification between boric acid (BA) and hydroxyl-terminated polydimethylsiloxane (PDMS-OH). In this supramolecular network, rigid PDMS chains provide mechanical strength, while flexible PBS chains impart self-healing via B-O dative bonds, hydrogen bonds, and topological entanglements. Four DCP variants with PDMS:PBS weight ratios ranging from 1:1 to 1:4 (denoted as DCP-1 to DCP-4) were synthesized to optimize overall performance. A characteristic absorption band at 1340 cm − 1 in the Fourier-transform infrared (FTIR) spectra (Fig. 2 b), attributed to the B-O stretching vibration in the Si-O-B bond 22 , confirms the successful formation of PBS chains. The absence of Si-H stretching vibration at 2160 cm − 1 indicates the complete consumption of Si-H groups in the original VPDMS (Fig. S1 ), confirming the formation of PDMS chains. Stability tests (Fig. 2 c) reveal that PBS undergoes macroscopic collapse within 40 min due to its fluidic nature, whereas all DCP variants, reinforced by rigid PDMS chains, maintain PDMS-like structural stability. Interestingly, only DCP-3 and DCP-4 fully healed within 120 min (Fig. 2 d), while DCP-1 and DCP-2 retained visible scratches even after 180 min, indicating a discernible difference in healing efficiency. Human skin typically ruptures at a strain of 35–115% 36 , whereas all DCP variants exceed this range, exhibiting elongations at break between 150% and 418% (Fig. 2 e). Among them, DCP-3 displays linear elasticity across a broad strain range of 10–150%, which is essential for ensuring consistent sensing performance under varying mechanical loads. Moreover, cyclic compression tests (55% strain, ~ 1 Hz) revealed that all DCP variants exhibited overlapping hysteresis loops but non-overlapping loading-unloading curves over 10 cycles, indicative of viscoelastic behavior (Fig. 2 f). However, DCP-4 exhibited the lowest compressive modulus and the most pronounced hysteresis under high strain, potentially resulting in a narrower detection threshold and delayed signal responses. Notably, rheological tests showed that all DCP variants exhibited the damping factors (tan δ ) below 1 across the full angular frequency range of 0.1–400 rad·s – 1 , reflecting a predominantly elastic response (Figs. 2 g and S2). Especially, DCP-3 displayed the most stable tan δ profile with minimal fluctuations, likely due to a well-balanced composition of PDMS and PBS chains. In addition, differential scanning calorimetry (DSC) analysis demonstrated that all DCP variants possess a glass transition temperature ( T g ) of approximately − 120°C, a melting temperature ( T m ) of around − 40°C, and a degradation temperature ( T d ) exceeding 250°C (Fig. 2 h and Table S1 ), indicating a wide thermal operating window. Taken together, the balance between self-healing efficiency and mechanical response positions DCP-3 as a promising polymer matrix for self-healing e-skin applications. Environmental adaptability and integration Earth’s diverse climates, ranging from polar glaciers to equatorial deserts, impose stringent demands on the environmental adaptability of e-skins (Fig. 3 a), where self-healing conditions and efficiency are essential. A notable advantage of DCP-3 lies in its ability to autonomously self-heal without external stimuli such as heat 14 or light 28 . The self-healing efficiency of its tensile strength reaches 84.4% after 48 h at -10°C (Fig. 3 b), 84.2% within 24 h at 25°C (Fig. 3 c), and up to 97.0% within 60 min at 60°C (Fig. 3 d), demonstrating excellent self-healing performance across a wide thermal range. To assess functional compatibility, MPs (20 wt%) and TPs (1 wt%) were also individually incorporated into DCP-3. The resulting composites retained nominal stress-strain profiles similar to the original matrix (Fig. S3 ). Especially after 24 hours of healing at 25°C, the self-healing efficiencies of tensile strength remained at 76.3% for MP-doped composite and 83.4% for TP-doped composite (Fig. S4 ), indicating that functionalization did not greatly compromise the mechanical response or healing capacity. ELM exhibits a low melting point, high conductivity, and good fluidity, and is widely used for fabricating self-healing circuits 37 . Particularly, doping with NPs can promote Ga 2 O 3 formation to improve wettability with polymer matrix 38 , and also may facilitate magnetic guidance for circuit repair. Meanwhile, we performed UV/ozone treatment to the DCP-3 surface, which forms stable -Ga-O-Si- bonds 39 with ELM and significantly enhances interfacial adhesion. Combined with shadow mask patterning, this approach can greatly improve the patterning resolution of NP-doped ELM onto DCP-3. As shown in Fig. 3 e, two NP-doped ELM circuits were patterned on DCP-3 surfaces doped with red and blue TPs, respectively, and their functionality was demonstrated by LED illumination. Precise alignment and magnetic assistance enabled the fractured red and blue segments to rapidly reconnect and restore stable conductivity, thereby lighting the LED. Simple stretching, bending, and twisting of the healed structure did not affect LED illumination, confirming the robust self-healing capability of both the polymer matrix and the circuit. Additionally, a more complex 3 × 3 electrode array was patterned on the surface of DCP-3 using a shallow mask (Fig. S5 ). The uncovered polymer surface, which retains B-O dative bonds and multiple hydrogen bonds, can provide inherent self-healing capability that facilitates layer-by-layer assembly (Fig. 3 f). The resulting five-layer integrated e-skin exhibits a stable structure without interfacial delamination, particle shedding, or ELM leakage even under large bending deformations, demonstrating excellent structural robustness (Figs. 3 g and 3 h). Finally, two e-skins were diagonally incised; then, through the combination of precise polymer layer alignment and magnetically guided ELM circuit reconnection, we achieved, for the first time, large-area structural and functional re-healing of a fully damaged e-skin (Fig. 3 i), marking a significant breakthrough in self-healing e-skin technology. Triboelectric tactile sensing and applications We further assessed the post-healing functional integrity of the diagonally incised e-skin. Benefiting from the compact stacked design, the 3 × 3 tactile sensing array consists of nine independent single-electrode triboelectric sensing units that operate in a vertical contact-separation mode. During human-machine interaction, triboelectric signals are generated through a combination of triboelectrification and electrostatic induction 34 . To be specific (Fig. 4 a), when a finger (the electropositive material) contacts the DCP-3 layer (the electronegative material), a contact potential difference forms at the interface due to their different work functions, driving electron transfer to the polymer surface; upon separation, the increasing gap enhances the polarization field, causing electrons to flow from the ELM electrode to the reference electrode until charge equilibrium is achieved at maximum separation; as the finger approaches again, the field weakens, inducing a reverse electron flow; and once full contact is restored, the charges balance again, completing the cycle. To evaluate adaptability to tactile pressures, real-time open-circuit voltage waveforms were recorded under constant contact pressures ranging from 0.2 to 16 kPa at 1 Hz (Fig. 4 b), and the correlation between the maximum open-circuit voltage and contact pressure was plotted (Fig. 4 c). It was observed that the open-circuit voltage increased nearly linearly with the applied pressure. When the pressure ranged from 0.2 to 4 kPa, the sensitivity of the triboelectric signal was 0.98 V·kPa – 1 , whereas for pressures ranged from 4 to 16 kPa, the sensitivity decreased to 0.083 V·kPa – 1 . The higher sensitivity at lower pressures can be attributed to an increased contact area on the DCP-3 layer, which facilitates greater charge transfer during contact. However, once the contact area reaches its limit, higher pressure is needed to induce electron cloud overlaps at the polymer-skin interface 34 , which reduces charge transfer efficiency and thus limits sensitivity. To further verify signal stability, over 500 repeated clicking cycles were conducted at pressures of 0.2, 1.2, and 4 kPa (Fig. 4 d). At all three pressure levels, the output waveforms remained consistent throughout the early, middle, and late stages, with amplitude retention exceeding 99% over the entire test duration (Fig. 4 e). The signals show similar waveforms under different pressures, with response times ranging from 100–125 ms during pressure application and 98–120 ms during pressure release, which are faster than human response time to tactile stimuli (139 ms) 40 . The high-sensitivity contact-separation triboelectric signals endow the e-skin with versatile tactile sensing capabilities. When attached to the wrist, it can monitor subtle muscle movements. During fist clenching (Fig. 4 g), muscle contraction causes separation between the e-skin and the wrist skin, creating instantaneous open-circuit voltage peaks whose magnitude correlates with the gripping force (Video S1). Similarly, swallowing movements can be readily detected (Video S2). Moreover, when attached to the foot arch, the e-skin can monitor gait frequency and arch loading conditions. During walking, jogging, and running (Fig. 4 h), changes in gait frequency alter the contact-separation rate between the arch skin and the e-skin, producing signals with distinct frequencies. Meanwhile, because the tester does not exhibit foot inversion or eversion, the magnitude of the open-circuit voltage remains stable. Furthermore, the 3 × 3 tactile sensing array endows the e-skin with spatial tactile mapping capability. When a tester sequentially taps the electrodes in the order of 1→5→6→3→2→4 (Fig. 4 i), the multi-channel data acquisition unit records time-correlated waveforms without crosstalk, demonstrating excellent tactile trajectory tracking performance. Additionally, under simulated conditions where multiple tactile stimuli with varying pressures and frequencies are applied simultaneously (Videos S3 and S4), the e-skin exhibits strong anti-interference capability and stable signal recognition, facilitating high-precision measurements in complex practical scenarios. Capacitive pressure sensing and applications The 3 × 3 pressure sensing array consists of nine independent parallel-plate capacitive sensing units. For each unit, the upper ELM electrode is shared with the triboelectric sensing unit, while the bottom ELM electrode is independent, with the intermediate DCP-3 layer acting as the dielectric. During human-machine interaction (Fig. 5 a), finger pressure compresses the DCP-3 layer, decreasing its thickness ( d ) and increasing the capacitance ( C ); at maximum pressure, the polymer reaches its minimum thickness ( d min ) and maximum capacitance ( C max ); upon release, the polymer gradually recovers, and the capacitance decreases accordingly; once contact separation occurs, both the thickness ( d 0 ) and capacitance ( C 0 ) return to their initial values. This process enables the establishment of a pressure-capacitance relationship through the relative capacitance variation (Δ C / C 0 = C / C 0 –1) 41 . Notably, the minimal vertically stacked structure facilitates self-healing of the e-skin; however, the triboelectric effect may interfere with capacitive signals 40 . To address this issue, a programmable frequency-sweeping strategy is suggested to dynamically coordinate the switching of each triboelectric and capacitive sensing unit based on the detected stimulus intensity, thereby enabling highly sensitive and spatially resolved tactile and pressure sensing. The peak values of relative capacitance variation under pressures ranging from 1 to 40 kPa at 1 Hz were extracted from temporal waveforms and displayed as bar plots (Fig. 5 b). A corresponding sensitivity curve was derived by plotting the maximum relative capacitance variation (Δ C / C 0 = C max / C 0 –1) as a function of pressure (Fig. 5 c). Data fitting revealed a linear sensitivity of 0.108 kPa – 1 within the 3–28 kPa range, while sensitivity declined markedly outside this window. This response may arise from distinct deformation mechanisms of the DCP-3 layer under varying pressure regimes (Fig. 2 e). At moderate pressures, elastic deformation dominates, yielding a linear pressure-thickness relationship. At low pressures, polymer chain slippage induces viscous flow and nonlinear thickness variation. At high pressures, chain rearrangement may restrict compressibility, while intensified electrode deformation enhances edge-field effects, leading to increased parasitic capacitance and a reduction in effective signal output. In addition, the upper pressure limit of 28 kPa corresponds to lifting a weight of approximately 100–150 pounds with both hands, basically fulfilling the practical requirements of human-machine interaction. Signal stability was assessed under pressures of 4, 16, and 28 kPa through more than 500 repeated clicking cycles (Fig. 5 d). Similar waveforms were maintained at all three pressures across the early, middle, and late stages (Fig. 5 e). Response times ranged from 264 to 292 ms during pressure application and from 244 to 270 ms during release across all pressure levels (Fig. 5 f). Note that the relaxation time decreased with increasing pressure, suggesting that the DCP-3 layer, as a viscoelastic polymer, operated within a hyperelastic regime in the 3–28 kPa range. By leveraging the complementary characteristics of both triboelectric and capacitive sensing units, the e-skin can achieve an extended pressure sensing range spanning from 0.2 to 28 kPa. Additionally, spatial pressure distribution of the capacitive sensing array was explored by placing a 100 g weight on the surface of the diagonally healed e-skin through plastic sheets shaped as “X”, “M”, or “U” (Fig. 4 g). When using the “X”-shaped plate, five sensing units were subjected to the weight, whereas in the “M” and “U” configurations, seven units were loaded. As a result, the average pressure per unit was higher in the “X” configuration, leading to larger relative capacitance variation compared to the other two cases. It is also noted that slight differences in the electrode coverage areas across the “X”, “M”, and “U”-shape plates resulted in minor variations in pressure distribution across the capacitive sensing array. The capacitive pressure sensing array and triboelectric tactile sensing array in our self-healing e-skin exhibit functional complementarity. Their synergistic operation enables more accurate responses to mechanical stimuli and provides comprehensive signal feedback for huma-machine interaction. Optical-waveguide nociceptive sensing and applications Inspired by the nociceptive and thermosensitive functions of free nerve endings, MPs and TPs were respectively integrated into the DCP-3 matrix to enable pain and temperature sensing, thereby greatly enhancing sensitivity and expanding the detectable stimulus range. ZnS:Cu, a representative mechanoluminescent material (Fig. S6 a), exhibits bright and reproducible green emission (~ 520 nm), primarily attributed to Cu 2+ luminescent centers 42 – 44 . The underlying mechanism involves (Fig. 6 a): (i) pressure-induced piezoelectric polarization in the non-centrosymmetric wurtzite ZnS lattice (Fig. S6 b); (ii) the resulting piezopotential tilts the conduction and valence bands, releasing electrons from shallow donor traps into the conduction band; (iii) these detrapped electrons migrate to Cu 2+ centers, where non-radiative recombination with holes promotes of Cu 2+ outer electrons from the 6 A 1 ground state to the metastable 3 T 1 state; (iv) subsequent spin-forbidden radiative relaxation from 3 T 1 to 6 A 1 , produces the characteristic green emission. Although MPs intrinsically require high stress to trigger light emission, recent research suggests that triboelectric effects can markedly enhance mechanoluminescent responses and reduce the activation threshold 45 . Jeong et al. 42 reported that under compression, interfacial delamination and air-gap formation occurred between rigid ZnS:Cu MPs and the softer polymer matrix due to mechanical mismatch. This delamination, combined with triboelectric charging arising from their distinct positions in the triboelectric series 46 , facilitates additional electron detrapping from shallow donor levels, thereby significantly amplifying light emission. Jeong’s findings challenge the conventional view that luminescence is solely governed by matrix elasticity and instead underscore the importance of interfacial triboelectric fields. Building on this, we simulated the deformation of a DCP-3 matrix embedded with a single ZnS:Cu MP under applied pressure to quantify the effects of interfacial strain mismatch on electric potential and field distributions. Assuming global electroneutrality and a uniform charge distribution, a fixed positive surface charge density of 10 − 4 C·m − 2 was assigned to the MP at the delamination interface 42 , balanced by a deformation-induced negative surface charge on the DCP-3 matrix. For simulation details, see Supplementary Information. Simulations revealed peak electric potential and field values at the horizontal symmetry center, coinciding with interfacial delamination zones (Figs. 6 b and 6 c). Surprisingly, the triboelectric voltage (i.e., the maximum electric potential difference) increased linearly with matrix strain, whereas the maximum delamination-induced air-gap distance ( d max ) exhibited a nonlinear strain dependence (Fig. 6 d). Our findings strongly indicate that the triboelectric voltage ( U ), rather than the triboelectric field ( E = U / d max ) in the Jeong’s study 42 , maintains a more robust and predictable correlation with matrix strain. This relationship provides a reliable basis for sustaining a strong coupling between strain and luminescence intensity, even within the inelastic deformation regime of the polymer matrix. By leveraging the strain-triboelectric voltage-mechanoluminescence (STM) coupling mechanism, our findings provide a foundational strategy for developing optical waveguide-nociceptive sensor capable of selectively responding to both elastic and non-elastic deformation in self-healing polymers. Unintended physical contact between robots and humans poses a significant safety concern in interactive environments 47 . Current industrial safety standards, such as ISO 10218 and ISO/TS 15066, impose conservative limits on robot speed, force, and power to minimize injury risk in shared workspaces 48 . While effective in preventing harm, these rigid constraints often compromise robotic efficiency and lack the ability to distinguish between benign and hazardous interactions in real time. To overcome these limitations, by quantifying the energy flux density (i.e., the energy transferred per unit area) during collision events, we developed a control framework for physical human-robot interaction grounded in biomechanical injury thresholds. Remarkably, based on STM coupling mechanism, the energy flux density exhibits a piecewise-linear relationship with optical intensity (Fig. 6 f and Video S5), consistent with the simulated response of triboelectric voltage to strain (Fig. 6 d). This correlation allows the maximum optical signal to serve as a real-time proxy for estimating energy flux and categorizing collision risk. Specifically, medium-risk collisions (2–6 mJ·mm − 2 ), which typically associates with contusions, are recommended to initiate a robotic deceleration protocol, while high-risk collisions (> 6 mJ·mm − 2 ), which may cause fractures, should trigger an emergency stop. Built upon its high-sensitivity, strong interference resistance, and rapid-response luminescence characteristics (Fig. 6 g), the optical waveguide nociceptive sensing unit can offer a robust and scalable solution for intelligent, adaptive, and safety-aware human-robot interaction. By enabling real-time biomechanical risk assessment, this strategy may achieve an effective balance between robotic operational performance and injury prevention, paving the way for intelligence in collaborative robotics. Optical-waveguide thermal sensing and applications As an essential nociceptive modality, thermal perception in e-skin provides critical feedback for mitigating thermal hazards during physical human-robot interaction. To achieve this, we employed TPs consisting of phase-change microcapsules embedded with a reversible thermochromic dye (Fig. S6 c). Within a specific temperature range, the absorption spectrum of the TPs shifts in response to temperature changes (Video S6), exhibiting excellent reversibility and durability. Due to the low thermal conductivity of the organic polymer shell, the colour-change response time of the TPs is governed more by heat transfer kinetics than by the intrinsic phase transition process 49 . Similarly, the low thermal conductivity of the DCP-3 matrix further delays the thermal response of the thermochromic layer in e-skin. During the thermochromic transition, the spectrometer collects reflected rather than transmitted light, necessitating subtraction of the initial-state reflection spectrum to obtain time-resolved absorption spectra. To avoid spectral overlap with the mechanoluminescent signals, we selected a red TP with a peak absorption at 591 nm that shifts to 577 nm upon activation (Fig. 6 h). A one-dimensional heat conduction model was established across the thermochromic layer between its top surface (in contact with the heat source) and its bottom surface (optically coupled to the waveguide): $$\frac{{\partial T(x,t)}}{{\partial t}}=\alpha \frac{{{\partial ^2}T(x,t)}}{{\partial {x^2}}}$$ 1 , where T ( x , t ) is the temperature at position x∈[0, w ] and time t , w is the thickness of the DCP-3 film, and α is the thermal diffusivity of the DCP-3. Under the assumption of a shallow surface transition (i.e., the contact time t → 0), we surprisingly found that the temperature of the heat source can be analytically estimated as: $$T={T_0}+S \cdot \frac{{I(t)}}{{\sqrt t }}$$ 2 , $$S=\frac{{w({T_0} - {T_{Air}})}}{{\sqrt {\pi \alpha } }}$$ 3 , where T 0 is the transition temperature, T Air is the ambient temperature, and I(t) is the normalized intensity of absorption spectra ( 0 ≤ I(t) ≤ 1 ). For derivation details, refer to the Supplementary Information. We further measured the relationship between the normalized absorption intensity and the square root of colour-change time under heat source temperatures ranging from 40 to 100°C (with ambient temperature T Air =25°C). The resulting curves exhibited an approximately linear slope in the early stage of the thermochromic transition (Fig. 6 i). By fitting the slope values obtained at different heat source temperatures, we observed a near-linear dependence with the slope S = 17.014, which is in excellent agreement with the analytical prediction from Eq. 2 (Fig. 6 j). This result confirms the feasibility of establishing an absorption intensity-temperature sensing strategy. In practical scenarios, the robot can utilize this optical waveguide temperature sensing unit to assess burn risk during human-robot interaction and issue early warnings when the surface temperature exceeds 50°C. This framework supports the deployment of caregiver robots in daily service scenarios with enhanced safety and responsiveness. Discussion In summary, we propose a biologically inspired, self-healable e-skin capable of quadruple-modal perception of tactile, pressure, nociceptive, and thermal stimuli. The e-skin adopts a vertically stacked five-layer architecture, comprising three DCP layers doped with functional MPs or TPs and two ELM layers doped with magnetic NPs. This synergistic configuration enables complete structural self-healing across a broad temperature range (–10°C to 60°C), offering a structurally simple, maintenance-friendly, and modular solution for intelligent human-machine interaction. Through the integration of triboelectric and capacitive sensing modules, the e-skin achieves high sensitivity and spatiotemporal resolution in tracking and responding to tactile and pressure inputs. Meanwhile, the incorporation of optical waveguide sensing modules establishes coupling relationships between energy flux density and mechanoluminescent emission spectra, as well as between heat source temperature and the absorption spectra of the thermochromic layer. These optical correlations enable the quantitative assessment of nociceptive (impact) and thermal (burn) risks, providing a viable strategy for real-time injury prevention in human-machine interactive environments. Overall, our work offers a safe, reliable, and multifunctional framework for next-generation interactive electronics. Future research will focus on integrating machine learning-assisted adaptive perception with full-body deployment on reconfigurable robotic surfaces, ultimately advancing applications in intelligent robotics. Methods Chemicals and materials PMHS ( M n =1700–3200), VPDMS (850–1150 cSt), PDMS-OH (3,500 cSt), BA (≥ 99.5%), and methanol (≥ 99.8%) were purchased from Sigma Aldrich and used as received. ZnS:Cu MPs (~ 5 µm) were obtained from Shanghai Kerun Optoelectronics Technology Co., Ltd. Red and Blue TPs (~ 3 µm, T 0 = 31 ℃) were supplied by Shenzhen Huancai Technology Co., Ltd. Ga (~ 0.5 mm, ≥ 99.99%), In (~ 0.5 mm, ≥ 99.99%), Sn (~ 0.5 mm, ≥ 99.99%), and Ni (1–3 µm, ≥ 99.99%) particles were provided by Helian Xiamen Technology Co., Ltd. Synthesis of polymer films The target polymer film was synthesized via a one-pot method by mechanically stirring the selected polymers with methanol solutions containing functional particles for 1 h. The resulting precursor was then cast into polytetrafluoroethylene (PTFE) molds (100 × 100 × 0.5 mm 3 ), vacuum-degassed at room temperature for 30 min, and subsequently crosslink-cured at 160°C for 12 h to yield the final polymer film. Specifically, pure PBS was synthesized by mixing BA and PDMS-OH at a mass ratio of 1:20, while pure PDMS was prepared by blending PMHS and VPDMS at a mass ratio of 1:10. DCP-1 to DCP-4 were formulated by combining PBS and PDMS precursors at mass ratios ranging from 1:1 to 1:4, depending on the desired composition. Additionally, the methanol solution (50 g·L – 1 ) was prepared by ultrasonically dispersing the corresponding particles in methanol for 15 min. In the final polymer films, MPs and TPs were incorporated at 20 wt% and 1 wt%, respectively, relative to the total polymer mass. Synthesis of particle-embedded ELM alloy Ga, In, and Sn particles were combined at mass ratios of 68.5 wt%, 21.5 wt%, and 10 wt%, respectively, and placed in a PTFE beaker. The mixture was heated to 250°C and held for 1 h under gentle stirring to promote uniform alloying. After naturally cooling to room temperature, 10 wt% of NPs were introduced, followed by vigorous stirring for 1 h to yield the final particle-embedded ELM alloy. E-skin fabrication and Integration A 0.05 mm-thick custom-fabricated 304 stainless steel shadow mask was used to define the surface treatment area. The mask was aligned with the geometry of the e-skin, featuring patterned openings corresponding to the electrode and lead regions for subsequent ELM brushing. Under the protection of this mask, DCP-3 films doped with functional particles were placed in a dark chamber equipped with a UV lamp ( λ max = 185 nm, power density = 85 µW·cm – 2 ) and subjected to UV/ozone treatment for 2 h. After treatment, the shadow mask was retained while the ELM was uniformly brushed onto the exposed regions and guided into patterned leads via magnetic alignment. The e-skin was then assembled by sequentially stacking: (i) an MPs-doped DCP-3 layer with an ELM pattern, (ii) a TPs-doped DCP-3 layer with an ELM pattern, and (iii) a TPs-doped DCP-3 layer without ELM. The electrode regions were carefully aligned and the leads positioned on opposite sides to prevent short-circuiting. In addition, Ag wires (0.1 mm in diameter) were connected to the ELM leads to enable subsequent electrical measurements. The assembled e-skin was vacuum-treated at 60°C for 1 h to ensure structural integration, followed by UV/ozone treatment for 1 h to remove surface tackiness. Characterization FTIR spectra were collected using a Nicolet iS50 spectrometer in attenuated total reflectance (ATR) mode over the range of 400–4000 cm − 1 . X-ray diffraction (XRD) patterns were collected using a Rigaku IV diffractometer with a Cu Kα radiation source ( λ = 1.5418 Å) over a 2 θ range of 10°–80°. SEM images were collected using a Hitachi SU-70 scanning electron microscope to characterize microstructures. Mechanical testing was performed using an Instron Microtester 5984 universal testing machine according to ISO 37:2024. Dumbbell-shaped specimens (length: 10 mm, thickness: 2 mm) were tested under tensile and compressive loading at a constant strain rate of 1 mm·s − 1 . Rheological properties were measured using an MCR302 rheometer (Anton Paar) with a PP25 plate and a sample thickness of 1 mm, in oscillatory angular frequency sweep mode from 0.1 to 400 rad·s − 1 at an initial strain of 0.5%. DSC curves were collected using a DSC 204 F1 analyzer over the temperature range of − 150°C to 100°C at a heating rate of 10°C·min − 1 . Triboelectric signals were recorded using a Keithley 6514B electrometer connected to a BNC-2110 preamplifier (National Instruments). Multichannel electrical signals were collected using a VK10x-MUL charge amplifier in conjunction with a Smacq USB-3200 data acquisition card. Capacitance signals were collected using a Keithley 7510 digital multimeter. Optical emission spectra were collected using a NOVA2S-EX thermoelectrically cooled back-illuminated array spectrometer (Focuslight Optics) operating across a wavelength range of 325–1100 nm with a slit width of 200 µm. Optical signals were transmitted using a FIB-1000-NIR optical fiber. Declarations Contributions H.W. and Z.Z. conceived and designed the study. H.W., H.X., Y.S., J.Y., Y.Z., J.L., Y.G., and L.L. carried out materials synthesis, device fabrication, and triboelectric/capacitive sensing experiments. H.C. performed the finite-element simulations and electrostatic modeling in COMSOL and contributed to data interpretation. H.W., H.X., Q.Y. and Y.S. conducted the optical waveguide experiments and the risk-classification analysis. Z.Z. analyzed the data and wrote the manuscript with input from all authors. J.S., S.L., and Z.Z. provided supervision and resources. All authors discussed the results, revised the manuscript, and approved the final version. Competing interests The authors declare no competing interests. Funding Declaration This work was supported by the Natural Science Foundation of Fujian Province, China (Grant No. 2022J01044) and the State Key Laboratory of Advanced Nuclear Energy Technology, Nuclear Power Institute of China (Grant No. YNSW-0224-0101-19-01). Author Contribution H.W. and Z.Z. conceived and designed the study. H.W., H.X., Y.S., J.Y., Y.Z., J.L., Y.G., and L.L. carried out materials synthesis, device fabrication, and triboelectric/capacitive sensing experiments. H.C. performed the finite-element simulations and electrostatic modeling in COMSOL and contributed to data interpretation. H.W., H.X., Q.Y. and Y.S. conducted the optical waveguide experiments and the risk-classification analysis. Z.Z. analyzed the data and wrote the manuscript with input from all authors. J.S., S.L., and Z.Z. provided supervision and resources. All authors discussed the results, revised the manuscript, and approved the final version. Data Availability All data generated in this study are provided in the Supplementary Information/Source Data file. Source data are provided with this paper. References Wang, M. et al. Artificial skin perception. Adv. Mater. 33, 2003014 (2021). Yuan, Z. & Shen, G. Materials and device architecture towards a multimodal electronic skin. Mater. Today 64, 165–179 (2023). Zhang, H. et al. Humanoid electronic-skin technology for the era of artificial intelligence of things. Matter 8, 102136 (2025). Zhang, X. et al. Highly stretchable electronic-skin sensors with porous microstructure for efficient multimodal sensing with wearable comfort. Adv. Mater. Interfaces 10, 2201958 (2023). Fu, X., Cheng, W., Wan, G., Yang, Z. & Tee, B. C. K. Toward an AI era: advances in electronic skins. Chem. Rev. 124, 9899–9948 (2024). Xu, C. et al. A physicochemical-sensing electronic skin for stress response monitoring. Nat. Electron. 7, 168–179 (2024). Shin, J. et al. Bioresorbable pressure sensors protected with thermally grown silicon dioxide for the monitoring of chronic diseases and healing processes. Nat. Biomed. Eng. 3, 37–46 (2019). Kim, T. et al. Ultrathin crystalline-silicon-based strain gauges with deep learning algorithms for silent speech interfaces. Nat. Commun. 13, 5815 (2022). Westerveld, W. J. et al. Sensitive, small, broadband and scalable optomechanical ultrasound sensor in silicon photonics. Nat. Photonics 15, 341–345 (2021). Kim, J. et al. Stretchable silicon nanoribbon electronics for skin prosthesis. Nat. Commun. 5, 5747 (2014). Kang, J., Tok, J. B. H. & Bao, Z. Self-healing soft electronics. Nat. Electron. 2, 144–150 (2019). Yang, J. C. et al. Electronic skin: recent progress and future prospects for skin-attachable devices for health monitoring, robotics, and prosthetics. Adv. Mater. 31, 1904765 (2019). Roels, E. et al. Processing of self-healing polymers for soft robotics. Adv. Mater. 34, 2104798 (2022). Bai, H., Kim, Y. S. & Shepherd, R. F. Autonomous self-healing optical sensors for damage intelligent soft-bodied systems. Sci. Adv. 8, eabq2104 (2022). Jung, J. et al. Self-healing electronic skin with high fracture strength and toughness. Nat. Commun. 15, 9763 (2024). Yang, L. et al. Self-Healing, reconfigurable, thermal-switching, transformative electronics for health monitoring. Adv. Mater. 35, 2207742 (2023). Liu, Y. et al. Applications of cellulose-based flexible self-healing sensors for human health monitoring. Nano Energy 127, 109790 (2024). Li, X. et al. Self-healing liquid metal hydrogel for human–computer interaction and infrared camouflage. Mater. Horiz. 10, 2945–2957 (2023). Sun, Y., Li, X., Hu, T., Zhang, S. & Niu, W. Dynamic acylhydrazone bonds cross-linked chromotropic photonic-ionic skin with self-healing and high resilience for interactive human-machine interface. Chem. Eng. J. 475, 146188 (2023). White, S. R. et al. Autonomic healing of polymer composites. Nature 409, 794–797 (2001). Song, F. et al. Tunable “soft and stiff”, self-healing, recyclable, thermadapt shape memory biomass polymers based on multiple hydrogen bonds and dynamic imine bonds. J. Mater. Chem. A 7, 13400–13410 (2019). Tang, M. et al. Autonomous self-healing, self-adhesive, highly conductive composites based on a silver-filled polyborosiloxane/polydimethylsiloxane double-network elastomer. J. Mater. Chem. A 7, 27278–27288 (2019). Chen, Y. et al. Hydrogel-based pressure sensors for electronic skin systems. Matter 8, 101992 (2025). Park, H. et al. Toughening self-healing elastomer crosslinked by metal–ligand coordination through mixed counter anion dynamics. Nat. Commun. 14, 5026 (2023). Chen, J. et al. Ladderphane copolymers for high-temperature capacitive energy storage. Nature 615, 62–66 (2023). Urban, M. W. et al. Key-and-lock commodity self-healing copolymers. Science 362, 220–225 (2018). Zhu, R., Zhu, D., Zheng, Z. & Wang, X. Tough double network hydrogels with rapid self-reinforcement and low hysteresis based on highly entangled networks. Nat. Commun. 15, 1344 (2024). Sun, F. et al. Vascular smooth muscle-inspired architecture enables soft yet tough self-healing materials for durable capacitive strain-sensor. Nat. Commun. 14, 130 (2023). Li, B., Cao, P.-F., Saito, T. & Sokolov, A. P. Intrinsically self-healing polymers: from mechanistic insight to current challenges. Chem. Rev. 123, 701–735 (2023). Li, S. et al. Bimodal capacitive tactile sensor assisted by shield effect of triboelectric nanogenerator. Nano Energy 118, 108946 (2023). Zhang, H., Li, H. & Li, Y. Biomimetic electronic skin for robots aiming at superior dynamic-static perception and material cognition based on triboelectric-piezoresistive effects. Nano Lett. 24, 4002–4011 (2024). Qiu, Y. et al. Quantitative softness and texture bimodal haptic sensors for robotic clinical feature identification and intelligent picking. Sci. Adv. 10, eadp0348 (2024). Ma, X. et al. Bimodal tactile sensor without signal fusion for user-interactive applications. ACS Nano 16, 2789–2797 (2022). Qu, X. et al. Artificial tactile perception smart finger for material identification based on triboelectric sensing. Sci. Adv. 8, eabq2521 (2022). Fang, H. et al. A flexible dual-mode photodetector for human–machine collaborative IR imaging. Nano-Micro Lett. 17, 229 (2025). Edwards, C. & Marks, R. Evaluation of biomechanical properties of human skin. Clin. Dermatol. 13, 375–380 (1995). Chen, J., Yi, D., Shen, B. & Zheng, W. Multifunctional liquid-metal composites for electromagnetic communication and attenuation. Adv. Mater. 37, 2404595 (2025). Guo, R. et al. Ni-GaIn amalgams enabled rapid and customizable fabrication of wearable and wireless healthcare electronics. Adv. Eng. Mater. 20, 1800054 (2018). Yu, L. et al. Transportable, endurable, and recoverable liquid metal powders with mechanical sintering conductivity for flexible electronics and electromagnetic interference shielding. ACS Appl. Mater. Interfaces 14, 48150–48160 (2022). Liu, Y. et al. Triboelectric tactile sensor for pressure and temperature sensing in high-temperature applications. Nat. Commun. 16, 383 (2025). Xia, S. Y. et al. High-performance Cu–Cu interconnects attained through air sintering of oleylamine-capped Cu nanoparticles for power electronics packaging. Rare Metals 44, 3281–3298 (2025). Jeong, H. I. et al. Super elastic and negative triboelectric polymer matrix for high performance mechanoluminescent platforms. Nat. Commun. 16, 854 (2025). Lin, C. et al. Electrically chargeable inorganic persistent luminescence in an alternating current driven electroluminescent device. Commun. Mater. 6, 25 (2025). Peka, P. & Schulz, H. J. Empirical one-electron model of optical transitions in Cu-doped ZnS and CdS. Physica B 193, 57–65 (1994). Wang, X. et al. Dynamic pressure mapping of personalized handwriting by a flexible sensor matrix based on the mechanoluminescence process. Adv. Mater. 27, 2324–2331 (2015). Pan, X. et al. Quantifying the interfacial triboelectricity in inorganic-organic composite mechanoluminescent materials. Nat. Commun. 15, 2673 (2024). Haddadin, S., De Luca, A. & Albu-Schäffer, A. Robot collisions: a survey on detection, isolation, and identification. IEEE T. Robot. 33, 1292–1312 (2017). Haddadin, S., Albu-Schäffer, A. & Hirzinger, G. The role of the robot mass and velocity in physical human-robot interaction—part I: non-constrained blunt impacts. in IEEE Int. Conf. Robot. Autom. 1331–1338 (2008). Zhang, Y., Liu, H., Niu, J., Wang, X. & Wu, D. Development of reversible and durable thermochromic phase-change microcapsules for real-time indication of thermal energy storage and management. Appl. Energ. 264, 114729, (2020). Additional Declarations No competing interests reported. Supplementary Files SupplementaryInformation.docx VideoS1.mp4 VideoS2.mp4 VideoS3.mp4 VideoS4.mp4 VideoS5.mp4 VideoS6.mp4 Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7503393","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":518167932,"identity":"fc824056-4663-4df2-a806-5c70c7abd8b8","order_by":0,"name":"Henghui Wang","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Henghui","middleName":"","lastName":"Wang","suffix":""},{"id":518167933,"identity":"8ca73dff-24de-4034-86e2-3e8b02e62d92","order_by":1,"name":"Hao Xue","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Hao","middleName":"","lastName":"Xue","suffix":""},{"id":518167934,"identity":"17d22384-27f7-4eb4-b7d5-4d4924aea582","order_by":2,"name":"Yapeng Song","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Yapeng","middleName":"","lastName":"Song","suffix":""},{"id":518167935,"identity":"c3ce0b94-9d76-419e-b382-408ae802a215","order_by":3,"name":"Junjie Yuan","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Junjie","middleName":"","lastName":"Yuan","suffix":""},{"id":518167936,"identity":"b8ef0848-8b97-4e62-9a82-87c16985b159","order_by":4,"name":"Huijun Cao","email":"","orcid":"","institution":"Xiamen City University","correspondingAuthor":false,"prefix":"","firstName":"Huijun","middleName":"","lastName":"Cao","suffix":""},{"id":518167937,"identity":"fbc17f8d-f659-40da-8acb-8f03180e0d6a","order_by":5,"name":"Quan Yuan","email":"","orcid":"","institution":"Sungkyunkwan University","correspondingAuthor":false,"prefix":"","firstName":"Quan","middleName":"","lastName":"Yuan","suffix":""},{"id":518167938,"identity":"301ae996-9663-4e8d-8fe2-7979c5b66a95","order_by":6,"name":"Yichen Zhu","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Yichen","middleName":"","lastName":"Zhu","suffix":""},{"id":518167939,"identity":"4cb5eaab-610f-4de3-858b-5bf30be29c73","order_by":7,"name":"Jonghwan Suhr","email":"","orcid":"","institution":"Sungkyunkwan University","correspondingAuthor":false,"prefix":"","firstName":"Jonghwan","middleName":"","lastName":"Suhr","suffix":""},{"id":518167940,"identity":"94b9e214-64b7-48f0-a363-1d98dbace223","order_by":8,"name":"Jintang Li","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Jintang","middleName":"","lastName":"Li","suffix":""},{"id":518167941,"identity":"207f1b60-1b71-43d9-ac7b-343d6ee19f2d","order_by":9,"name":"Siwei Li","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Siwei","middleName":"","lastName":"Li","suffix":""},{"id":518167942,"identity":"2c4aae1f-61b0-4c6e-a898-7ae03e696e03","order_by":10,"name":"Yingjie Guo","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Yingjie","middleName":"","lastName":"Guo","suffix":""},{"id":518167943,"identity":"f45db1ca-6436-4574-b73e-abd6405a910b","order_by":11,"name":"Liyun Li","email":"","orcid":"","institution":"Xiamen University","correspondingAuthor":false,"prefix":"","firstName":"Liyun","middleName":"","lastName":"Li","suffix":""},{"id":518167944,"identity":"49d27c4b-5b6c-4d39-9bea-579a933f833a","order_by":12,"name":"Zhihao Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0ElEQVRIiWNgGAWjYDACCQY2ECUH4bGRoMWYdC2JDURrkZ/d/uzBzx216RuunTFg+FB2mIF/dgN+LYxzDqQb9p45njtzdo4B44xzhxkk7hzAr4VZIuGYBG/bsdx+6RwDZt62wwwGEgn4tbBJJLZJ/m07ls4G0vKXGC08Esls0rxtNQn8IC2MxGiRkEhjk5ZtO2A4c3ZawcGec+k8EjcIaJGfkf5M8m1bnbzB7eSND36UWcvxzyCgBQoOg8kDIJcSpR4I6ohVOApGwSgYBSMRAADpGD4uq8kP6QAAAABJRU5ErkJggg==","orcid":"","institution":"Xiamen University","correspondingAuthor":true,"prefix":"","firstName":"Zhihao","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2025-09-01 01:53:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7503393/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7503393/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":91985125,"identity":"13d2e4de-2e93-4b45-82c6-ce0588e9781f","added_by":"auto","created_at":"2025-09-23 11:48:51","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":9347501,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBioinspired design of a self-healing, quadruple-modal e-skin.\u003c/strong\u003e The compact five-layer integrated architecture comprises: optical waveguide temperature sensing units mimicking free nerve endings (Layer 1 and Layer 3); triboelectric tactile sensing units mimicking Meissner’s corpuscles (Layers 1 and 2); capacitive pressure sensing units mimicking Merkel disks and Ruffini endings (Layers 2–4); and optical waveguide nociceptive sensing units mimicking free nerve endings (Layer 5).\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/9411f00364f157bcc0cf5d3b.png"},{"id":91985127,"identity":"9bae9ecf-2079-4d2e-9b79-f5c58134dc15","added_by":"auto","created_at":"2025-09-23 11:48:52","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":13744546,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMaterial design and characterization of DCP variants. \u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Schematic illustration of the one-pot synthesis of supramolecular DCP for self-healing e-skin, using PMHS, VPDMS, BA, and PDMS-OH as precursors. Key self-healable features are highlighted, including hydrogen bonding, dative (B-O) bonding, and topological entanglement. (\u003cstrong\u003eb\u003c/strong\u003e) FTIR spectra (500-4000 cm\u003csup\u003e-1\u003c/sup\u003e). (\u003cstrong\u003ec\u003c/strong\u003e) Optical images showing structural stability over 0-40 min. (\u003cstrong\u003ed\u003c/strong\u003e) Time-dependent optical images of scratch healing behavior (0-180 min). (\u003cstrong\u003ee\u003c/strong\u003e) Tensile stress-strain curves. (\u003cstrong\u003ef\u003c/strong\u003e) Compressive cyclic stress-strain curves over 10 cycles at 55% strain. (\u003cstrong\u003eg\u003c/strong\u003e) Damping factor (tan\u003cem\u003eδ\u003c/em\u003e) curves from dynamic frequency sweeps (0.1-130 Hz). The tan\u003cem\u003eδ\u003c/em\u003e was represented as a ratio of loss modulus (G′′) and storage modulus (G′). (\u003cstrong\u003eh\u003c/strong\u003e) DSC curves from -150 °C to 250 °C.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/cfe3b250fe5e71276b24311b.png"},{"id":91986201,"identity":"c907f957-03f6-49ed-9e38-47cc91229e29","added_by":"auto","created_at":"2025-09-23 11:56:52","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":19312067,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSelf-healing capability evaluation and integrated fabrication. \u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Schematic illustration of the environmental adaptability of DCP-3. (\u003cstrong\u003eb–d\u003c/strong\u003e) Tensile stress-strain curves and corresponding bar charts of self-healing efficiency versus healing time for DCP-3 at –10 °C, 25 °C, and 60 °C, respectively. (\u003cstrong\u003ee\u003c/strong\u003e) LED demonstration of a self-healing circuit constructed with ELM and DCP-3. The circuit was fabricated using ELM doped with NPs. Prior to patterning, UV/ozone treatment was applied to the uncovered polymer surface through a shadow mask to enhance adhesion between ELM and DCP-3. (\u003cstrong\u003ef\u003c/strong\u003e) Schematic illustration of the layer-by-layer fabrication of e-skin enabled by self-healing DCP-3. The top and middle polymer layers were doped with TPs (red or blue), while the bottom polymer layer contained MPs. (\u003cstrong\u003eg\u003c/strong\u003e) Top view of the fabricated e-skin. (\u003cstrong\u003eh\u003c/strong\u003e) Cross-sectional view after diagonal cutting and large-angle bending. Regions A-C indicate the TP-doped DCP-3 layer, the NP-doped ELM layer, and the MP-doped polymer layer, respectively. (\u003cstrong\u003ei\u003c/strong\u003e) Top view of the e-skin after diagonal cutting and re-healing.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/4377078ed5d9f09425a8448d.png"},{"id":91986200,"identity":"6a1df752-aa05-4bad-8982-cb1ce8e8cc8d","added_by":"auto","created_at":"2025-09-23 11:56:51","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":6945078,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTactile sensing and applications. \u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Working mechanism of triboelectric sensing based on the single-electrode vertical contact-separation mode. (\u003cstrong\u003eb\u003c/strong\u003e) Real-time open-circuit voltage waveforms under constant contact pressures ranging from 0.2 to 16 kPa at 1 Hz. (\u003cstrong\u003ec\u003c/strong\u003e) Sensing sensitivity (\u003cem\u003eS = \u003c/em\u003ed\u003cem\u003eV\u003c/em\u003e/d\u003cem\u003eP\u003c/em\u003e, where \u003cem\u003eV\u003c/em\u003e is the maximum open-circuit voltage and \u003cem\u003eP \u003c/em\u003eis the applied pressure). (\u003cstrong\u003ed\u003c/strong\u003e)\u003cstrong\u003e \u003c/strong\u003eWaveform stability tests of open-circuit voltage at pressures of 0.2, 1.2, and 4 kPa. (\u003cstrong\u003ee\u003c/strong\u003e)\u003cstrong\u003e \u003c/strong\u003eWaveforms at the early, middle, and late stages. (\u003cstrong\u003ef\u003c/strong\u003e) Response times (\u003cstrong\u003eg\u003c/strong\u003e) Muscle movement detection during fist clenching and corresponding waveforms. (\u003cstrong\u003eh\u003c/strong\u003e) Gait frequency and arch loading detection and corresponding waveforms. (\u003cstrong\u003ei\u003c/strong\u003e) Spatial tactile trajectory tracking and corresponding time-correlated waveforms.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/df82e04cce7f36334e53d2a2.png"},{"id":91985128,"identity":"cbccf2d3-b998-4e2d-997b-e775e5384f7d","added_by":"auto","created_at":"2025-09-23 11:48:52","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":12875525,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePressure sensing and applications. \u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Working mechanism of capacitive sensing based on the parallel-plate capacitive configuration. (\u003cstrong\u003eb\u003c/strong\u003e) Bar chart of maximum relative capacitance variation under constant contact pressures ranging from 1 to 40 kPa at 1 Hz. (\u003cstrong\u003ec\u003c/strong\u003e) Sensing sensitivity (\u003cem\u003eS = \u003c/em\u003eΔ\u003cem\u003eC\u003c/em\u003e/(C\u003csub\u003e0\u003c/sub\u003e·\u003cem\u003eP\u003c/em\u003e), where Δ\u003cem\u003eC\u003c/em\u003e is the capacitance variation, \u003cem\u003eC\u003c/em\u003e\u003csub\u003e0\u003c/sub\u003e is the initial capacitance, and \u003cem\u003eP \u003c/em\u003eis the applied pressure). (\u003cstrong\u003ed\u003c/strong\u003e)\u003cstrong\u003e \u003c/strong\u003eWaveform stability tests of maximum relative capacitance variation at pressures of 4, 16, and 28 kPa. (\u003cstrong\u003ee\u003c/strong\u003e)\u003cstrong\u003e \u003c/strong\u003eWaveforms at the early, middle, and late stages. (\u003cstrong\u003ef\u003c/strong\u003e) Response times (\u003cstrong\u003eg\u003c/strong\u003e) Photographs and corresponding feedback diagrams of the diagonally healed pressure sensing array upon loading with 100 g weights via “X”, “M”, and “U”-shaped plastic sheets.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/37d9d5344d7cd2309fab6f17.png"},{"id":91986577,"identity":"f80672ad-dc35-4a22-b8d7-f221e8b359ae","added_by":"auto","created_at":"2025-09-23 12:04:52","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":7919812,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNociceptive and thermal sensing and applications. \u003c/strong\u003e(\u003cstrong\u003ea\u003c/strong\u003e) Mechanoluminescence mechanism of DCP-3 matrix embedded with ZnS:Cu MPs under applied stress. (\u003cstrong\u003eb\u003c/strong\u003e) Finite element simulation of triboelectric potential and field distributions induced by interfacial strain mismatch (Fig. S7 and Table S2). (\u003cstrong\u003ec\u003c/strong\u003e) Schematic illustration of an interfacial delamination zone. (\u003cstrong\u003ed\u003c/strong\u003e)\u003cstrong\u003e \u003c/strong\u003eRelationships between triboelectric field, air-gap distance, and matrix strain. (\u003cstrong\u003ee\u003c/strong\u003e)\u003cstrong\u003e \u003c/strong\u003eCorrelation between energy flux density and mechanoluminescent intensity. (\u003cstrong\u003ef\u003c/strong\u003e) Sensing sensitivity map and classification of collision risk levels. (\u003cstrong\u003eg\u003c/strong\u003e) Emission spectra under representative energy flux densities (Fig. S8). (\u003cstrong\u003eh\u003c/strong\u003e) Absorption spectra of the thermochromic layer under various heat source temperatures. (\u003cstrong\u003ei\u003c/strong\u003e) Time evolution of normalized absorption intensity under different heat source temperatures. (\u003cstrong\u003ej\u003c/strong\u003e) Sensing sensitivity map and classification of thermal injury risks.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/07e14cd5b02f1fbcf42253cc.png"},{"id":91988096,"identity":"d8dfcceb-d649-419a-a3d4-d8322ad0216e","added_by":"auto","created_at":"2025-09-23 12:13:23","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":66148355,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/abee0f39-9f8c-4cfd-be3c-61d768e96690.pdf"},{"id":91985133,"identity":"0bc7b32f-481d-4eff-917a-be714193b412","added_by":"auto","created_at":"2025-09-23 11:48:52","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":9977543,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/b54282473101d42ee30c76fe.docx"},{"id":91985132,"identity":"1d30b0d3-cde9-474b-a27c-8b4502cee572","added_by":"auto","created_at":"2025-09-23 11:48:52","extension":"mp4","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":18576310,"visible":true,"origin":"","legend":"","description":"","filename":"VideoS1.mp4","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/b041213241e4fb74448b8788.mp4"},{"id":91985134,"identity":"0cfd47a6-26fd-4aff-a927-f3ace68b7ba2","added_by":"auto","created_at":"2025-09-23 11:48:52","extension":"mp4","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":19399900,"visible":true,"origin":"","legend":"","description":"","filename":"VideoS2.mp4","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/63ad4c73d4b9c07ce4566747.mp4"},{"id":91985138,"identity":"60c5b398-a8fc-4595-8cb8-4b0db0a2e553","added_by":"auto","created_at":"2025-09-23 11:48:56","extension":"mp4","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":92322051,"visible":true,"origin":"","legend":"","description":"","filename":"VideoS3.mp4","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/046c2941d46a394257a4b6c5.mp4"},{"id":91985135,"identity":"0fa458a9-477a-477b-a534-76172fe56198","added_by":"auto","created_at":"2025-09-23 11:48:53","extension":"mp4","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":29468246,"visible":true,"origin":"","legend":"","description":"","filename":"VideoS4.mp4","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/3d72efdc57001c3a7450d95d.mp4"},{"id":91985136,"identity":"736446e7-4763-4e93-80b6-78b10bca3378","added_by":"auto","created_at":"2025-09-23 11:48:54","extension":"mp4","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":35042048,"visible":true,"origin":"","legend":"","description":"","filename":"VideoS5.mp4","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/ea74985d28dcff88a292cf3e.mp4"},{"id":91985137,"identity":"e5cd682e-f94c-44e0-80bc-3baaa84d31ef","added_by":"auto","created_at":"2025-09-23 11:48:55","extension":"mp4","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":61094802,"visible":true,"origin":"","legend":"","description":"","filename":"VideoS6.mp4","url":"https://assets-eu.researchsquare.com/files/rs-7503393/v1/b3a2ad6b43d3e0625b71a698.mp4"}],"financialInterests":"No competing interests reported.","formattedTitle":"A self-healing e-skin for quadruple-modal sensing","fulltext":[{"header":"Introduction","content":"\u003cp\u003eSkin, the largest peripheral sensory organ in biological systems, exhibits complex sensory functionalities mediated by an array of cutaneous receptors\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. These include free nerve endings, Meissner's corpuscles, Ruffini endings, and Merkel disks, which collectively enable multimodal perception of thermal, nociceptive, tactile, and pressure stimuli. Inspired by these biological sensing modalities, the development of e-skin with integrated multimodal sensing capabilities has emerged as a critical frontier in the research of artificial sensory systems\u003csup\u003e\u003cspan additionalcitationids=\"CR3 CR4 CR5\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Rigid or semi-rigid materials are typically employed in sensor construction to ensure signal stability, combined with structural engineering to enhance sensitivity and expand detection range\u003csup\u003e\u003cspan additionalcitationids=\"CR8\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Nevertheless, mechanical incompatibility with biological tissues\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e, along with wear and accidental damage during operation\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e, remains a primary cause of device failure. These challenges underscore an urgent need for e-skin possessing both flexibility and intrinsic self-healing capabilities.\u003c/p\u003e\u003cp\u003eTo bridge the functional gap between e-skin and natural skin, flexible, self-healing polymers with varied electrical properties (insulators, semiconductors, and electronic/ionic conductors\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e) have been employed in the construction of e-skins for applications in prosthetics\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e, robotics\u003csup\u003e\u003cspan additionalcitationids=\"CR14\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e, health monitoring\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, and human-machine interfaces\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. The self-healing mechanisms of these polymers primarily involve: (i) extrinsic repair, in which embedded microcapsules rupture to release healing agents\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e; and (ii) intrinsic repair, which relies on the reconstruction of dynamic covalent bonds (e.g., disulfide\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e, imine\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e, and boronic ester\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e bonds), supramolecular interactions (e.g., hydrogen bonds\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, metal-ligand coordination\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e, π-π stacking\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e, and van der Waals forces\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e), and topological entanglement\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e within supramolecular networks. Ideally, e-skin should combine efficient self-healing with sufficient toughness to ensure robust sensing performance. However, a trade-off exists between healing efficiency and bond strength\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. To address this limitation, recent research has focused on molecular design to develop polymers that integrate multiple self-healing mechanisms\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. Despite these advances at the material level, the system-level integration of diverse sensing units, comprising distinct materials\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e and architectures\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, remains a major obstacle, impeding the development of e-skins capable of reliable multisignal acquisition while retaining flexibility and self-healing functionality.\u003c/p\u003e\u003cp\u003eAddressing system-level integration challenges requires not only innovations in materials but also advances in sensing architecture. Unimodal configurations often entail trade-offs among functionality, sensitivity, and detection range, while multimodal strategies may offer a more promising route toward comprehensive and high-fidelity perception. Thus far, progress in e-skin development has largely centred on bimodal sensing architectures, leading to configurations such as triboelectric-capacitive tactile sensors\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, triboelectric-piezoresistive tactile sensors\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, piezoelectric-piezoresistive tactile sensors\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, and light-resistance-based tactile-temperature sensors\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Despite significant progress in compact integration and signal decoupling, these architectures remain limited in sensory diversity and largely overlook self-healing functionality. Notably, triboelectric sensing can mimic the function of Meissner's corpuscles, enabling the detection of surface contact and frictional stimuli\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, as well as material discrimination\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. Capacitive sensing may resemble the roles of Merkel discs or Ruffini endings in the dermis, allowing for sustained pressure detection under moderate strain\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, and when configured as arrays, spatial resolution\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. Optical waveguide sensing can emulate the function of free nerve endings by capturing light spectrum variations induced by external stimuli\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. When doped with thermochromic or mechanoluminescent particles, it enables the perception of thermal\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e or nociceptive\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e stimuli under large strains. Given the functional complementarity among triboelectric, capacitive, and optical waveguide sensing modalities, system-level integration may offer a promising strategy for developing a quadruple-modal e-skins capable of simultaneously sensing tactile, pressure, nociceptive, and thermal stimuli.\u003c/p\u003e\u003cp\u003eHere, we present a self-healing e-skin featuring a vertically stacked, five-layer architecture (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The 1st, 3rd, and 5th layers serve as insulated layers composed of a double-network cross-linked polymer (DCP) formed from polydimethylsiloxane (PDMS) and polyborosiloxane (PBS). Red or blue thermochromic particles (TPs) are embedded in the 1st and 3rd layers, while ZnS:Cu mechanoluminescent particles (MPs) are incorporated into the 5th layer. The 2nd and 4th layers consist of identical 3 \u0026times; 3 electrode arrays, fabricated by screen-printing eutectic Ga-In-Sn liquid metal (ELM) doped with Ni particles (NPs) onto the underlying insulating layers. Functionally, the 1st layer and each electrode in the 2nd layer forms a single-electrode triboelectric unit for tactile sensing; each pair of parallel electrodes between the 2nd and 4th layers forms a parallel-plate capacitive unit for pressure sensing; and the 1st and 3rd layers, and the 5th layer, independently serve as optical waveguide units for temperature and nociceptive sensing, respectively. By integrating triboelectric, capacitive, and optical waveguide sensing modalities into a unified platform, our e-skin can achieve compact quadruple-modal perception (touch, pressure, nociception and temperature) alongside self-healing, high sensitivity, spatiotemporal resolution, and scalable fabrication. These features mark a critical step toward the practical implementation of e-skin technologies.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n\u003ch2\u003eMaterial design and characterization\u003c/h2\u003e\n\u003cp\u003eTo enable a self-healing e-skin with quadruple-modal sensing functionality, the polymer matrix must rapidly self-repair under harsh conditions, maintain structural integrity and optical transparency, and exhibit mechanical responsiveness akin to natural skin. Both PDMS and PBS are silicone-based polymers. PDMS exhibits excellent toughness and optical transparency but lacks self-healing functionality, whereas PBS is viscoelastic, transparent, and self-repairable. To integrate these complementary properties, we synthesized the supramolecular DCP via one-pot strategy comprising both PDMS and PBS chains (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ea). The PDMS chains were formed through hydrosilylation between polymethylhydrogensiloxane (PMHS) and dimethylvinyl-terminated dimethylsiloxane (VPDMS), while the PBS chains were generated by esterification between boric acid (BA) and hydroxyl-terminated polydimethylsiloxane (PDMS-OH). In this supramolecular network, rigid PDMS chains provide mechanical strength, while flexible PBS chains impart self-healing via B-O dative bonds, hydrogen bonds, and topological entanglements.\u003c/p\u003e\n\u003cp\u003eFour DCP variants with PDMS:PBS weight ratios ranging from 1:1 to 1:4 (denoted as DCP-1 to DCP-4) were synthesized to optimize overall performance. A characteristic absorption band at 1340 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e in the Fourier-transform infrared (FTIR) spectra (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eb), attributed to the B-O stretching vibration in the Si-O-B bond\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e, confirms the successful formation of PBS chains. The absence of Si-H stretching vibration at 2160 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e indicates the complete consumption of Si-H groups in the original VPDMS (Fig. \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e), confirming the formation of PDMS chains. Stability tests (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ec) reveal that PBS undergoes macroscopic collapse within 40 min due to its fluidic nature, whereas all DCP variants, reinforced by rigid PDMS chains, maintain PDMS-like structural stability. Interestingly, only DCP-3 and DCP-4 fully healed within 120 min (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ed), while DCP-1 and DCP-2 retained visible scratches even after 180 min, indicating a discernible difference in healing efficiency.\u003c/p\u003e\n\u003cp\u003eHuman skin typically ruptures at a strain of 35\u0026ndash;115%\u003csup\u003e36\u003c/sup\u003e, whereas all DCP variants exceed this range, exhibiting elongations at break between 150% and 418% (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ee). Among them, DCP-3 displays linear elasticity across a broad strain range of 10\u0026ndash;150%, which is essential for ensuring consistent sensing performance under varying mechanical loads. Moreover, cyclic compression tests (55% strain, ~\u0026thinsp;1 Hz) revealed that all DCP variants exhibited overlapping hysteresis loops but non-overlapping loading-unloading curves over 10 cycles, indicative of viscoelastic behavior (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ef). However, DCP-4 exhibited the lowest compressive modulus and the most pronounced hysteresis under high strain, potentially resulting in a narrower detection threshold and delayed signal responses.\u003c/p\u003e\n\u003cp\u003eNotably, rheological tests showed that all DCP variants exhibited the damping factors (tan\u003cem\u003e\u0026delta;\u003c/em\u003e) below 1 across the full angular frequency range of 0.1\u0026ndash;400 rad\u0026middot;s\u003csup\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e, reflecting a predominantly elastic response (Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eg and S2). Especially, DCP-3 displayed the most stable tan\u003cem\u003e\u0026delta;\u003c/em\u003e profile with minimal fluctuations, likely due to a well-balanced composition of PDMS and PBS chains. In addition, differential scanning calorimetry (DSC) analysis demonstrated that all DCP variants possess a glass transition temperature (\u003cem\u003eT\u003c/em\u003e\u003csub\u003eg\u003c/sub\u003e) of approximately \u0026minus;\u0026thinsp;120\u0026deg;C, a melting temperature (\u003cem\u003eT\u003c/em\u003e\u003csub\u003em\u003c/sub\u003e) of around \u0026minus;\u0026thinsp;40\u0026deg;C, and a degradation temperature (\u003cem\u003eT\u003c/em\u003e\u003csub\u003ed\u003c/sub\u003e) exceeding 250\u0026deg;C (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eh and Table \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e), indicating a wide thermal operating window. Taken together, the balance between self-healing efficiency and mechanical response positions DCP-3 as a promising polymer matrix for self-healing e-skin applications.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv class=\"Heading\"\u003e\u003cstrong\u003eEnvironmental adaptability and integration\u003c/strong\u003e\u003c/div\u003e\n\u003cp\u003eEarth\u0026rsquo;s diverse climates, ranging from polar glaciers to equatorial deserts, impose stringent demands on the environmental adaptability of e-skins (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea), where self-healing conditions and efficiency are essential. A notable advantage of DCP-3 lies in its ability to autonomously self-heal without external stimuli such as heat\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e or light\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. The self-healing efficiency of its tensile strength reaches 84.4% after 48 h at -10\u0026deg;C (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb), 84.2% within 24 h at 25\u0026deg;C (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ec), and up to 97.0% within 60 min at 60\u0026deg;C (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ed), demonstrating excellent self-healing performance across a wide thermal range. To assess functional compatibility, MPs (20 wt%) and TPs (1 wt%) were also individually incorporated into DCP-3. The resulting composites retained nominal stress-strain profiles similar to the original matrix (Fig. \u003cspan class=\"InternalRef\"\u003eS3\u003c/span\u003e). Especially after 24 hours of healing at 25\u0026deg;C, the self-healing efficiencies of tensile strength remained at 76.3% for MP-doped composite and 83.4% for TP-doped composite (Fig. \u003cspan class=\"InternalRef\"\u003eS4\u003c/span\u003e), indicating that functionalization did not greatly compromise the mechanical response or healing capacity.\u003c/p\u003e\n\u003cp\u003eELM exhibits a low melting point, high conductivity, and good fluidity, and is widely used for fabricating self-healing circuits\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. Particularly, doping with NPs can promote Ga\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e formation to improve wettability with polymer matrix\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e, and also may facilitate magnetic guidance for circuit repair. Meanwhile, we performed UV/ozone treatment to the DCP-3 surface, which forms stable -Ga-O-Si- bonds\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e with ELM and significantly enhances interfacial adhesion. Combined with shadow mask patterning, this approach can greatly improve the patterning resolution of NP-doped ELM onto DCP-3. As shown in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ee, two NP-doped ELM circuits were patterned on DCP-3 surfaces doped with red and blue TPs, respectively, and their functionality was demonstrated by LED illumination. Precise alignment and magnetic assistance enabled the fractured red and blue segments to rapidly reconnect and restore stable conductivity, thereby lighting the LED. Simple stretching, bending, and twisting of the healed structure did not affect LED illumination, confirming the robust self-healing capability of both the polymer matrix and the circuit.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAdditionally, a more complex 3 \u0026times; 3 electrode array was patterned on the surface of DCP-3 using a shallow mask (Fig. \u003cspan class=\"InternalRef\"\u003eS5\u003c/span\u003e). The uncovered polymer surface, which retains B-O dative bonds and multiple hydrogen bonds, can provide inherent self-healing capability that facilitates layer-by-layer assembly (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ef). The resulting five-layer integrated e-skin exhibits a stable structure without interfacial delamination, particle shedding, or ELM leakage even under large bending deformations, demonstrating excellent structural robustness (Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eg and \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eh). Finally, two e-skins were diagonally incised; then, through the combination of precise polymer layer alignment and magnetically guided ELM circuit reconnection, we achieved, for the first time, large-area structural and functional re-healing of a fully damaged e-skin (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ei), marking a significant breakthrough in self-healing e-skin technology.\u003c/p\u003e\n\u003ch3\u003eTriboelectric tactile sensing and applications\u003c/h3\u003e\n\u003cp\u003eWe further assessed the post-healing functional integrity of the diagonally incised e-skin. Benefiting from the compact stacked design, the 3 \u0026times; 3 tactile sensing array consists of nine independent single-electrode triboelectric sensing units that operate in a vertical contact-separation mode. During human-machine interaction, triboelectric signals are generated through a combination of triboelectrification and electrostatic induction\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. To be specific (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ea), when a finger (the electropositive material) contacts the DCP-3 layer (the electronegative material), a contact potential difference forms at the interface due to their different work functions, driving electron transfer to the polymer surface; upon separation, the increasing gap enhances the polarization field, causing electrons to flow from the ELM electrode to the reference electrode until charge equilibrium is achieved at maximum separation; as the finger approaches again, the field weakens, inducing a reverse electron flow; and once full contact is restored, the charges balance again, completing the cycle.\u003c/p\u003e\n\u003cp\u003eTo evaluate adaptability to tactile pressures, real-time open-circuit voltage waveforms were recorded under constant contact pressures ranging from 0.2 to 16 kPa at 1 Hz (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eb), and the correlation between the maximum open-circuit voltage and contact pressure was plotted (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ec). It was observed that the open-circuit voltage increased nearly linearly with the applied pressure. When the pressure ranged from 0.2 to 4 kPa, the sensitivity of the triboelectric signal was 0.98 V\u0026middot;kPa\u003csup\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e, whereas for pressures ranged from 4 to 16 kPa, the sensitivity decreased to 0.083 V\u0026middot;kPa\u003csup\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. The higher sensitivity at lower pressures can be attributed to an increased contact area on the DCP-3 layer, which facilitates greater charge transfer during contact. However, once the contact area reaches its limit, higher pressure is needed to induce electron cloud overlaps at the polymer-skin interface\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e, which reduces charge transfer efficiency and thus limits sensitivity. To further verify signal stability, over 500 repeated clicking cycles were conducted at pressures of 0.2, 1.2, and 4 kPa (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ed). At all three pressure levels, the output waveforms remained consistent throughout the early, middle, and late stages, with amplitude retention exceeding 99% over the entire test duration (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ee). The signals show similar waveforms under different pressures, with response times ranging from 100\u0026ndash;125 ms during pressure application and 98\u0026ndash;120 ms during pressure release, which are faster than human response time to tactile stimuli (139 ms)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eThe high-sensitivity contact-separation triboelectric signals endow the e-skin with versatile tactile sensing capabilities. When attached to the wrist, it can monitor subtle muscle movements. During fist clenching (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eg), muscle contraction causes separation between the e-skin and the wrist skin, creating instantaneous open-circuit voltage peaks whose magnitude correlates with the gripping force (Video S1). Similarly, swallowing movements can be readily detected (Video S2). Moreover, when attached to the foot arch, the e-skin can monitor gait frequency and arch loading conditions. During walking, jogging, and running (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eh), changes in gait frequency alter the contact-separation rate between the arch skin and the e-skin, producing signals with distinct frequencies. Meanwhile, because the tester does not exhibit foot inversion or eversion, the magnitude of the open-circuit voltage remains stable. Furthermore, the 3 \u0026times; 3 tactile sensing array endows the e-skin with spatial tactile mapping capability. When a tester sequentially taps the electrodes in the order of 1\u0026rarr;5\u0026rarr;6\u0026rarr;3\u0026rarr;2\u0026rarr;4 (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ei), the multi-channel data acquisition unit records time-correlated waveforms without crosstalk, demonstrating excellent tactile trajectory tracking performance. Additionally, under simulated conditions where multiple tactile stimuli with varying pressures and frequencies are applied simultaneously (Videos S3 and S4), the e-skin exhibits strong anti-interference capability and stable signal recognition, facilitating high-precision measurements in complex practical scenarios.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003ch3\u003eCapacitive pressure sensing and applications\u003c/h3\u003e\n\u003cp\u003eThe 3 \u0026times; 3 pressure sensing array consists of nine independent parallel-plate capacitive sensing units. For each unit, the upper ELM electrode is shared with the triboelectric sensing unit, while the bottom ELM electrode is independent, with the intermediate DCP-3 layer acting as the dielectric. During human-machine interaction (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ea), finger pressure compresses the DCP-3 layer, decreasing its thickness (\u003cem\u003ed\u003c/em\u003e) and increasing the capacitance (\u003cem\u003eC\u003c/em\u003e); at maximum pressure, the polymer reaches its minimum thickness (\u003cem\u003ed\u003c/em\u003e\u003csub\u003emin\u003c/sub\u003e) and maximum capacitance (\u003cem\u003eC\u003c/em\u003e\u003csub\u003emax\u003c/sub\u003e); upon release, the polymer gradually recovers, and the capacitance decreases accordingly; once contact separation occurs, both the thickness (\u003cem\u003ed\u003c/em\u003e\u003csub\u003e0\u003c/sub\u003e) and capacitance (\u003cem\u003eC\u003c/em\u003e\u003csub\u003e0\u003c/sub\u003e) return to their initial values. This process enables the establishment of a pressure-capacitance relationship through the relative capacitance variation (\u0026Delta;\u003cem\u003eC\u003c/em\u003e/\u003cem\u003eC\u003c/em\u003e\u003csub\u003e0\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;\u003cem\u003eC\u003c/em\u003e/\u003cem\u003eC\u003c/em\u003e\u003csub\u003e0\u003c/sub\u003e\u0026ndash;1)\u003csup\u003e41\u003c/sup\u003e. Notably, the minimal vertically stacked structure facilitates self-healing of the e-skin; however, the triboelectric effect may interfere with capacitive signals\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. To address this issue, a programmable frequency-sweeping strategy is suggested to dynamically coordinate the switching of each triboelectric and capacitive sensing unit based on the detected stimulus intensity, thereby enabling highly sensitive and spatially resolved tactile and pressure sensing.\u003c/p\u003e\n\u003cp\u003eThe peak values of relative capacitance variation under pressures ranging from 1 to 40 kPa at 1 Hz were extracted from temporal waveforms and displayed as bar plots (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eb). A corresponding sensitivity curve was derived by plotting the maximum relative capacitance variation (\u0026Delta;\u003cem\u003eC\u003c/em\u003e/\u003cem\u003eC\u003c/em\u003e\u003csub\u003e0\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;\u003cem\u003eC\u003c/em\u003e\u003csub\u003emax\u003c/sub\u003e/\u003cem\u003eC\u003c/em\u003e\u003csub\u003e0\u003c/sub\u003e\u0026ndash;1) as a function of pressure (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ec). Data fitting revealed a linear sensitivity of 0.108 kPa\u003csup\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e within the 3\u0026ndash;28 kPa range, while sensitivity declined markedly outside this window. This response may arise from distinct deformation mechanisms of the DCP-3 layer under varying pressure regimes (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ee). At moderate pressures, elastic deformation dominates, yielding a linear pressure-thickness relationship. At low pressures, polymer chain slippage induces viscous flow and nonlinear thickness variation. At high pressures, chain rearrangement may restrict compressibility, while intensified electrode deformation enhances edge-field effects, leading to increased parasitic capacitance and a reduction in effective signal output. In addition, the upper pressure limit of 28 kPa corresponds to lifting a weight of approximately 100\u0026ndash;150 pounds with both hands, basically fulfilling the practical requirements of human-machine interaction.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSignal stability was assessed under pressures of 4, 16, and 28 kPa through more than 500 repeated clicking cycles (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ed). Similar waveforms were maintained at all three pressures across the early, middle, and late stages (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ee). Response times ranged from 264 to 292 ms during pressure application and from 244 to 270 ms during release across all pressure levels (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003ef). Note that the relaxation time decreased with increasing pressure, suggesting that the DCP-3 layer, as a viscoelastic polymer, operated within a hyperelastic regime in the 3\u0026ndash;28 kPa range. By leveraging the complementary characteristics of both triboelectric and capacitive sensing units, the e-skin can achieve an extended pressure sensing range spanning from 0.2 to 28 kPa. Additionally, spatial pressure distribution of the capacitive sensing array was explored by placing a 100 g weight on the surface of the diagonally healed e-skin through plastic sheets shaped as \u0026ldquo;X\u0026rdquo;, \u0026ldquo;M\u0026rdquo;, or \u0026ldquo;U\u0026rdquo; (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eg). When using the \u0026ldquo;X\u0026rdquo;-shaped plate, five sensing units were subjected to the weight, whereas in the \u0026ldquo;M\u0026rdquo; and \u0026ldquo;U\u0026rdquo; configurations, seven units were loaded. As a result, the average pressure per unit was higher in the \u0026ldquo;X\u0026rdquo; configuration, leading to larger relative capacitance variation compared to the other two cases. It is also noted that slight differences in the electrode coverage areas across the \u0026ldquo;X\u0026rdquo;, \u0026ldquo;M\u0026rdquo;, and \u0026ldquo;U\u0026rdquo;-shape plates resulted in minor variations in pressure distribution across the capacitive sensing array. The capacitive pressure sensing array and triboelectric tactile sensing array in our self-healing e-skin exhibit functional complementarity. Their synergistic operation enables more accurate responses to mechanical stimuli and provides comprehensive signal feedback for huma-machine interaction.\u003c/p\u003e\n\u003ch3\u003eOptical-waveguide nociceptive sensing and applications\u003c/h3\u003e\n\u003cp\u003eInspired by the nociceptive and thermosensitive functions of free nerve endings, MPs and TPs were respectively integrated into the DCP-3 matrix to enable pain and temperature sensing, thereby greatly enhancing sensitivity and expanding the detectable stimulus range. ZnS:Cu, a representative mechanoluminescent material (Fig. \u003cspan class=\"InternalRef\"\u003eS6\u003c/span\u003ea), exhibits bright and reproducible green emission (~\u0026thinsp;520 nm), primarily attributed to Cu\u003csup\u003e2+\u003c/sup\u003e luminescent centers\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e42\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. The underlying mechanism involves (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003ea): (i) pressure-induced piezoelectric polarization in the non-centrosymmetric wurtzite ZnS lattice (Fig. \u003cspan class=\"InternalRef\"\u003eS6\u003c/span\u003eb); (ii) the resulting piezopotential tilts the conduction and valence bands, releasing electrons from shallow donor traps into the conduction band; (iii) these detrapped electrons migrate to Cu\u003csup\u003e2+\u003c/sup\u003e centers, where non-radiative recombination with holes promotes of Cu\u003csup\u003e2+\u003c/sup\u003e outer electrons from the \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003eA\u003csub\u003e1\u003c/sub\u003e ground state to the metastable \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003eT\u003csub\u003e1\u003c/sub\u003e state; (iv) subsequent spin-forbidden radiative relaxation from \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003eT\u003csub\u003e1\u003c/sub\u003e to \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003eA\u003csub\u003e1\u003c/sub\u003e, produces the characteristic green emission.\u003c/p\u003e\n\u003cp\u003eAlthough MPs intrinsically require high stress to trigger light emission, recent research suggests that triboelectric effects can markedly enhance mechanoluminescent responses and reduce the activation threshold\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Jeong et al.\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e reported that under compression, interfacial delamination and air-gap formation occurred between rigid ZnS:Cu MPs and the softer polymer matrix due to mechanical mismatch. This delamination, combined with triboelectric charging arising from their distinct positions in the triboelectric series\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e, facilitates additional electron detrapping from shallow donor levels, thereby significantly amplifying light emission. Jeong\u0026rsquo;s findings challenge the conventional view that luminescence is solely governed by matrix elasticity and instead underscore the importance of interfacial triboelectric fields.\u003c/p\u003e\n\u003cp\u003eBuilding on this, we simulated the deformation of a DCP-3 matrix embedded with a single ZnS:Cu MP under applied pressure to quantify the effects of interfacial strain mismatch on electric potential and field distributions. Assuming global electroneutrality and a uniform charge distribution, a fixed positive surface charge density of 10\u003csup\u003e\u0026minus;\u0026thinsp;4\u003c/sup\u003e C\u0026middot;m\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e was assigned to the MP at the delamination interface\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, balanced by a deformation-induced negative surface charge on the DCP-3 matrix. For simulation details, see Supplementary Information. Simulations revealed peak electric potential and field values at the horizontal symmetry center, coinciding with interfacial delamination zones (Figs.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eb and \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003ec). Surprisingly, the triboelectric voltage (i.e., the maximum electric potential difference) increased linearly with matrix strain, whereas the maximum delamination-induced air-gap distance (\u003cem\u003ed\u003c/em\u003e\u003csub\u003emax\u003c/sub\u003e) exhibited a nonlinear strain dependence (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003ed). Our findings strongly indicate that the triboelectric voltage (\u003cem\u003eU\u003c/em\u003e), rather than the triboelectric field (\u003cem\u003eE\u003c/em\u003e\u0026thinsp;=\u0026thinsp;\u003cem\u003eU\u003c/em\u003e/\u003cem\u003ed\u003c/em\u003e\u003csub\u003emax\u003c/sub\u003e) in the Jeong\u0026rsquo;s study\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, maintains a more robust and predictable correlation with matrix strain. This relationship provides a reliable basis for sustaining a strong coupling between strain and luminescence intensity, even within the inelastic deformation regime of the polymer matrix. By leveraging the strain-triboelectric voltage-mechanoluminescence (STM) coupling mechanism, our findings provide a foundational strategy for developing optical waveguide-nociceptive sensor capable of selectively responding to both elastic and non-elastic deformation in self-healing polymers.\u003c/p\u003e\n\u003cp\u003eUnintended physical contact between robots and humans poses a significant safety concern in interactive environments\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e. Current industrial safety standards, such as ISO 10218 and ISO/TS 15066, impose conservative limits on robot speed, force, and power to minimize injury risk in shared workspaces\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. While effective in preventing harm, these rigid constraints often compromise robotic efficiency and lack the ability to distinguish between benign and hazardous interactions in real time. To overcome these limitations, by quantifying the energy flux density (i.e., the energy transferred per unit area) during collision events, we developed a control framework for physical human-robot interaction grounded in biomechanical injury thresholds. Remarkably, based on STM coupling mechanism, the energy flux density exhibits a piecewise-linear relationship with optical intensity (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003ef and Video S5), consistent with the simulated response of triboelectric voltage to strain (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003ed). This correlation allows the maximum optical signal to serve as a real-time proxy for estimating energy flux and categorizing collision risk. Specifically, medium-risk collisions (2\u0026ndash;6 mJ\u0026middot;mm\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e), which typically associates with contusions, are recommended to initiate a robotic deceleration protocol, while high-risk collisions (\u0026gt;\u0026thinsp;6 mJ\u0026middot;mm\u003csup\u003e\u0026minus;\u0026thinsp;2\u003c/sup\u003e), which may cause fractures, should trigger an emergency stop. Built upon its high-sensitivity, strong interference resistance, and rapid-response luminescence characteristics (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eg), the optical waveguide nociceptive sensing unit can offer a robust and scalable solution for intelligent, adaptive, and safety-aware human-robot interaction. By enabling real-time biomechanical risk assessment, this strategy may achieve an effective balance between robotic operational performance and injury prevention, paving the way for intelligence in collaborative robotics.\u003c/p\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n\u003ch2\u003eOptical-waveguide thermal sensing and applications\u003c/h2\u003e\n\u003cp\u003eAs an essential nociceptive modality, thermal perception in e-skin provides critical feedback for mitigating thermal hazards during physical human-robot interaction. To achieve this, we employed TPs consisting of phase-change microcapsules embedded with a reversible thermochromic dye (Fig. \u003cspan class=\"InternalRef\"\u003eS6\u003c/span\u003ec). Within a specific temperature range, the absorption spectrum of the TPs shifts in response to temperature changes (Video S6), exhibiting excellent reversibility and durability. Due to the low thermal conductivity of the organic polymer shell, the colour-change response time of the TPs is governed more by heat transfer kinetics than by the intrinsic phase transition process\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e. Similarly, the low thermal conductivity of the DCP-3 matrix further delays the thermal response of the thermochromic layer in e-skin.\u003c/p\u003e\n\u003cp\u003eDuring the thermochromic transition, the spectrometer collects reflected rather than transmitted light, necessitating subtraction of the initial-state reflection spectrum to obtain time-resolved absorption spectra. To avoid spectral overlap with the mechanoluminescent signals, we selected a red TP with a peak absorption at 591 nm that shifts to 577 nm upon activation (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eh). A one-dimensional heat conduction model was established across the thermochromic layer between its top surface (in contact with the heat source) and its bottom surface (optically coupled to the waveguide):\u003c/p\u003e\n\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\n\u003cdiv id=\"FileID_Equ1\" class=\"mathdisplay\"\u003e$$\\frac{{\\partial T(x,t)}}{{\\partial t}}=\\alpha \\frac{{{\\partial ^2}T(x,t)}}{{\\partial {x^2}}}$$\u003c/div\u003e\n\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\n\u003c/div\u003e\n\u003cp\u003e,\u003c/p\u003e\n\u003cp\u003ewhere \u003cem\u003eT\u003c/em\u003e(\u003cem\u003ex\u003c/em\u003e, \u003cem\u003et\u003c/em\u003e) is the temperature at position x\u0026isin;[0, \u003cem\u003ew\u003c/em\u003e] and time \u003cem\u003et\u003c/em\u003e, w is the thickness of the DCP-3 film, and \u003cem\u003e\u0026alpha;\u003c/em\u003e is the thermal diffusivity of the DCP-3. Under the assumption of a shallow surface transition (i.e., the contact time \u003cem\u003et\u003c/em\u003e \u0026rarr; 0), we surprisingly found that the temperature of the heat source can be analytically estimated as:\u003c/p\u003e\n\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\n\u003cdiv id=\"FileID_Equ2\" class=\"mathdisplay\"\u003e$$T={T_0}+S \\cdot \\frac{{I(t)}}{{\\sqrt t }}$$\u003c/div\u003e\n\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\n\u003c/div\u003e\n\u003cp\u003e,\u003c/p\u003e\n\u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\n\u003cdiv id=\"FileID_Equ3\" class=\"mathdisplay\"\u003e$$S=\\frac{{w({T_0} - {T_{Air}})}}{{\\sqrt {\\pi \\alpha } }}$$\u003c/div\u003e\n\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\n\u003c/div\u003e\n\u003cp\u003e,\u003c/p\u003e\n\u003cp\u003ewhere \u003cem\u003eT\u003c/em\u003e\u003csub\u003e0\u003c/sub\u003e is the transition temperature, \u003cem\u003eT\u003c/em\u003e\u003csub\u003eAir\u003c/sub\u003e is the ambient temperature, and \u003cem\u003eI(t)\u003c/em\u003e is the normalized intensity of absorption spectra (\u003cem\u003e0\u0026thinsp;\u0026le;\u0026thinsp;I(t)\u0026thinsp;\u0026le;\u0026thinsp;1\u003c/em\u003e). For derivation details, refer to the Supplementary Information.\u003c/p\u003e\n\u003cp\u003eWe further measured the relationship between the normalized absorption intensity and the square root of colour-change time under heat source temperatures ranging from 40 to 100\u0026deg;C (with ambient temperature \u003cem\u003eT\u003c/em\u003e\u003csub\u003eAir\u003c/sub\u003e=25\u0026deg;C). The resulting curves exhibited an approximately linear slope in the early stage of the thermochromic transition (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003ei). By fitting the slope values obtained at different heat source temperatures, we observed a near-linear dependence with the slope \u003cem\u003eS\u003c/em\u003e\u0026thinsp;=\u0026thinsp;17.014, which is in excellent agreement with the analytical prediction from Eq.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003ej). This result confirms the feasibility of establishing an absorption intensity-temperature sensing strategy. In practical scenarios, the robot can utilize this optical waveguide temperature sensing unit to assess burn risk during human-robot interaction and issue early warnings when the surface temperature exceeds 50\u0026deg;C. This framework supports the deployment of caregiver robots in daily service scenarios with enhanced safety and responsiveness.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn summary, we propose a biologically inspired, self-healable e-skin capable of quadruple-modal perception of tactile, pressure, nociceptive, and thermal stimuli. The e-skin adopts a vertically stacked five-layer architecture, comprising three DCP layers doped with functional MPs or TPs and two ELM layers doped with magnetic NPs. This synergistic configuration enables complete structural self-healing across a broad temperature range (\u0026ndash;10\u0026deg;C to 60\u0026deg;C), offering a structurally simple, maintenance-friendly, and modular solution for intelligent human-machine interaction. Through the integration of triboelectric and capacitive sensing modules, the e-skin achieves high sensitivity and spatiotemporal resolution in tracking and responding to tactile and pressure inputs. Meanwhile, the incorporation of optical waveguide sensing modules establishes coupling relationships between energy flux density and mechanoluminescent emission spectra, as well as between heat source temperature and the absorption spectra of the thermochromic layer. These optical correlations enable the quantitative assessment of nociceptive (impact) and thermal (burn) risks, providing a viable strategy for real-time injury prevention in human-machine interactive environments. Overall, our work offers a safe, reliable, and multifunctional framework for next-generation interactive electronics. Future research will focus on integrating machine learning-assisted adaptive perception with full-body deployment on reconfigurable robotic surfaces, ultimately advancing applications in intelligent robotics.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eChemicals and materials\u003c/h2\u003e\u003cp\u003ePMHS (\u003cem\u003eM\u003c/em\u003e\u003csub\u003en\u003c/sub\u003e=1700\u0026ndash;3200), VPDMS (850\u0026ndash;1150 cSt), PDMS-OH (3,500 cSt), BA (\u0026ge;\u0026thinsp;99.5%), and methanol (\u0026ge;\u0026thinsp;99.8%) were purchased from Sigma Aldrich and used as received. ZnS:Cu MPs (~\u0026thinsp;5 \u0026micro;m) were obtained from Shanghai Kerun Optoelectronics Technology Co., Ltd. Red and Blue TPs (~\u0026thinsp;3 \u0026micro;m, \u003cem\u003eT\u003c/em\u003e\u003csub\u003e0\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;31 ℃) were supplied by Shenzhen Huancai Technology Co., Ltd. Ga (~\u0026thinsp;0.5 mm, \u0026ge;\u0026thinsp;99.99%), In (~\u0026thinsp;0.5 mm, \u0026ge;\u0026thinsp;99.99%), Sn (~\u0026thinsp;0.5 mm, \u0026ge;\u0026thinsp;99.99%), and Ni (1\u0026ndash;3 \u0026micro;m, \u0026ge;\u0026thinsp;99.99%) particles were provided by Helian Xiamen Technology Co., Ltd.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eSynthesis of polymer films\u003c/h2\u003e\u003cp\u003eThe target polymer film was synthesized via a one-pot method by mechanically stirring the selected polymers with methanol solutions containing functional particles for 1 h. The resulting precursor was then cast into polytetrafluoroethylene (PTFE) molds (100 \u0026times; 100 \u0026times; 0.5 mm\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e), vacuum-degassed at room temperature for 30 min, and subsequently crosslink-cured at 160\u0026deg;C for 12 h to yield the final polymer film.\u003c/p\u003e\u003cp\u003eSpecifically, pure PBS was synthesized by mixing BA and PDMS-OH at a mass ratio of 1:20, while pure PDMS was prepared by blending PMHS and VPDMS at a mass ratio of 1:10. DCP-1 to DCP-4 were formulated by combining PBS and PDMS precursors at mass ratios ranging from 1:1 to 1:4, depending on the desired composition.\u003c/p\u003e\u003cp\u003eAdditionally, the methanol solution (50 g\u0026middot;L\u003csup\u003e\u0026ndash;\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e) was prepared by ultrasonically dispersing the corresponding particles in methanol for 15 min. In the final polymer films, MPs and TPs were incorporated at 20 wt% and 1 wt%, respectively, relative to the total polymer mass.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003eSynthesis of particle-embedded ELM alloy\u003c/h2\u003e\u003cp\u003eGa, In, and Sn particles were combined at mass ratios of 68.5 wt%, 21.5 wt%, and 10 wt%, respectively, and placed in a PTFE beaker. The mixture was heated to 250\u0026deg;C and held for 1 h under gentle stirring to promote uniform alloying. After naturally cooling to room temperature, 10 wt% of NPs were introduced, followed by vigorous stirring for 1 h to yield the final particle-embedded ELM alloy.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003eE-skin fabrication and Integration\u003c/h2\u003e\u003cp\u003eA 0.05 mm-thick custom-fabricated 304 stainless steel shadow mask was used to define the surface treatment area. The mask was aligned with the geometry of the e-skin, featuring patterned openings corresponding to the electrode and lead regions for subsequent ELM brushing. Under the protection of this mask, DCP-3 films doped with functional particles were placed in a dark chamber equipped with a UV lamp (\u003cem\u003eλ\u003c/em\u003e\u003csub\u003emax\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;185 nm, power density\u0026thinsp;=\u0026thinsp;85 \u0026micro;W\u0026middot;cm\u003csup\u003e\u0026ndash;\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e) and subjected to UV/ozone treatment for 2 h.\u003c/p\u003e\u003cp\u003eAfter treatment, the shadow mask was retained while the ELM was uniformly brushed onto the exposed regions and guided into patterned leads via magnetic alignment. The e-skin was then assembled by sequentially stacking: (i) an MPs-doped DCP-3 layer with an ELM pattern, (ii) a TPs-doped DCP-3 layer with an ELM pattern, and (iii) a TPs-doped DCP-3 layer without ELM. The electrode regions were carefully aligned and the leads positioned on opposite sides to prevent short-circuiting. In addition, Ag wires (0.1 mm in diameter) were connected to the ELM leads to enable subsequent electrical measurements. The assembled e-skin was vacuum-treated at 60\u0026deg;C for 1 h to ensure structural integration, followed by UV/ozone treatment for 1 h to remove surface tackiness.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003eCharacterization\u003c/h2\u003e\u003cp\u003eFTIR spectra were collected using a Nicolet iS50 spectrometer in attenuated total reflectance (ATR) mode over the range of 400\u0026ndash;4000 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. X-ray diffraction (XRD) patterns were collected using a Rigaku IV diffractometer with a Cu Kα radiation source (\u003cem\u003eλ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.5418 \u0026Aring;) over a 2\u003cem\u003eθ\u003c/em\u003e range of 10\u0026deg;\u0026ndash;80\u0026deg;. SEM images were collected using a Hitachi SU-70 scanning electron microscope to characterize microstructures. Mechanical testing was performed using an Instron Microtester 5984 universal testing machine according to ISO 37:2024. Dumbbell-shaped specimens (length: 10 mm, thickness: 2 mm) were tested under tensile and compressive loading at a constant strain rate of 1 mm\u0026middot;s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Rheological properties were measured using an MCR302 rheometer (Anton Paar) with a PP25 plate and a sample thickness of 1 mm, in oscillatory angular frequency sweep mode from 0.1 to 400 rad\u0026middot;s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e at an initial strain of 0.5%. DSC curves were collected using a DSC 204 F1 analyzer over the temperature range of \u0026minus;\u0026thinsp;150\u0026deg;C to 100\u0026deg;C at a heating rate of 10\u0026deg;C\u0026middot;min\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Triboelectric signals were recorded using a Keithley 6514B electrometer connected to a BNC-2110 preamplifier (National Instruments). Multichannel electrical signals were collected using a VK10x-MUL charge amplifier in conjunction with a Smacq USB-3200 data acquisition card. Capacitance signals were collected using a Keithley 7510 digital multimeter. Optical emission spectra were collected using a NOVA2S-EX thermoelectrically cooled back-illuminated array spectrometer (Focuslight Optics) operating across a wavelength range of 325\u0026ndash;1100 nm with a slit width of 200 \u0026micro;m. Optical signals were transmitted using a FIB-1000-NIR optical fiber.\u003c/p\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003ch2\u003eContributions\u003c/h2\u003e\u003cp\u003eH.W. and Z.Z. conceived and designed the study. H.W., H.X., Y.S., J.Y., Y.Z., J.L., Y.G., and L.L. carried out materials synthesis, device fabrication, and triboelectric/capacitive sensing experiments. H.C. performed the finite-element simulations and electrostatic modeling in COMSOL and contributed to data interpretation. H.W., H.X., Q.Y. and Y.S. conducted the optical waveguide experiments and the risk-classification analysis. Z.Z. analyzed the data and wrote the manuscript with input from all authors. J.S., S.L., and Z.Z. provided supervision and resources. All authors discussed the results, revised the manuscript, and approved the final version.\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003ch2\u003eCompeting interests\u003c/h2\u003e\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e\u003cp\u003eDeclaration\u003c/p\u003e\u003cp\u003eThis work was supported by the Natural Science Foundation of Fujian Province, China (Grant No. 2022J01044) and the State Key Laboratory of Advanced Nuclear Energy Technology, Nuclear Power Institute of China (Grant No. YNSW-0224-0101-19-01).\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eH.W. and Z.Z. conceived and designed the study. H.W., H.X., Y.S., J.Y., Y.Z., J.L., Y.G., and L.L. carried out materials synthesis, device fabrication, and triboelectric/capacitive sensing experiments. H.C. performed the finite-element simulations and electrostatic modeling in COMSOL and contributed to data interpretation. H.W., H.X., Q.Y. and Y.S. conducted the optical waveguide experiments and the risk-classification analysis. Z.Z. analyzed the data and wrote the manuscript with input from all authors. J.S., S.L., and Z.Z. provided supervision and resources. All authors discussed the results, revised the manuscript, and approved the final version.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eAll data generated in this study are provided in the Supplementary Information/Source Data file. Source data are provided with this paper.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eWang, M. \u003cem\u003eet al.\u003c/em\u003e Artificial skin perception. \u003cem\u003eAdv. Mater.\u003c/em\u003e 33, 2003014 (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYuan, Z. \u0026amp; Shen, G. Materials and device architecture towards a multimodal electronic skin. \u003cem\u003eMater. Today\u003c/em\u003e 64, 165\u0026ndash;179 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang, H. \u003cem\u003eet al.\u003c/em\u003e Humanoid electronic-skin technology for the era of artificial intelligence of things. \u003cem\u003eMatter\u003c/em\u003e 8, 102136 (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang, X. \u003cem\u003eet al.\u003c/em\u003e Highly stretchable electronic-skin sensors with porous microstructure for efficient multimodal sensing with wearable comfort. \u003cem\u003eAdv. Mater. Interfaces\u003c/em\u003e 10, 2201958 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFu, X., Cheng, W., Wan, G., Yang, Z. \u0026amp; Tee, B. C. K. Toward an AI era: advances in electronic skins. \u003cem\u003eChem. Rev.\u003c/em\u003e 124, 9899\u0026ndash;9948 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXu, C. \u003cem\u003eet al.\u003c/em\u003e A physicochemical-sensing electronic skin for stress response monitoring. \u003cem\u003eNat. Electron.\u003c/em\u003e 7, 168\u0026ndash;179 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShin, J. \u003cem\u003eet al.\u003c/em\u003e Bioresorbable pressure sensors protected with thermally grown silicon dioxide for the monitoring of chronic diseases and healing processes. \u003cem\u003eNat. Biomed. Eng.\u003c/em\u003e 3, 37\u0026ndash;46 (2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKim, T. \u003cem\u003eet al.\u003c/em\u003e Ultrathin crystalline-silicon-based strain gauges with deep learning algorithms for silent speech interfaces. \u003cem\u003eNat. Commun.\u003c/em\u003e 13, 5815 (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWesterveld, W. J. \u003cem\u003eet al.\u003c/em\u003e Sensitive, small, broadband and scalable optomechanical ultrasound sensor in silicon photonics. \u003cem\u003eNat. Photonics\u003c/em\u003e 15, 341\u0026ndash;345 (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKim, J. \u003cem\u003eet al.\u003c/em\u003e Stretchable silicon nanoribbon electronics for skin prosthesis. \u003cem\u003eNat. Commun.\u003c/em\u003e 5, 5747 (2014).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKang, J., Tok, J. B. H. \u0026amp; Bao, Z. Self-healing soft electronics. \u003cem\u003eNat. Electron.\u003c/em\u003e 2, 144\u0026ndash;150 (2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYang, J. C. \u003cem\u003eet al.\u003c/em\u003e Electronic skin: recent progress and future prospects for skin-attachable devices for health monitoring, robotics, and prosthetics. \u003cem\u003eAdv. Mater.\u003c/em\u003e 31, 1904765 (2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRoels, E. \u003cem\u003eet al.\u003c/em\u003e Processing of self-healing polymers for soft robotics. \u003cem\u003eAdv. Mater.\u003c/em\u003e 34, 2104798 (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBai, H., Kim, Y. S. \u0026amp; Shepherd, R. F. Autonomous self-healing optical sensors for damage intelligent soft-bodied systems. \u003cem\u003eSci. Adv.\u003c/em\u003e 8, eabq2104 (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJung, J. \u003cem\u003eet al.\u003c/em\u003e Self-healing electronic skin with high fracture strength and toughness. \u003cem\u003eNat. Commun.\u003c/em\u003e 15, 9763 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYang, L. \u003cem\u003eet al.\u003c/em\u003e Self-Healing, reconfigurable, thermal-switching, transformative electronics for health monitoring. \u003cem\u003eAdv. Mater.\u003c/em\u003e 35, 2207742 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu, Y. \u003cem\u003eet al.\u003c/em\u003e Applications of cellulose-based flexible self-healing sensors for human health monitoring. \u003cem\u003eNano Energy\u003c/em\u003e 127, 109790 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, X. \u003cem\u003eet al.\u003c/em\u003e Self-healing liquid metal hydrogel for human\u0026ndash;computer interaction and infrared camouflage. \u003cem\u003eMater. Horiz.\u003c/em\u003e 10, 2945\u0026ndash;2957 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSun, Y., Li, X., Hu, T., Zhang, S. \u0026amp; Niu, W. Dynamic acylhydrazone bonds cross-linked chromotropic photonic-ionic skin with self-healing and high resilience for interactive human-machine interface. \u003cem\u003eChem. Eng. J.\u003c/em\u003e 475, 146188 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWhite, S. R. \u003cem\u003eet al.\u003c/em\u003e Autonomic healing of polymer composites. \u003cem\u003eNature\u003c/em\u003e 409, 794\u0026ndash;797 (2001).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSong, F. \u003cem\u003eet al.\u003c/em\u003e Tunable \u0026ldquo;soft and stiff\u0026rdquo;, self-healing, recyclable, thermadapt shape memory biomass polymers based on multiple hydrogen bonds and dynamic imine bonds. \u003cem\u003eJ. Mater. Chem. A\u003c/em\u003e 7, 13400\u0026ndash;13410 (2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTang, M. \u003cem\u003eet al.\u003c/em\u003e Autonomous self-healing, self-adhesive, highly conductive composites based on a silver-filled polyborosiloxane/polydimethylsiloxane double-network elastomer. \u003cem\u003eJ. Mater. Chem. A\u003c/em\u003e 7, 27278\u0026ndash;27288 (2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen, Y. \u003cem\u003eet al.\u003c/em\u003e Hydrogel-based pressure sensors for electronic skin systems. \u003cem\u003eMatter\u003c/em\u003e 8, 101992 (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePark, H. \u003cem\u003eet al.\u003c/em\u003e Toughening self-healing elastomer crosslinked by metal\u0026ndash;ligand coordination through mixed counter anion dynamics. \u003cem\u003eNat. Commun.\u003c/em\u003e 14, 5026 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen, J. \u003cem\u003eet al.\u003c/em\u003e Ladderphane copolymers for high-temperature capacitive energy storage. \u003cem\u003eNature\u003c/em\u003e 615, 62\u0026ndash;66 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eUrban, M. W. \u003cem\u003eet al.\u003c/em\u003e Key-and-lock commodity self-healing copolymers. \u003cem\u003eScience\u003c/em\u003e 362, 220\u0026ndash;225 (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhu, R., Zhu, D., Zheng, Z. \u0026amp; Wang, X. Tough double network hydrogels with rapid self-reinforcement and low hysteresis based on highly entangled networks. \u003cem\u003eNat. Commun.\u003c/em\u003e 15, 1344 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSun, F. \u003cem\u003eet al.\u003c/em\u003e Vascular smooth muscle-inspired architecture enables soft yet tough self-healing materials for durable capacitive strain-sensor. \u003cem\u003eNat. Commun.\u003c/em\u003e 14, 130 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, B., Cao, P.-F., Saito, T. \u0026amp; Sokolov, A. P. Intrinsically self-healing polymers: from mechanistic insight to current challenges. \u003cem\u003eChem. Rev.\u003c/em\u003e 123, 701\u0026ndash;735 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, S. \u003cem\u003eet al.\u003c/em\u003e Bimodal capacitive tactile sensor assisted by shield effect of triboelectric nanogenerator. \u003cem\u003eNano Energy\u003c/em\u003e 118, 108946 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang, H., Li, H. \u0026amp; Li, Y. Biomimetic electronic skin for robots aiming at superior dynamic-static perception and material cognition based on triboelectric-piezoresistive effects. \u003cem\u003eNano Lett.\u003c/em\u003e 24, 4002\u0026ndash;4011 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQiu, Y. \u003cem\u003eet al.\u003c/em\u003e Quantitative softness and texture bimodal haptic sensors for robotic clinical feature identification and intelligent picking. \u003cem\u003eSci. Adv.\u003c/em\u003e 10, eadp0348 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMa, X. \u003cem\u003eet al.\u003c/em\u003e Bimodal tactile sensor without signal fusion for user-interactive applications. \u003cem\u003eACS Nano\u003c/em\u003e 16, 2789\u0026ndash;2797 (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQu, X. \u003cem\u003eet al.\u003c/em\u003e Artificial tactile perception smart finger for material identification based on triboelectric sensing. \u003cem\u003eSci. Adv.\u003c/em\u003e 8, eabq2521 (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFang, H. \u003cem\u003eet al.\u003c/em\u003e A flexible dual-mode photodetector for human\u0026ndash;machine collaborative IR imaging. \u003cem\u003eNano-Micro Lett.\u003c/em\u003e 17, 229 (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEdwards, C. \u0026amp; Marks, R. Evaluation of biomechanical properties of human skin. \u003cem\u003eClin. Dermatol.\u003c/em\u003e 13, 375\u0026ndash;380 (1995).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen, J., Yi, D., Shen, B. \u0026amp; Zheng, W. Multifunctional liquid-metal composites for electromagnetic communication and attenuation. \u003cem\u003eAdv. Mater.\u003c/em\u003e 37, 2404595 (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGuo, R. \u003cem\u003eet al.\u003c/em\u003e Ni-GaIn amalgams enabled rapid and customizable fabrication of wearable and wireless healthcare electronics. \u003cem\u003eAdv. Eng. Mater.\u003c/em\u003e 20, 1800054 (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYu, L. \u003cem\u003eet al.\u003c/em\u003e Transportable, endurable, and recoverable liquid metal powders with mechanical sintering conductivity for flexible electronics and electromagnetic interference shielding. \u003cem\u003eACS Appl. Mater. Interfaces\u003c/em\u003e 14, 48150\u0026ndash;48160 (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu, Y. \u003cem\u003eet al.\u003c/em\u003e Triboelectric tactile sensor for pressure and temperature sensing in high-temperature applications. \u003cem\u003eNat. Commun.\u003c/em\u003e 16, 383 (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXia, S. Y. \u003cem\u003eet al.\u003c/em\u003e High-performance Cu\u0026ndash;Cu interconnects attained through air sintering of oleylamine-capped Cu nanoparticles for power electronics packaging. \u003cem\u003eRare Metals\u003c/em\u003e 44, 3281\u0026ndash;3298 (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJeong, H. I. \u003cem\u003eet al.\u003c/em\u003e Super elastic and negative triboelectric polymer matrix for high performance mechanoluminescent platforms. \u003cem\u003eNat. Commun.\u003c/em\u003e 16, 854 (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLin, C. \u003cem\u003eet al.\u003c/em\u003e Electrically chargeable inorganic persistent luminescence in an alternating current driven electroluminescent device. \u003cem\u003eCommun. Mater.\u003c/em\u003e 6, 25 (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePeka, P. \u0026amp; Schulz, H. J. Empirical one-electron model of optical transitions in Cu-doped ZnS and CdS. \u003cem\u003ePhysica B\u003c/em\u003e 193, 57\u0026ndash;65 (1994).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang, X. \u003cem\u003eet al.\u003c/em\u003e Dynamic pressure mapping of personalized handwriting by a flexible sensor matrix based on the mechanoluminescence process. \u003cem\u003eAdv. Mater.\u003c/em\u003e 27, 2324\u0026ndash;2331 (2015).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePan, X. \u003cem\u003eet al.\u003c/em\u003e Quantifying the interfacial triboelectricity in inorganic-organic composite mechanoluminescent materials. \u003cem\u003eNat. Commun.\u003c/em\u003e 15, 2673 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHaddadin, S., De Luca, A. \u0026amp; Albu-Sch\u0026auml;ffer, A. Robot collisions: a survey on detection, isolation, and identification. \u003cem\u003eIEEE T. Robot.\u003c/em\u003e 33, 1292\u0026ndash;1312 (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHaddadin, S., Albu-Sch\u0026auml;ffer, A. \u0026amp; Hirzinger, G. The role of the robot mass and velocity in physical human-robot interaction\u0026mdash;part I: non-constrained blunt impacts. in \u003cem\u003eIEEE Int. Conf. Robot. Autom.\u003c/em\u003e 1331\u0026ndash;1338 (2008).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang, Y., Liu, H., Niu, J., Wang, X. \u0026amp; Wu, D. Development of reversible and durable thermochromic phase-change microcapsules for real-time indication of thermal energy storage and management. \u003cem\u003eAppl. Energ.\u003c/em\u003e 264, 114729, (2020).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7503393/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7503393/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eE-skins capable of multimodal perception are essential for intelligent human-machine interaction, yet integrating real-time responsiveness with structural self-healing across multiple sensing modalities remains a significant challenge. Here, we report a biologically inspired, self-healing e-skin that enables the perception of tactile, pressure, nociceptive, and thermal stimuli. The device adopts a vertically stacked compact architecture comprising double-network cross-linked polymer layers and eutectic liquid metal layers, enabling rapid and complete structural healing even after severe mechanical damage over a wide temperature range. We demonstrate that the e-skin retains its functional integrity after healing, with triboelectric and capacitive sensing units enabling high-sensitivity, spatiotemporally resolved tracking of tactile and pressure stimuli, respectively. Meanwhile, by quantitatively analyzing the mechanoluminescent and thermochromic spectra, the optical waveguide sensing units enable real-time optical encoding of nociceptive and thermal stimuli, thereby allowing effective classification of impact and burn injury risks. Our work lays a solid foundation for the development of intelligent robotic systems capable of adaptive perception and injury prevention in complex and dynamic human-machine environments.\u003c/p\u003e","manuscriptTitle":"A self-healing e-skin for quadruple-modal sensing","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-23 11:48:47","doi":"10.21203/rs.3.rs-7503393/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"55f4c655-e5e4-4e45-bf39-708bff617ca8","owner":[],"postedDate":"September 23rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":55064689,"name":"Physical sciences/Materials science"},{"id":55064690,"name":"Physical sciences/Optics and photonics"},{"id":55064691,"name":"Physical sciences/Physics"}],"tags":[],"updatedAt":"2025-09-23T11:48:47+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-23 11:48:47","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7503393","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7503393","identity":"rs-7503393","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}