A Microphase‐Separated Design toward an All‐Round Ionic Hydrogel with Discriminable and Anti‐Disturbance Multisensory Functions

Stretchable ionic hydrogels with superior all‐round properties that can detect multimodal sensations with excellent discriminability and robustness against external disturbances are highly required for artificial electronic skinapplications. However, some critical material parameters exhibit intrinsic tradeoffs with each other for most ionic hydrogels. Here, a microphase‐separated hydrogel is demonstrated by combining three strategies: (1) using of a low crosslinker/monomer ratio to obtain highly entangled polymer chains as the first network; (2) the introduction of zwitterions into the first network; (3) the synthesis of an ultrasoft polyelectrolyte as the second network. This all‐round elastic ionic hydrogel exhibits a low Young's modulus (< 60 kPa), large stretchability (> 900%), high resilience (> 95%), unique strain‐stiffening behavior, excellent fatigue tolerance, high ionic conductivity (> 2.0 S m⁻1), and anti‐freezing capability, which have not been achieved before. These properties allow the ionic hydrogel to operate as a stretchable multimodal sensor that can detect and decouple multiple stimuli (temperature, pressure, and proximity) with excellent discriminability, high sensitivity, and strong sensing‐robustness against strains or temperature perturbations. The ionic hydrogel sensor exhibits great potential for intelligent electronic skin applications such as reliable health monitoring and accurate object identification.


Introduction
Skin, the body's all-round and most versatile organ, provides protection and receives external sensory stimuli. [1]Artificial electronic skin is designed to mimic the multisensory functions and unique mechanical properties of natural skin and has been applied in advanced robotics, intelligent medical diagnostics, skin-attachable wearables, and human-machine interfaces. [2]s the primary part of the somatosensory system, the human skin can feel multiple sensations with high discriminability to decouple multiple stimuli without interference via different mechanoreceptors, which are composed of ion conductors in the skin. [3]Moreover, the skin has intriguing mechanical properties, including softness, elasticity, fatigue resistance, and strain-stiffening behavior, which allow the embedded mechanoreceptors to perceive with high sensing accuracy, reliability, and superior robustness against mechanical disturbances.Despite these impressive characteristics, its physiological functions are susceptible to harsh climatic conditions. [4]For instance, the human skin lacks an anti-freezing capability, and low environmental temperatures can decrease its sensing and protective functions and increase its susceptibility to mechanical stresses. [4]Thus, artificial electronic skins have been developed to mimic skin's comprehensive sensory and mechanical properties while also adding new desired properties such as freezing tolerance. [5][8][9] Various ionic hydrogels have been applied in the design of wearable sensing devices that can detect stimuli such as strain, pressure, or temperature. [10]espite notable advances in the development of ionic hydrogel sensors, there has been no report of a stretchable ionic hydrogel that simultaneously possesses robust and discriminable multisensory functions, skin-like mechanical properties, and skinabsent, yet highly desirable, anti-freezing capability.Previous studies have realized multisensory functions in one stretchable ionic hydrogel, but multiple stimuli were usually converted into one type of electrical signal output, such as a resistance or capacitance change. [3,6,10,11]Thus, they could not decouple the interference of multiple signals and lacked stimuli discriminability and sensing robustness. [11]hile stretchable and ionic-conductive hydrogels have been frequently reported, most cannot combine skin-like mechanical and physical properties and anti-freezing properties since some of the critical material parameters exhibited intrinsic tradeoffs with each other.[14] High resilience and low hysteresis are essential for obtaining stable and durable stretchable materials where repetitive movements or loadings are required. [13]However, polymer hydrogels with good stretchability typically have difficulty achieving a low hysteresis and high resilience due to the relaxation of dangling chains or chemical bond breakage. [7,14][17] A key property of polyelectrolyte hydrogels is their ionic conductivity, [18] but they are often mechanically weak due to their high swelling ratios and lack of stress dissipation mechanisms. [19]Double-network toughening strategies include the introduction of a dense neutral polymer into the polyelectrolyte matrix or the use of another polyelectrolyte with equivalent opposite charges to form a dense polyelectrolyte complex. [19]These approaches have been developed to improve the stretchability and toughness of polyelectrolyte hydrogels, but they typically employ a polyelectrolyte as the brittle and sacrificial first network to dissipate energy upon stretching, thus inevitably leading to irreversible mechanical degradation, large hysteresis, and a decrease in ionic conductivity. [19]hird, there is a tradeoff between the anti-freezing performance and mechanical properties of conventional hydrogels. [20]Almost all reported anti-freezing gels are either organogels (including organo/hydrogels) or hydrophilic hydrogels that contain antifreezing additives such as ionic compounds, natural proteins, and carbon nanomaterials. [21][24] Till now, there have been no reports on stretchable ionic hydrogels with skinlike distinguishable multisensory functions, superior mechanical properties, and skin-absent anti-freezing properties.
Herein, we report the design and fabrication of a highly elastic, superb robust, and anti-freezing artificial multisensory skin based on an interpenetrating network (IPN) ionic polymer hydrogel with a microphase-separated microstructure by combining three strategies: (1) the use of a highly-entangled copolymer as the first network, (2) the introduction of zwitterions into the first network, and (3) the use of an ionic conductive polyelectrolyte as the second network.The highly entangled structure together with strong intermolecular interactions between the interpenetrating networks facilitated the formation of microphase-separated domains, which greatly improved the elasticity and robustness of the hydrogel.The resultant ionic hydrogel showed a high ionic conductivity (> 2.0 S/m), large stretchability (> 900%), superb elasticity (resilience > 95%, hysteresis < 5%), skin-like softness (Young's modulus < 60 kPa), unique strain-stiffening behavior, excellent fatigue resistance, and desirable moisture-retention and anti-freezing properties.Similar to skin tactile, the ionic hydrogel skin could detect and decouple multiple stimuli (temperature, pressure, and proximity) via an ionic thermoelectric effect, piezocapacitive effect, and fringe field effect, respectively.The combination of mechanical and physical effects enabled the multimodal hydrogel sensor to exhibit high-capacitive pressure sensitivity (up to 7.5 kPa −1 ), high-temperature sensing resolution within 0.1°C, and excellent sensing robustness against strain and temperature perturbations.This ionic hydrogel showed great promise in reliable health monitoring and accurate object identification applications.

Design and Fabrication of the Hydrogels
In contrast to the conventional double-network toughening strategy, [25] we combined three approaches to fabricate our ionic hydrogel skin to overcome the tradeoff issues.First, a singlenetwork copolymer hydrogel was synthesized by using neutral acrylamide (AM) and zwitterionic sulfobetaine methacrylate (SBMA) as monomers and N, N′-methylene bis(acrylamide) (MBAA) as a cross-linker in a molar ratio of AM: SBMA: MBAA = 2175: 188: 1. Due to the low value of the crosslinker/monomer molar ratio (0.00042) of poly(AM-co-SBMA), this first network hydrogel was highly-entangled and was mainly crosslinked by these entanglements (Figure 1a), giving it a high degree of elasticity. [26,27]Second, the poly(AM-co-SBMA) hydrogel was immersed in 45 wt.% diallyldimethyl-ammonium chloride (DAD-MAC) precursor solution, and the first network was swollen in the DADMAC solution with a swelling ratio of ≈200% due to the polyelectrolyte nature and osmotic pressure of SBMA (Figure S1, Supporting Information). [18,19,28]After swelling, DADMAC was polymerized to form poly(diallyldimethyl-ammonium chloride (PDADMAC) in the presence of a diluted first network to obtain a poly(AM-co-SBMA)/PDADMAC hydrogel with an IPN structure (denoted as PAS/PD-IPN hydrogel).The molar ratio of AM: SBMA: DADMAC was optimized to be 2175: 188: 3760 (discussed below).PDADMAC was chosen as the polyelectrolyte second network because of its high ionic conductivity [18,29], ultrasoft nature, [19] and antibacterial quaternary ammonium structure, [30] which are advantageous for wearable applications.Third, the zwitterionic SBMA in the highly entangled first network formed strong intermolecular bonds (SBMA-DADMAC ionic complex) with PDADMAC (Figure 1a). [31,32]These allowed the formation of microphase-separated domains inside the hydrogels, which endowed the hydrogel with excellent elasticity and robustness. [33,34]or comparison, a poly(AM-co-SBMA) single network hydrogel (denoted PAS-SN), PDADMAC single network hydrogel (denoted PD-SN), and poly(acrylamide)/PDADMAC IPN hydrogel (denoted PA/PD-IPN) were also fabricated.

Characterization
The highly entangled structure and strong intermolecular interactions together contributed to the formation of microphaseseparated domains, as reflected by the macroscopic morphological change from transparent to opaque.Figure 1b presents photos of the PAS/PD-IPN hydrogel showing its opacity, which was caused by interfacial reflections and a biphase refractive index difference due to the existence of phase-separated regions. [35]n contrast, the PAS-SN and PD-SN with uniform microstructures showed a greater light penetration depth and were entirely transparent.To further verify microphase separation, scanning electron microscopy (SEM) was conducted to dissect local structural differences between the PAS-SN and PAS/PD-IPN hydrogels.The SEM image in Figure 1c shows that PAS/PD-IPN presented a microporous skeleton with microphase-separated domains, while PAS-SN in Figure 1d possessed a loose and porous network, which confirmed the internal attraction-induced structural changes.
To accurately investigate this phenomenon from a microscopic perspective, the internal microstructure of phase-separated hydrogel was visualized using in situ nondestructive nanocomputed tomography (Nano CT) scanning.In the 3D reconstruction (Figure 1e), the light element of the solvent phase (colored grey) and the heavy element of the polymer phase (colored blue) in the PAS/PD-IPN hydrogel exhibited a clear solid-liquid interphase boundary and a denser solid polymer-rich region.The binary images in Figure S2a (Supporting Information) reveal that similar to traditional microphase separation, the polymer-rich region and solvent-rich region formed a bicontinuous, microscale structure.The distance between adjacent polymer-rich regions calculated from Figure S2b (Supporting Information) increased from 25 μm (PAS-SN) to 61 μm (PAS/PD-IPN), which qualitatively indicates phase separation in the PAS/PD-IPN hydrogel. [36]ccording to our initial design, interchain interactions and a highly entangled network are both necessary for jointly accelerating the formation of a microphase-separated structure.Therefore, to verify the interchain interactions between PAS-SN, PD-SN, and PAS/PD-IPN, Fourier-transform infrared (FT-IR) spectroscopy and X-ray photoelectron spectroscopy (XPS) characterizations were applied.For PAS-SN hydrogel, the characteristic absorption bands, which were assigned to the S = O asymmetric stretching bands of -SO 3− in SBMA, shifted from 1180 cm −1 to 1186 cm −1 after forming the IPN hydrogel with PDADMAC (Figure 1f). [28]This shift implies a strong attraction between the anionic sulfonate group of SBMA and the cationic quaternary ammonium group of PDADMAC.Moreover, the results in Figure 1g indicated that due to the strong interactions between the quaternary ammonium groups and -SO 3− , the spinorbit doublet of S2p of -SO 3− species from the zwitterionic moiety of PAS-SN shifted from 168.68 and 167.46 eV to 168.5 and 167.3 eV, respectively. [37,38]These suggest the existence of the SBMA-DADMAC ionic complex in the PAS/PD-IPN hydrogel.
We further studied how different AM-to-SBMA molar ratios and DADMAC concentrations in precursor solution contributed to the degree of microphase separation in PAS/PD-IPN hydrogel.AM and SBMA ratio adjustments were essential to the construction of the PA x S y -SN (x and y represent the molar ratio of AM and SBMA, respectively).We fixed the total mass of AM and SBMA monomers and initiators, and changed the molar ratio of AM and SBMA from 2894: 0 to 719: 548 (denoted as PA 2894 S 0 , PA 2175 S 188 , PA 1678 S 308 , PA 1199 S 428 , and PA 719 S 548 , respectively) (Figure S3, Supporting Information).In this case, only PA 2894 S 0 , PA 2175 S 188 , PA 1678 S 308 can form highly entangled hydrogels due to their high crosslinker/monomer molar ratio. [26,27]After soaking the PA x S y -SN into DADMAC precursor solutions (fixed at 45 wt.% of DADMAC) without polymerization, all hydrogels remained in high transparency (Figure S3, Supporting Information), indicating no phase separation was generated.After DADMAC was polymerized to PD to form IPN hydrogel, opacity phenomenon (Figure S4, Supporting Information) and microphase-separated domains (Figure S5, Supporting Information) appeared in PA 2175 S 188 /PD-IPN, PA 1678 S 308 /PD-IPN, and PA 1199 S 428 /PD-IPN hydrogels.Among them, PA 2175 S 188 /PD-IPN hydrogel exhibited the lowest transparency (Figure S4, Supporting Information) and had the largest microphase-separated domains (Figure S5, Supporting Information).In contrast, microphaseseparated domains are hardly found in PA 2894 S 0 /PD-IPN and PA 719 S 548 /PD-IPN hydrogels that exhibited high transparency.This suggests that single chain-entanglement (PA 2894 S 0 /PD-IPN) or only SBMA-DADMAC interactions (PA 719 S 548 /PD-IPN) did not lead to microphase separation and that the optimal AM-to-SBMA ratio was 2175: 188.Next, we fixed the AM-to-SBMA molar ratio at 2175: 188 but changed the concentration of DADMAC in precursor solutions from 30 to 60 wt.% to construct PA 2175 S 188 /PD z -IPN hydrogel (z represents initial DADMAC concentration in precursor solutions, denoted from PA 2175 S 188 /PD 30 to PA 2175 S 188 /PD 60 -IPN) (Figure S6, Supporting Information).As the DADMAC contents increased, the transparency of the hydrogels showed a mild decrease from PA 2175 S 188 /PD 30 -IPN to PA 2175 S 188 /PD 40 -IPN, followed by a rapid decrease to PA 2175 S 188 /PD 45 -IPN (Figure S7, Supporting Information).As shown in SEM images, PA 2175 S 188 /PD 45 -IPN hydrogel had the most and largest microphase-separated domains in its porous skeleton (Figure S8, Supporting Information).Macroscopic phase separation appeared in PA 2175 S 188 /PD 60 hydrogel with DADMAC content reaching 60 wt.% (Figure S6, Supporting Information).Different from microphase-separated domains, such large macroscopic phase-separated structures would bring a negative effect on the mechanical properties of the hydrogels.All these results reconfirmed that strong interactions between the two skeletons and the high entanglement of the first PAS-SN network were jointly responsible for phase separation.

Mechanical and Anti-Freezing Properties
Next, tensile stress-strain tests were performed to investigate the mechanical elasticity of PA x S y /PD z -IPN hydrogels with different molar ratios of AM and SBMA and different contents of DADMAC.Hysteresis and resilience are two of the most prominent parameters used to assess the viscoelasticity of hydrogels and elastomers. [13]Hysteresis is defined as the energy lost during energy storage and dissipation during material stretching and release. [13]Resilience is defined as the ratio of energy recovered during unloading to the work done to the material during loading. [27]We measured the cyclic loading-unloading tensile curves of the PA x S y /PD z -IPN hydrogels under 200% strains to demonstrate their elasticity.As shown in Figure S9 (Supporting Information), the PA 2175 S 188 /PD 45 -IPN exhibited the highest resilience value (> 96%) among the PA x S y /PD 45 -IPN hydrogels with varied AM to SBMA molar ratios.As to the PA 2175 S 188 /PD z -IPN hydrogels with different DADMAC contents, the PA 2175 S 188 /PD 45 -IPN hydrogel also showed the best resilience values (Figure S10, Supporting Information).This is consistent with the optimal microphase separation phenomenon as discussed above.The microphase-separated domains can work as strong crosslinking points in the IPN hydrogels, which makes it possible to improve the mechanical elasticity upon stress or compression loading. [33,34]For convenience, PAS/PD-IPN was utilized to represent the optimal PA 2175 S 188 /PD 45 -IPN in this work.
Tensile stress-strain tests were then performed to quantitatively examine the mechanical properties of all PAS/PD-IPN, PAS-SN, PD-SN, and PA/PD-IPN hydrogels.As shown in Figure 2a, the PD-SN electrolyte hydrogel showed a low Young's modulus, tensile strength, strain at break, and toughness (9.6 kPa, 18.3 kPa, 250%, and 22.4 kJ m −3 , respectively).After forming an interpenetrating network with PAS, the PAS/PD-IPN hydrogel exhibited a significantly higher Young's modulus, tensile strength, strain at break, and toughness (60.2 kPa, 220.2 kPa, 900%, and 1023 kJ m −3 , respectively) than the PD-SN hydrogels (Figure 2a).The introduction of PD played a significant role on decreasing the Young's modulus of the hydrogels (Figure 2a; Table S1, Supporting Information).Soft nature is favorable for enhancing capacitive pressure sensitivity of the hydrogel sensor.When compared with the PAS-SN hydrogel, the PAS/PD-IPN hydrogel was softer yet stronger (Table S1, Supporting Information).While PAS/PD-IPN and PAS-SN hydrogels showed similar strain at break values, the higher strength of the PAS/PD-IPN hydrogel was attributed to its unique strain-stiffening behavior. [39]he differential modulus first decreases from 60.2 to 14.7 kPa in the strain range of 0-350%, and then increases to 27.9 kPa at the maximum 900% strain, corresponding to the typical characteristic of stiffness variation.The PAS/PD-IPN hydrogel was soft to the touch due to its low Young's modulus at low strains. [6]How-ever, it could stiffen due to a rapid increase in its Young's modulus at large strains. [40]The initial softness at a low strain state was mainly due to the low modulus of the PDADMAC network and the microphase-separated microstructure. [19,36]After stretching to a large strain, the alignment and subsequent fragmentation of fragile PDADMAC resulted in stiffening behavior under a large strain. [32]e next performed dynamic mechanical analysis (DMA) on the hydrogels to study their viscoelastic properties.As shown in Figure 2b, the storage modulus (G′) was about one order of magnitude larger than the loss modulus (G″) over a frequency range of 1-100 rad s −1 at ambient temperature for PAS/PD-IPN hydrogel.This indicates the elastic-dominated mechanical behavior of the ionic hydrogel. [41,42]The weak frequency dependence of both values reveals the stability of the PAS/PD-IPN hydrogel, which was attributed to the strong microphaseseparated domains and highly entangled polymer networks in the hydrogels. [43]e further measured the cyclic loading-unloading tensile curves of the PAS-SN, PA/PD-IPN, and PAS/PD-IPN hydrogels under different strains and deformation ratios to demonstrate their elasticity.The resilience values of the PAS/PD-IPN hydrogel were in the range of 95-99.5% up to a strain of 400% (Figure 2c).These values were much higher than those for the PAS-SN and PA/PD-IPN hydrogels (Figure 2d; Table S1, Supporting Information).This comparison indicates that the design containing zwitterionic monomers and the formation of microphaseseparated domains played critical roles in improving the elasticity of the ionic hydrogel.Moreover, the PAS/PD-IPN hydrogel showed no hysteresis loops under a cyclic strain of 100% and exhibited low hysteresis (< 5%) when the strain increased to 300 and 400% (Figure S11, Supporting Information).The tensile response of the PAS/PD-IPN hydrogel was nearly independent of the stretching rate, in contrast to conventional polymer hydrogels (Figure 2e). [44,45]The PAS/PD-IPN hydrogel exhibited small hysteresis loops (< 3%), with a resilience higher than 94% at strain rates in the range of 20-200 mm/min at 200% strain (Figure 2f), which are much better than those for PAS-SN and PA/PD-IPN hydrogels without microphase-separated domains (Figures S12 and S13, Supporting Information).The relaxation of dangling chains and chemical bond breakage would lead to energy dissipation, which would decrease the resilience and increase the hysteresis of conventional hydrogels or elastomers.Very differently, the microphase-separated domains can work as strong crosslinking points and prevent the entangled chain from relaxing and chemical bonds from rupturing inside the domains, thus significantly limiting the energy dissipation.As a result, the PAS/PD-IPN hydrogel exhibited high resilience, low hysteresis, and stretching rate-independent tensile response.The resilience and hysteresis values of PAS/PD-IPN hydrogel belong to one of the best values for polymer hydrogels and elastomers (Table S2, Supporting Information).
With low energy dissipation and excellent resilience, the IPN ionic hydrogel was considered to exhibit good mechanical fatigue resistance. [13]We demonstrate this fatigue tolerance behavior by performing cyclic tensile tests at a fixed strain of 100% under a deformation rate of 100 mm/min.As displayed in Figure 2g, the loading-unloading tensile curves were almost unchanged, with low hysteresis and high resilience for more than 1 000 strain cycles.In comparison, the loading-unloading tensile curves of the PD-SN hydrogel exhibited large residual strain for more than 1 000 strain cycles, and hysteresis increased with the number of strain cycles for the PAS-SN hydrogel (Figure S14, Supporting Information).The PAS/PD-IPN hydrogel also exhibited excellent resilience and fatigue resistance during consecutive compressive testing.Cyclic stress-strain tests under gradually increasing compressive strains from 20% to 80% (Figure 2h) and consecutive compressive cycles at a strain of 80% (Figure 2i) were conducted for PD-SN, PAS-SN, and PAS/PD-IPN hydrogels.The loading-unloading curves of the PAS/PD-IPN ionic hydrogel during different cycles almost overlapped with each other after the first ten cycles and exhibited little residual strain during 200 cycles (Figure 2i).In contrast, the PD-SN hydrogel broke after one compressive cycle at a strain of 80% (Figure S15a, Supporting Information).The PAS-SN hydrogel exhibited large residual strain in its compressive curves during multiple loading-unloading cycles (Figure S15b).The excellent elasticity of the PAS/PD-IPN hydrogel was strongly related to the microphase-separated domains within the hydrogel. [33,34]n addition to its superior elasticity, the ionic hydrogel skin also maintained its mechanical flexibility at subzero temperatures. [46]emperature-dependent dynamic mechanical analysis (DMA) in both tensile and compressive modes was conducted to investigate the anti-freezing property of the PAS/PD-IPN hydrogel.Figure 2j,k show that the storage modulus (E′) and loss modulus (E″) of the PAS/PD-IPN hydrogel remained stable when the temperature decreased from 50 to −10 °C.Further decreasing the temperature only slightly increased E′ and E″, indicating that the freezing point of the PAS/PD-IPN hydrogel was ≈−16 °C.This is consistent with the differential scanning calorimetry (DSC) results in Figure 2l.The exothermic peaks corresponding to water crystallization during the cooling process appeared at −20 and 0 °C for the PAS/PD-IPN and PAS-SN hydrogels, respectively, indicating that the PAS-SN hydrogel did not possess anti-freezing properties.Interestingly, negligible heat flow could be detected in the PD-SN hydrogel, indicating no ice formation. [46,47]Thus, the anti-freezing property of the PAS/PD-IPN hydrogel mainly arose from the polyelectrolyte in the IPN hydrogel.The PDADMAC network contained numerous ionic groups, which can strongly bind to water via ion-induced solvation. [21,48]These associated ions (salts) in the hydrogel depressed the freezing point of water, behaving similarly to an ion-induced anti-freezing mechanism reported for salt-added hydrogels. [49,50]Due to this anti-freezing property, the PAS/PD-IPN hydrogel showed similar stress-strain behavior when measured at ambient temperature (20 °C) and at a subzero temperature of −10 °C (Figure 2m).The resilience value of the PAS/PD-IPN hydrogel was higher than 96% up to strain of 300% at a subzero temperature of −10 °C (Figure 2n; Figure S16, Supporting Information).The PAS/PD-IPN hydrogel remained mechanically elastic and could be reversibly bent and twisted without fracturing even at −20 °C (Figure 2o).Moreover, the PAS/PD-IPN hydrogels exhibited superior water retention to the PAS-SN hydrogel (Figure S17, Supporting Information).

Ionic Conductivity and Electrophysiological Signal Sensing
To investigate the ion-transport properties of the PAS/PD-IPN hydrogel, the ionic conductivity and ionic Seebeck coefficient were measured.Figure 3a demonstrates the ionic conductivity of PAS/PD-IPN hydrogel measured at temperatures ranging from 25 to −10 °C.The ionic conductivity was obtained by fitting the electrochemical impedance spectroscopy (EIS) spectra to an equivalent circuit (Figure S18, Supporting Information).The PAS/PD-IPN hydrogel exhibited an ionic conductivity of 2.3 S m −1 at 25 °C, which was higher than that for PAS-SN, and PA/PD-IPN hydrogels (Figure S19, Supporting Information) and surpassed most previously reported conductive  S3, Supporting Information).While the ionic conductivity showed a gradual decrease along with the temperature, it remained at 0.9 S m −1 at −10 °C, indicating its antifreezing property.The ionic conductivity of the PAS/PD-IPN hydrogel under different stretching states was also characterized (Figure 3b).The results showed that the tensile strain had little impact on the ionic conductivity, and the PAS/PD-IPN hydrogel still delivered an ionic conductivity of ≈2.0 S m −1 at a tensile strain of 200%, revealing its resistance to strain perturbations.This was mainly attributed to the microphase-separated domains that acted as a hard phase that maintained the polymer's structural and functional integrity inside its domains upon stretching. [34]he combination of high ionic conductivity and superior elasticity makes the PAS/PD-IPN hydrogel an ideal electrode material for measuring electrophysiological signals (Figure 3c), such as electrocardiograms (ECGs) and electromyograms (EMGs).By using a commercial three-lead ECG sensing setup, the ECG waveforms were first recorded with our PAS/PD-IPN and PAS-SN hydrogel electrodes and then with a commercial Ag/AgCl gel electrode as a comparison.As shown in Figure 3d, the PAS/PD-IPN hydrogel electrode delivered high-quality ECG signals with characteristic P, Q, R, S, and T peaks identical to those obtained from a commercial Ag/AgCl gel electrode.The P wave, QRS complex, and T wave correspond to the activation of the atria, activation, and depolarization of the ventricles, and repolarization of the ventricles, respectively.A signal-to-noise ratio (SNR) of 36.9 dB was calculated from the ECG signals recorded by the PAS/PD-IPN hydrogel (calculation details in the Methods section), which was higher than that of the commercial Ag/AgCl gel electrode (≈24.3 dB) and PAS-SN hydrogel (≈25.2 dB) with lower ionic conductivity (Figure S20, Supporting Information).

hydrogels (Table
The mechanical robustness, low Young's modulus, and adhesion enabled the PAS/PD-IPN hydrogel to be firmly attached to the skin to measure ECG signals during exercise and long-term monitoring.Figure 3e shows that the heart rates of a human volunteer increased from 80 beats/min at rest to 120 beats/min after exercise.The ECG signals measured using the PAS/PD-IPN hydrogel electrode during skin stretching and squeezing all showed higher SNR values than those obtained from the commercial Ag/AgCl electrode (Figure 3f,g).During long-term monitoring, ECG signals collected from the PAS/PD-IPN hydrogel electrode showed only slightly lower SNR values after twelve h of continuous recording.These values were still much higher than the initial SNR value obtained using the commercial Ag/AgCl electrode (Figure S21, Supporting Information).In addition, the PAS/PD-IPN and PAS-SN hydrogel electrodes were adhered to the skin of a volunteer's forearm to monitor the EMG signals generated from muscle fibers.The hand motions of clenching and loosening a dynamometer with 5, 10, 20, 40, and 60 kg force were characterized (Figure 3h).The corresponding intensity of measured EMG signals increased with the gripping force (Figure S22, Supporting Information).The PAS/PD-IPN hydrogel electrode also produced higher-quality EMG signals and higher SNR values than the commercial Ag/AgCl electrode and PAS-SN hydrogel electrode (Figure 3i).

Ionic Thermoelectric Thermal Sensing Performance
Because the PAS/PD-IPN hydrogel skin is an ion conductor, it can form a concentration gradient of ions in response to a tem-perature differential (ΔT) to generate thermopower based on the Soret effect. [51]When ΔT formed in the PAS/PD-IPN hydrogel skin, the primary mobile ion diffused from the hot side to the cold side through the immobile polymer chains (Figure 4a). [52]he mobile ions were mainly transported through the immobile dimethyl diallyl ammonium groups of PDADMAC. [29,52]Therefore, we evaluated the ionic thermoelectric thermal-sensing properties of the PAS/PD-IPN hydrogel under different conditions.Two Peltier modules were connected in parallel to either heat or cool the PAS/PD-IPN hydrogel skin and were applied on opposite ends of the hydrogel film to generate a temperature gradient across it (Figure 4a).According to ionic thermoelectric mechanisms, the ionic Seebeck coefficient (S i ) of the ionic hydrogel was obtained by fitting the slope of open-circuit voltage (V therm ) versus temperature difference (ΔT) plots. [53,54]One Peltier module was used to control one side of the hydrogel film at a fixed initial temperature (T 0 ) and one Peltier module was used to apply a temperature (heating or cooling) on the other side of the hydrogel (T s ), thus generating ΔT across the hydrogel sample.As shown in Figure 4b inset, at a fixed T 0 of 20 °C and 50% relative humidity (RH), various ΔT of 14.0, 8.1, 3.1, −3.1, −8.7, and −13.8 °C were generated across the PAS/PD-IPN hydrogel.Their corresponding V therm reached stable values of 16.03, 10.02, 3.33, −3.78, −11.08, and −17.90 mV, respectively.The maximum S i of the PAS/PD-IPN hydrogel was calculated to be 1.21 mV K −1 (Figure 4b), which showed no major change (S i = 1.17 mV K −1 ), even when T 0 was decreased to −10 °C (Figure 4c).This indicates that the IPN ionic hydrogel skin had an anti-freezing property and exhibited a temperature sensitivity at both ambient and subzero temperatures.The S i value of PAS/PD-IPN hydrogel was much higher than that for PAS-SN hydrogel without the introduction of an ionic PD network, and comparable to that for PA/PD-IPN and PD-SN hydrogels (Figure S23, Supporting Information).
When a tensile strain was applied, the V therm values of the PAS/PD-IPN hydrogel remained relatively stable.The S i values of PAS/PD-IPN hydrogel at strains of 20 and 40% exhibited a very slight increase to 1.29 and 1.33 mV K −1 , respectively, from an initial value of ≈1.21 mV K −1 at 0% strain (Figure 4d,e).The stability of the S i upon cyclic stretch-release tests at 200% strain was also evaluated (Figure 4f).Remarkably, after subjecting the hydrogel to 200 stretch-release cycles, the S i value remained as high as 1.43 mV K −1 .These results indicate the ability of the PAS/PD-IPN hydrogel to retain its temperature sensitivity under large strain deformation, which mainly stems from its superior elasticity.Temperature resolution and response time are two other critical parameters for gauging the performance of the sensing material.As shown in Figure 4g, the PAS/PD-IPN hydrogel demonstrated a detection accuracy of ΔT = 0.1 °C with V therm = 0.10 and 0.12 mV under 0 and 100% strain, respectively.This temperature resolution was attributed to the high S i value of PAS/PD-IPN hydrogel in both the original and stretched state.Moreover, at a ΔT of 0.1 °C, the PAS/PD-IPN hydrogel exhibited a response time of ≈20 s under 0 and 100% strain (Figure 4g inset).
The PAS/PD-IPN hydrogel also exhibited stable and robust temperature sensing performance against pressure disturbances.Figure 4h presents the real-time output voltage and a PAS/PD-IPN hydrogel sensor at a temperature differential of 0 and 6.5 °C under the disturbance of pressure stimuli.The sensor was first monitored with no temperature differential applied for 900 s.In this neutral state, a series of pressures from 2.0 to 10.0 kPa were applied onto the hydrogel sensor but no variation in V therm was observed in real time.Following this monitoring period, the two ends of the sensing device were heated on one end to generate a ΔT = 6.5 °C in the subsequent 900 s.During the introduction of this temperature differential, V therm gradually increased and stabilized at ≈8 mV.In this heated state, a series of pressures were also applied onto the hydrogel sensor, and again no change in V therm was observed in real time.These results indicated that the PAS/PD-IPN hydrogel can discriminate temperature stimuli from pressure stimuli.
To evaluate the practical thermal sensing behavior of the PAS/PD-IPN hydrogel, we further measured the time-dependent variation in V therm by dropping water droplets at different temperatures (10, 15, 20, 25, and 30 °C) onto the PAS/PD-IPN hydrogel at 20 °C.V therm rapidly dropped by −1.21 and −0.20 mV after contacting the cold-water droplets (10 and 15 °C) due to the Soret effect.Then, it gradually increased due to the warming of the cold-water droplet over time and stabilized at 0 mV upon reaching thermal equilibrium (Figure 4i,j).When warm-water droplets (25 and 30 °C) were dropped onto the PAS/PD-IPN hydrogel, the instantaneous values of V therm increased to 1.01 and 2.03 mV.Thus, variations in thermal stimuli from water droplets could be detected by the different instantaneous V therm values.

Capacitive Pressure Sensing Performance
Next, to evaluate the electromechanical characteristics of the PAS/PD-IPN hydrogel when used as a capacitive pressure sensor, it was sandwiched between two flexible electrodes, which were both gold/polyimide (Au/PI) films (Figure 5a).Because the sensitivity of a capacitive pressure sensor mainly depends on the dielectric constant and deformation of the dielectric material, the high dielectric constant and low Young's modulus of the PAS/PD-IPN hydrogel can help increase the capacitive pressure sensitivity. [55]To further improve the sensitivity, the PAS/PD-IPN hydrogel was micropatterned (Figure 5b).In most conventional capacitive pressure sensors, the baseline capacitance (initial capacitance, C 0 ) is usually on the order of tens of femtofarads due to the low dielectric constant of the sensing material. [55]Such a low baseline capacitance leads to a low signal-to-noise ratio, which makes the capacitive sensor highly vulnerable to nearby parasitic effects and/or in the readout circuitry. [55]As shown in Figure 5c, the measured initial value of the PAS/PD-IPN hydrogel sensor was on the order of microfarads as the applied pressure increased from 1 to 400 Pa.These high capacitance values can be easily measured by using low-cost electronics. [56]To ensure reliable measurements, the C 0 was set at 1 μF to measure the capacitive pressure sensing performance.
The relative capacitance change (ΔC = C p -C 0 , C p corresponding to the capacitance under pressure) as a function of applied pressure of the PAS/PD-IPN hydrogel sensor was measured over a pressure range of 0-10.0 kPa, as shown in Figure 5d.The capacitive response curves contained three linear regions: 0-0.2, 0.2-2.0, and 2.0-10.0kPa.The corresponding linearity and sensitivity (S = (ΔC/C 0 )/P, P denotes the applied pressure) were calculated to be 0.94, 0.96, and 0.97, and 7.92 kPa −1 , 1.92 kPa −1 , and 0.74 kPa −1 , respectively.A higher sensitivity was obtained in the low-pressure region because the micropatterned surface of the PAS/PD-IPN hydrogel was more prone to a low pressure and the modulus of the hydrogel was low at small compressive strains. [57,58]Significantly, the sensitivity of 7.92 kPa −1 within the 0-0.2 kPa range and 1.92 kPa −1 within the 0.2-2.0kPa range of the PAS/PD-IPN hydrogel sensor surpassed those of most previously reported capacitive pressure sensors in the pressure range below 2.0 kPa (Table S4, Supporting Information).These sensitivity values for PAS/PD-IPN hydrogel were also much higher than those for PAS-SN and PA/PD-IPN hydrogel with higher Young's modulus (Figure S24, Supporting Information).
To evaluate the dynamic response of the hydrogel sensor, the PAS/PD-IPN hydrogel was subjected to cyclic transient pressure by increasing the loading from 0.1 to 10.0 kPa.As illustrated in Figure 5e, the capacitance response remained stable under various pressure loading and unloading cycles.To measure the sensing reliability of the hydrogel sensor, its sensing curves under various compression rates were measured (Figure 5f).Under a loading pressure of 0.1 kPa, the peak values of C p were unchanged upon increasing the loading rate from 1 to 10 mm/min.Moreover, the sensing stability of the PAS/PD-IPN hydrogel was investigated by subjecting the hydrogel sensor to long-term loading-unloading cycles (Figure 5g).The hydrogel sensor exhibited a stable and reliable capacitive response and withstood more than 1 000 repeated compressions up to 0.4 kPa.The response time to pressure stimuli was ≈150 ms (Figure 5h).These results reveal the sensing reliability, durability, stability, and accuracy of the PAS/PD-IPN hydrogel sensor.We further demonstrated the capacitive pressure sensing performance of the PAS/PD-IPN hydrogel sensor by successively dropping water droplets onto it.The results showed that the sensor could differentiate pressures induced by six water droplets applied one after another (Figure 5i).
The electromechanical robustness of the PAS/PD-IPN hydrogel sensor against strains and temperature perturbations was further investigated.The electromechanical characteristics of the hydrogel sensor were measured by applying a tensile strain to the PAS/PD-IPN hydrogel.Notably, as shown in Figure 5j, the pressure sensitivity of the hydrogel sensors remained virtually constant even at strains up to 50%, demonstrating the ability to against strain disturbance, which may root from low differential modulus for the hydrogel in this strain range.When the hydrogel sensor was under 100% strain, the sensitivity of the device only slightly decreased to 5.65 kPa −1 within the range of 0-0.2 kPa (Figure S25 and Table S5, Supporting Information).This value was still higher than those of most reported unstretched capacitive pressure sensors (Table S4, Supporting Information).The sensitivity of the hydrogel sensor even exhibited a mild increase to 8.68 kPa −1 within the range of 0-0.2 kPa after being subjected to 200 stretching-releasing cycles to a strain of up to 200% (Figure 5k; Table S5, Supporting Information).Figure 5l further demonstrates the ability to isolate pressure from temperature stimuli, illustrating the capacitive change of the sensing device changes only as a function of applied pressure, remaining independent of temperature variations from 20 to 30 °C.Moreover, compared with the test performed at ambient temperature (25 °C), the hydrogel sensor at 0 and −5 °C showed a lower capacitive response under stimulation by a high pressure (Figure S25b, Supporting Information).However, the sensitivity of the hydrogel sensor remained as high as 6.12 kPa −1 and 4.42 kPa −1 within the low-pressure range of 0-0.2 kPa at 0 and −5 °C, respectively (Table S5, Supporting Information).Such high sensitivity in the low-pressure range under subzero temperature was due to the excellent anti-freezing property of the PAS/PD-IPN hydrogel.

Proximity Sensing Performance
In addition to capacitive pressure sensing and ionic thermoelectric thermal sensing capabilities, the PAS/PD-IPN hydrogel skin could discern an approaching object, if the object is conductive or has a dielectric constant different from that of the air. [59]o determine the proximity sensing ability, a fringe field sensor was constructed by placing a pair of parallel electrodes on one surface of the PAS/PD-IPN hydrogel with an elastic insulating tape (VHB 4905, 3 M Inc.) laminated between one electrode and the hydrogel (Figure 6a).In theory, the capacitance change during proximity sensing originates from a disturbance in the fringing electric field.When an object approaches a fringing field sensor, the fringing electric field is intercepted, leading to a drop in mutual capacitance (C m ). [41]Figure 6b exhibits continuous C m changes as a conductive Cu bulk (10.0 × 3.0 × 0.2 cm) repeatedly approached from an initial distance (D 0 ) of 30 cm to various detection distances (D t ) from the sensor and then moved away back to D 0 .Interestingly, the PAS/PD-IPN hydrogel-based proximity sensor detected the conductive object as far away as 15 cm from the sensor.C m showed a clear drop as D t decreased.The proximity sensor could also detect dielec-tric objects such as solvents.Figure 6c illustrates the relative C m changes as a function of D t when a glass tube containing water (30 mL) approached the sensor.The relative C m change became more pronounced and reached −1.65% at D t = 0.5 cm (Figure 6c; Figure S26, Supporting Information).The relative capacitance changes at specific D m for PAS/PD-IPN hydrogelbased proximity sensor was much larger than those for PAS-SN hydrogel sensor (Figure S26, Supporting Information).This indicated that PAS/PD-IPN hydrogel with higher ionic conductivity exhibited better proximity sensitivity than that of PAS-SN hydrogel.Moreover, this proximity-sensing performance for PAS/PD-IPN exhibited almost unchanged as the hydrogel sensor was stretched to 50% strain (Figure 6c; Figure S27, Supporting Information), indicating its robust sensing performance against strain perturbation.The proximity-sensing ability of the PAS/PD-IPN hydrogel sensor was independent of the movement speed of target objects (Figure 6d).The response time to an approaching water droplet was as fast as ≈70 ms (Figure 6e).In addition to distance, the fringe field sensor was sensitive to changes in the dielectric constant (ɛ) of objects. [60]Continuous changes in C m were recorded when 11 types of solvents, including water, DMF, and ether, with different dielectric constants and a fixed volume, repeatedly approached and moved away from the PAS/PD-IPN sensor, as shown in Figure 6f

Discriminable Multiple Sensing Performance
Our PAS/PD-IPN hydrogel sensor can perceive proximity, temperature, and pressure stimuli with high sensitivity, high accuracy, fast response, and good reliability.[63] As shown in Figure 7a, a multimodal sensor was fabricated by simply patterning three pairs of electrodes onto the PAS/PD-IPN hydrogel, which endowed the sensing signals with three distinguishable outputs: a capacitive change for pressure sensing, a fringing field change for proximity sensing, and a thermal voltage change for temperature sensing.The uniqueness of our ionic hydrogel sensor is its ability to differentiate material types with different physical properties (i.e., density (), ɛ, or temperature).To demonstrate this, we used solvent identification as an example.Figure 7b shows the quantification process and different output signals when a solvent droplet with a fixed volume was dropped from a fixed distance onto the surface of the PAS/PD-IPN multimodal sensor.When the solvent droplet began to approach the sensor surface, the C m value monitored by the fringing field sensing unit in the multimodal sensor sharply decreased and finally reached a stable value immediately after the solvent contacted the sensor.When the solvent contacted the sensor, the capacitance monitored by the pressure-sensing unit instantaneously increased and stabilized due to the pressure applied by the solvent droplet.The thermal voltage output of the thermal-sensing unit gradually changed due to heat transfer between the solvent droplet and the sensor.
To demonstrate the material identification capability of our hydrogel-based multimodal sensor, we measured timedependent variations in C m , C p , and V therm by dropping two types of solvent droplets (0.5 mL) onto the sensor at different temperatures.The first group dropped onto PAS/PD-IPN sensor at T 0 = 30 °C included two types of solvents with similar ɛ values but very different  values: cyclohexane (ɛ = 2.02 F m −1 ,  = 0.77 g mL −1 , 20 °C) and dioxane (ɛ = 2.22 F m −1 ,  = 1.03 g mL −1 , 25 °C) (Figure 7c).While the output signals of C m from the fringing field sensing unit could not identify cyclohexane or dioxane due to their very similar ɛ values (Figure 7d), they could be differentiated by changes in the output signals of ΔC p from the pressure-sensing unit because of their large differences in  (Figure 7e).Moreover, variations in the initial temperature of the solvent droplets could also be detected by the difference in the instantaneous and equilibrated V therm values.As shown in Figure 7f, when cyclohexane (20 °C) and dioxane (25 °C) droplets were dropped onto the sensor at 30 °C, the V therm values decreased to −0.8 and −0.6 mV after 30 s, respectively.The other two solvents chosen for identification in the second group were cyclohexane (ɛ = 2.02 F m −1 ,  = 0.77 g mL −1 , 30 °C) and acetone (ɛ = 21.01F m −1 ,  = 0.79 g mL −1 , 35 °C), which have similar  values but had an order of magnitude difference in their ɛ values (Figure 7g).When cyclohexane and acetone droplets were dropped onto the sensor at 20 °C, their respective ultimate ΔC m values reached −0.45 pF and −0.75 pF, exhibiting a large difference in output signals, thus allowing them to be identified (Figure 7h).The output ΔC p signals from the pressure-sensing unit of both solvents were identical (Figure 7i) due to their similar  values.Moreover, variations in thermal stimuli from cyclohexane and acetone could be detected by differences in their V therm values (Figure 7j).These results show that our PAS/PD-IPN hydrogel sensor could detect and discriminate different types of materials without crosstalk due to its decoupled multimodal sensing ability.
To further demonstrate the real-world application potential of our hydrogel-based multimodal sensor on human skin, we attached the multimodal sensor onto a hand wrist to monitor the human arterial pulse, temperature change induced by cold object touching, and metal object approaching/leaving from the pressure sensing, temperature sensing, and proximity sensing units in the multimodal sensor, respectively (Figures S29-S31, Supporting Information).Considering the detectable pressure range of the PAS/PD-IPN pressure sensor (from 1.0 Pa to 10.0 kPa), the applicability of our multimodal sensor as artificial skin mainly focuses on monitoring subtle pressure signals, such as arterial pulse, and detecting gentle touch. [64]The soft nature of PAS/PD-IPN hydrogel can help the multimodal sensor tightly contact with soft human skin, allowing it to monitor tiny pulse signals precisely and stably from the pressure sensor (Figure S29, Supporting Information).Moreover, the V therm showed a rapid change when a cold object (10 °C) contacting the temperature sensing unit (20 °C) of the multimodal sensor, followed by reaching a thermal equilibrium after ≈50 s (Figure S30, Supporting Information).In addition, the proximity sensing unit in the multimodal sensor can detect the approaching and leaving of a metal object repeatedly (Figure S31, Supporting Information).The multimodal sensor would deform and stretch as the human skin stretched when hand wrist bending.The multimodal sensor exhibited stable and consistent pressure, temperature, and proximity sensing performance either the hand wrist was under relaxing state or bending state (Figures S29-S31, Supporting Information), indicating the ability of anti-strain disturbance for the PAS/PD-IPN hydrogel.

Conclusion
Here, we proposed a unique strategy for designing and fabricating an all-round multisensory ionic hydrogel with a microphaseseparated structure.The highly-entangled first network copolymerized with zwitterions, which allowed it to form strong interactions with the polyelectrolyte second network.These interactions facilitated the formation of a microphase-separated mi-crostructure in the ionic hydrogel, which contributed to its excellent mechanical properties including skin-like Young's modulus (< 60 kPa), high stretchability (> 900%), superior elasticity (resilience > 95%, hysteresis < 5%), strain-stiffening behavior, and excellent fatigue tolerance.The presence of zwitterions and a polyelectrolyte also gave the ionic hydrogel desirable moistureretention and anti-freezing properties.The ionic hydrogel retained a high ionic conductivity and ionic Seebeck coefficient under large tensile strain and subzero temperatures.Most importantly, the ionic hydrogel skin could detect and decouple multiple stimuli (temperature, pressure, and proximity) via an ionic thermoelectric effect, piezo-capacitive effect, and fringing field effect, respectively.The synergistic mechanical and physical effects gave the ionic hydrogel sensor a high capacitive pressure sensitivity of up to 7.5 kPa −1 with a detection limit of 1 Pa, a temperature sensing resolution of 0.1 °C, and excellent sensing robustness against large strain (> 200%) and subzero temperature perturbations.This ionic hydrogel sensor shows great potential for intelligent electronic skin applications, such as reliable health monitoring and accurate object identification.
Characterization: The morphologies of hydrogels were characterized at the Institute of Seawater Desalination and Multipurpose Utilization (MNR, Tianjin).A high-resolution field emission scanning electron microscopy (SEM, Hitachi Regulus 8220, Japan) equipped with a refrigerated transmission system (Quorum PP3010) was used.Hydrogels were first frozen using liquid nitrogen slush.Then, section-making and the sublimation process (−90 °C, 15 min) were carried out orderly followed by the in situ test using different magnifications.Nano computed tomography (Nano CT, Bruker SkyScan 2211, USA) was applied to visualize the solidliquid phase distribution.Fourier-transform infrared (FT-IR) spectroscopy was performed using a Thermo Scientific Nicolet IS50 FT-IR spectrometer in the wavenumber range of 4000-400 cm −1 to characterize the formation of a second network and interchain interactions inside the interpenetrating network.The interchain interactions between the dual-network were also confirmed by X-ray photoelectron spectroscopy (XPS, ESCALAB 250Xi, Thermo Scientific, USA).To evaluate the viscoelastic properties of the dual-network structure, dynamic mechanical analysis (DMA, TA Q800, USA) was conducted at a heating rate of 5 °C/min and a frequency of 1 Hz.Differential scanning calorimetry (DSC, DSC25, TA Instruments, USA) was used to determine the boundaries of the phase regions.The microstructure of hydrogel was observed using an ultra-depth three-dimensional microscope (LY-WN-YH8500, Chengdu Liyang, China).
Preparation of PAS-SN Hydrogel and PD-SN Hydrogel: Single-network hydrogels were prepared using a photo-polymerization method using I2959 as the initiator.First, 9.0 g AM and 3.0 g SBMA were dissolved in 18.0 mL water (molar ratio of AM/SBMA = 2175: 188), followed by the addition of 9.0 mg MBAA as the crosslinker and 270.0 mg I2959 as the photoinitiator.After ultrasonic dissolution, the solution was poured into a glass container and then sealed.After 365 nm UV irradiation for 1 h, the PAS-SN hydrogel was obtained.Similarly, to prepare PD-SN hydrogel, 20.0 g DADMAC, 100.0 mg MBAA, and 600.0 mg I2959 were dissolved in 24.4 g ultrapure water, and the method for preparing the PD-SN hydrogel was the same as above.
Preparation of PAS/PD-IPN Hydrogel: The PAS/PD-IPN hydrogel was prepared based on the first-prepared PAS-SN hydrogel.Briefly, the precursor solution of the second network was prepared with 80.0 g DADMAC, 0.4 g MBAA, 2.4 g I2959, and 97.ultrapure water (the concentration of DADMAC in the precursor solution was 45 wt.%).Then, the PAS-SN hydrogel with a width and length of 6.0 cm respectively was immersed in the prepared precursor solution for 24 h.Finally, the soaked hydrogel was polymerized under 365 nm ultraviolet irradiation for 1 h.The method for preparing the PAS/PD-IPN hydrogel with different contents of PD was the same as above.The molar ratio of AM:SBMA:DADMAC was optimized to be 2175: 188: 3760 in PAS/PD-IPN hydrogel.
Mechanical Property Tests: The mechanical properties under stretching and compression were tested by using a universal tensile testing machine (AGS-X, Shimadzu, Japan).For the stretching test, the hydrogel samples were stretched until fracture under a constant rate of 100 mm/min with an initial distance of 10 mm while the stress-strain curves were recorded at the same time.The conditions for the compression test were identical to those of stretching, except the rate was 20 mm/min.The conditions of other relevant tests during the stretching or compression process are illustrated in the manuscript.Due to its anti-freezing properties, we also evaluated the mechanical performance of PAS/PD-IPN hydrogel at 20 and −10 °C, and the test conditions were the same as those above.
Measurement of the Ionic Conductivity of PAS/PD-IPN Hydrogel: The ionic conductivity was evaluated under dynamic conditions, including different environmental temperatures (−10, −5, 0, 5, 15, and 25 °C) and different strains (0%, 50%, 100%, and 200%).The electrochemical impedance spectroscopy (EIS) analysis of the PAS/PD-IPN hydrogels was performed in a frequency range of 1-1 × 10 6 Hz using a CHI660 electrochemical workstation (Chenhua Co., China).The size of the PAS/PD-IPN hydrogel used for EIS detection was 10 × 10 × 2.6 mm.The ionic conductivity () was calculated using the following equation [54] : where d is the thickness of the hydrogel sample (cm); R is the body impedance of the measured sample (ohm), which was obtained from the intersection of the semi-arc and oblique lines in the Nyquist plot of the EIS spectrum; S is the effective cross-section area between the electrodes and hydrogel (cm 2 ).The ionic conductivity of the PAS/PD-IPN hydrogel at different temperatures was measured using a Peltier temperature control unit.To investigate the conductivity at different strains, the hydrogels were stretched to 0%, 50%, 100%, and 200% from an initial length of 2 cm under a constant temperature.
Fabrication of Microstructured IPN Hydrogel and Its Pressure-Sensing Performance: To amplify the pressure-sensing performance, a microstructure was introduced into the dual-network hydrogels using the same fabrication process as for the PAS/PD-IPN hydrogels.The only difference was that the prepared precursor solution was transferred to a glass container with one side pasted with sandpaper (120 mesh).The sensing performance was dramatically enhanced due to the more-sensitive structure deformation at the same pressure.The pressure-sensing properties of the hydrogel were characterized by applying an external load using a compression testing machine (AGS-X, Shimadzu, Japan), in combination with an LCR meter (TH2838H, Tonghui Electronic, China).This setup was used to record the capacitance values at a maximum voltage of 1 V and a sinusoidal signal of 1 kHz.The recognition of liquids was similar to the temperature difference detection.Water was selected for a demonstration and dropped onto the silver electrode.A dramatic increase in the capacitance was recorded using a TH2838H LCR with a frequency of 1 kHz.
Measurement of Non-Contact Proximity Sensing Performance: Solvents with different dielectric constants were encapsulated into identical glass test tubes and then fixed onto a motorized linear stage with a built-in con-troller from Zolix Inc.The solvents were cyclically moved over the hydrogel at a distance of 1 cm at a speed of 10 mm −1 s.For the cycling test, the distance and speed were set to 1 cm and 20 mm −1 s, respectively.For liquid sorting applications, different liquids encapsulated in glass tubes were moved circularly at a speed of 20 mm/min with a distance of 1 cm between the tube and electrodes.The capacitance was recorded using an LCR meter (TH2832, Tonghui Electronic, China) with a maximum voltage of 1 V and a 100 kHz sinusoidal signal.
Sensing Performance of PAS/PD-IPN Hydrogel under Dynamic Conditions: The evaluation of sensing performance under dynamic conditions included different environmental temperatures and different stretched states.The hydrogel-based sensors were characterized using three parameters (temperature, pressure, and proximity) at different temperatures (−10, −5, 0, 5, 15, 20, 25, and 30 °C) and strains (0-50% and 200%).For the temperature-variable sensing measurements, the hydrogel was first placed onto a temperature-control module, which was assembled and adjusted using a Peltier module to precisely control the temperature.For the sensing tests under stretching, the hydrogel was fixed onto a tensile platform.The thermovoltage and capacitance signals versus temperature or tensile condition were recorded, and the other test conditions were kept constant to provide a comparison.
Demonstration of Real-Time Solvent-Recognition Ability: For a real-time sensing demonstration, we compared dioxane, cyclohexane, and acetone because they possess different densities, dielectric constants, and temperatures.The scheme of the testing device is shown in Figure 7a.First, for proximity sensing, glass droppers containing 1 mL of different liquids at the same temperature were moved to approach the electrode quickly, and then the distance was kept at 1 cm.The capacitance was recorded by a TH2832 LCR with a frequency of 100 kHz.Then, for pressure sensing, 0.5 mL of the liquids was dropped, and the capacitance was recorded by a TH2838H LCR with a frequency of 1 kHz.Finally, during temperature sensing, the temperature was kept constant on one side, and then 0.5 mL of liquids at different temperatures were dropped onto the other side to form a temperature gradient.Then, the thermovoltage was recorded by a Keithley 2000 multimeter.The distance between the two electrodes and thermocouples was 5 mm.
Ethical Review for Committee Approval and Informed Written Consent of all Participants: Rules or permissions on wearable electronics from the relevant national or local authorities are not in place in the country where the experiments were performed.All experiments based on wearable electronics including ECG/EMG test, sensing demonstration upon human skin, and other related tests were conducted to verify the excellent sensing performances of PAS/PD-IPN hydrogel compared with others.All of these were conducted based on permissions obtained at the initial stage and in safety methods as well.During the test, detailed explanations of the sensing process, the need for cooperation, and any potential risks involved have been provided to guarantee a full understanding from all participants.In addition, informed written consent has also been obtained from the volunteers as supporting.

Figure 1 .
Figure 1.Design principle and characterization of PAS/PD-IPN hydrogel.a) Schematic diagram of the design and fabrication of the PAS/PD-IPN hydrogel.b) Photographs of the PAS/PD-IPN, PAS-SN, and PD-SN hydrogels.SEM images of c) PAS/PD-IPN and d) PAS-SN hydrogels.e) 3D Nano CT images of PAS/PD-IPN hydrogel based on differences in density.The blue and grey areas represent the heavy (polymer-rich region) and light (solvent-rich region) regions, respectively.Characterization of interchain interactions by f) FT-IR spectroscopy and (g) XPS.

Figure 2 .
Figure 2. Mechanical and anti-freezing properties of PAS/PD-IPN hydrogel.a) Tensile stress-strain curves of PAS-SN, PD-SN, and PAS/PD-IPN hydrogels at 100 mm/min.b) Frequency-dependent DMA result of the PAS/PD-IPN hydrogel at 20 °C.c) Loading-unloading tensile curves of PAS/PD-IPN hydrogels at strains ranging from 50 to 400% at a deformation rate of 100 mm/min.The curves are horizontally offset for clarity.The final strains are shown on the curves.Insets show the superposition of the stretching-relaxation curves at different strains.d) Resilience of PAS-SN, PA/PD-IPN, and PAS/PD-IPN hydrogels at each strain cycle.e) Loading-unloading tensile curves of PAS/PD-IPN hydrogel at strain rates from 20 to 200 mm/min at 200% strain.The curves are horizontally offset for clarity.The stretching rates are shown on the curves.Insets show the superposition of the stretching-relaxation curves at different stretching rates.f) Resilience and hysteresis of PAS/PD-IPN hydrogels at each strain cycle under different strain rates.g) 1 000 loadingunloading tensile cycles of PAS/PD-IPN hydrogel at 100% strain.The curves are horizontally offset for clarity.The cycling numbers are shown on the curves.Insets show the superposition of the stretching-relaxation curves for specific cycling number.h) Compressive stress-strain curves of PAS/PD-IPN hydrogel at strains ranging from 20 to 80%.i) 200 loading-unloading compressive cycles of PAS/PD-IPN hydrogel at 80% strain.Temperature-dependent DMA results under j) tensile and k) compressive mode for PAS/PD-IPN hydrogels at a frequency of 1 Hz at temperatures from −30 to 50 °C.l) DSC curves of the hydrogels.m) Tensile stress-strain curves for PAS/PD-IPN hydrogels measured at 20°C and −10 °C.n) Loading-unloading tensile curves of PAS/PD-IPN hydrogels at strains ranging from 100% to 300% measured under −10 °C.The curves are horizontally offset for clarity.The final strains are shown on the curves.Insets show the superposition of the stretching-relaxation curves at different strains.o) Photographs of PAS/PD-IPN hydrogels being bent and twisted at −20 °C.

Figure 3 .
Figure 3. Ionic conductivity and application of PAS/PD-IPN hydrogel for electrophysiological signal detection.a) Ionic conductivity of the PAS/PD-IPN hydrogel measured at temperatures from 25 to −10 °C.b) Ionic conductivity of IPN ionic hydrogel measured under different stretching states.c) Schematic illustration of the IPN hydrogel used to detect electrophysiological signals, including ECG and EMG.d) Comparison of ECG signals between the PAS/PD-IPN hydrogel and commercial Ag/AgCl gel electrodes.e) ECG signals detected by PAS/PD-IPN hydrogel electrodes at rest and after exercise.f) Comparison of the ECG signals between the PAS/PD-IPN hydrogel and commercial electrodes in the original, skin squeezing, and skin stretching states and g) corresponding SNR values.h) Comparison of EMG signals among PAS-SN hydrogel, PAS/PD-IPN hydrogel, and commercial Ag/AgCl gel electrodes when clenching and loosening a dynamometer by applying 5, 10, 20, 40, and 60 kg force by hand and i) corresponding SNR values.

Figure 4 .
Figure 4. Thermal sensing performance of PAS/PD-IPN hydrogel.a) The schematic illustration of the ion thermal diffusion in the PAS/PD-IPN hydrogel at various temperature differentials.b) Thermovoltage of the PAS/PD-IPN hydrogel as a function of ΔT at T 0 = 20 °C at relative humidity (RH) of 50%.Inset: thermovoltage change versus time at T 0 = 20 °C at RH = 50%.c) Thermovoltage of the PAS/PD-IPN hydrogel as a function of ΔT at T 0 = −10 °C at RH = 50%.Inset: thermovoltage change versus time at T 0 = −10 °C at RH = 50%.d) Thermovoltage of the PAS/PD-IPN hydrogel as a function of ΔT under strains of 0%, 10%, 20%, 30%, and 40%.e) The average S i values for PAS/PD-IPN hydrogel under strains of 0%, 10%, 20%, 30%, and 40%.More than five samples were measured for each strain state, shaded regions correspond to max and min S i values at each strain state.f) Thermovoltage as a function of ΔT for the PAS/PD-IPN hydrogel after 200 stretch-release cycles between 0-200% strain.Inset: thermovoltage change versus time for the PAS/PD-IPN hydrogel after 200 stretch-release cycles between 0-200% strain.T 0 and RT were fixed at 20 °C and 50%, respectively.g) Thermovoltage response to ΔT = 0.1 °C for the PAS/PD-IPN hydrogel.Inset: thermovoltage change was ≈0.1 mV, and the response time was < 20 s.T 0 and RT were fixed at 20 °C and 50%, respectively.h) Real-time output voltage of a PAS/PD-IPN hydrogel sensor at temperature differential of 0 and 6.5 °C against various pressure disturbances.i) Schematic illustration for detection of liquid with different temperatures.j) Time-dependent variation of thermovoltage when water with different temperatures was dropped onto the PAS/PD-IPN hydrogel at T 0 = 20 °C.

Figure 5 .
Figure 5. Capacitive pressure sensing performance of PAS/PD-IPN hydrogel.a) Schematic illustration of a capacitive pressure sensor based on PAS/PD-IPN hydrogel.b) Ultra-depth 3D microscopy image of the micropatterned surface of PAS/PD-IPN hydrogels.c) Capacitance changes over time of a PAS/PD-IPN hydrogel sensor in response to a gradient static pressure in the range of 1-400 Pa.d) Capacitive response as a function of applied pressure and a corresponding sensitivity to different pressures of the PAS/PD-IPN hydrogel sensor.The inset shows the magnified view of the sensitivity curve below 0.2 kPa with a calculated sensitivity of 7.92 kPa −1 .e) The capacitance change over time of the PAS/PD-IPN hydrogel sensor by varying the pressure in the range of 0.1-10.0kPa.f) Relative capacitance changes of PAS/PD-IPN hydrogel sensor under incremental loading rate cycling between 0 and 0.1 kPa.g) Capacitance changes of the PAS/PD-IPN hydrogel over 1 000 pressure loading-unloading cycles with a loading pressure of 0.4 kPa.Inset: Detailed capacitance change curves recorded between 3 000 and 3 100 s. h) Instant capacitive response of the PAS/PD-IPN hydrogel sensor exhibiting a response time of ≈150 ms during pressure loading.i) Detection of six successive water drops on the PAS/PD-IPN hydrogel sensor.j) Capacitive response as a function of applied pressure to the PAS/PD-IPN hydrogel under tensile strains of 0%, 30%, and 50%.Lines depict the average curve (N > 5), shape regions correspond to max and min capacitive changes at different pressures.k) Capacitive response as a function of applied pressure for the PAS/PD-IPN hydrogel before and after 200 stretching-releasing cycles between 0-200% strain.l) Capacitive response as a function of applied pressure of the PAS/PD-IPN hydrogel at 20, 25, and 30 °C.

Figure 6 .
Figure 6.Proximity sensing performance of PAS/PD-IPN hydrogel.a) Schematic illustration of a fringe field proximity sensor based on PAS/PD-IPN hydrogel.b) Continuous changes in mutual capacitance when a Cu bulk repeatedly approached to various D t = 3, 5, 10, and 15 cm and were moved away from the PAS/PD-IPN hydrogel sensor.c) Relative capacitance changes as a function of D m when the PAS/PD-IPN hydrogel sensor was subjected to 0% and 100% strain.d) Mutual capacitance changes when a glass tube containing water repeatedly approached from 10 to 1 cm away from the hydrogel sensor at incremental speeds.e) Instant mutual capacitive response when the tube containing water approached from 10 to 1 cm away from the PAS/PD-IPN hydrogel sensor.f) Continuous changes in mutual capacitance when water, DMF, and ether repeatedly approached and were moved away from the PAS/PD-IPN hydrogel sensor.g) The recorded mutual capacitance values of various solvents at D t = 1 cm.
and Figure S28 (Supporting Information).The C m changes of water (ɛ = 80.1), DMF (ɛ = 38.3),and ether (ɛ = 4.27) at D t = 1 cm were measured to be −0.125,−0.029, and −0.0074, respectively, demonstrating that the sensor could differentiate solvents with different ɛ values in a non-contact mode.The ɛ and C m relationship for 11 typical solvents (Figure S28, Supporting Information) is summarized in Figure 6g.The measured C m changes of most solvents were well fitted to their intrinsic ɛ, except for solvents with very similar ɛ values, such as cyclohexane (2.02) and dioxane (2.22).

Figure 7 .
Figure 7. Multiple sensing for substance identification.a) Schematic illustration of a multimodal sensor based on PAS/PD-IPN hydrogel.b) Schematic illustration of the decoupled multiple sensing mechanisms for fringing field proximity sensing, capacitive pressure sensing, and ionic thermoelectric thermal sensing.c) Dielectric constant and mass density for cyclohexane and dioxane.The PAS/PD-IPN hydrogel multi-sensor (T 0 = 30 °C) d) used to detect approaching cyclohexane and dioxane droplets via fringing field proximity sensing, e) detection of contacting cyclohexane and dioxane droplets by capacitive pressure sensing, and f) detection of the temperature of cyclohexane (20 °C) and dioxane (25 °C) by ionic thermoelectric thermal sensing capabilities.g) Dielectric constant and mass density of cyclohexane and acetone.Use of the PAS/PD-IPN hydrogel multi-modal sensor (T 0 = 20 °C) h) to detect the approach of cyclohexane and acetone droplets by fringing field proximity sensing, i) for detection of contacting cyclohexane and acetone droplets by capacitive pressure sensing, and j) for detection of the temperature of cyclohexane (30 °C) and acetone (35 °C) by ionic thermoelectric thermal sensing.