Enhanced high-strength, temperature-resistant PVA hydrogel sensors with silica/xanthan/glycerol for posture monitoring and handwriting recognition using deep learning†

Fanchen Luo, Yafei Qin*, Xi Wang, Xuanmo Zhao, Kedi Chen and Weichen Huang
Faculty of Mechanical and Electrical Engineering, Kunming University of Science and technology, Kunming 650500, China. E-mail: qinyafei_kmust@foxmail.com

First published on 20th August 2024

---

---

1. Introduction

Compared to traditional rigid sensors, flexible sensors offer inherent advantages such as greater flexibility, scalability, and biocompatibility. While conventional rigid sensors are highly effective in terms of gauge factor and sensitivity, they struggle to conform precisely to complex, flexible, curved interfaces. Flexible sensing materials, however, provide a solution to these limitations.^1–3 Among the various flexible sensing materials, hydrogels distinguish themselves with their wide application prospects and numerous advantages.^4–6 However, to become ideal flexible sensing materials, hydrogels must overcome specific challenges, including the trade-off between high conductivity and superior mechanical properties,^7,8 long electrical response times,⁹ and variations in electrical and mechanical performance under extreme temperature conditions.^10,11 Nonetheless, with a Young's modulus similar to that of human skin, hydrogels continue to demonstrate significant potential in the fields of human health monitoring and electronic skin.^12–14

Conductive hydrogels are generally classified into two types: electronic and ionic.^15,16 Electronic-type conductive hydrogels typically incorporate nano-conductive fillers, such as carbon nanotubes and graphene oxide. Hydrogels infused with these nano-fillers generally exhibit enhanced mechanical properties, larger specific surface areas, and improved conductivity. However, carbon-based nano-fillers often demonstrate toxicity to human cells, rendering them unsuitable for skin-friendly wearable electronics and implants. In contrast, the conductivity of ionic conductive hydrogels is achieved through the migration of ions within the hydrogel, making this mechanism ideal for sensing and monitoring physiological activities. This ionic conductivity also offers better skin compatibility compared to electronic conductive hydrogels.

The primary materials for ionic conductive hydrogels generally include polyvinyl alcohol (PVA),^17,18 polyacrylamide (PAM),^19,20 and polyethylene glycol (PEG).^21,22 Notably, PVA is characterized by its robust mechanical properties, including high elastic modulus and tensile strength, enhanced biocompatibility, and superior water absorption capacity. PVA hydrogels are synthesized via chemical crosslinking, using agents like boric acid,²³ has been restricted due to the resultant toxicity to human cells. In contrast, PVA hydrogels crafted through freeze–thaw cycles not only retain their mechanical excellence but also exhibit remarkable cell viability.^24,25 For instance, Miao et al. fabricated MXene/PVA hydrogels that exhibited exceptional long-term lubricity and load-bearing capabilities, attributed to the salting-out effects and a multimodal crosslinking approach.²⁶ In a study by Li et al., a conductive ionic hydrogel was created by combining PVA and ethylene glycol with MgCl₂ after repeated freezing and thawing. This hydrogel demonstrated remarkable resilience, withstanding over 6000 times its own weight and exhibiting excellent fatigue resistance.²⁷ Despite these advancements, the application of PVA-based hydrogel composites in extreme temperature conditions remains relatively unexplored.

At low temperatures, hydrogels are prone to freezing due to their high-water content, while at high temperatures, evaporation can cause the hydrogel network to shrink. Thus, the limited availability and stability of ionic conductive hydrogels restrict their application in flexible sensors, particularly under extreme temperature conditions.²⁸ Currently, introducing inorganic salts to lower the freezing point of water in hydrogels is a common strategy to enhance their resistance to extremely low temperatures. Zhou et al. synthesized antifreeze hydrogels with dual networks of polyvinyl alcohol (PVA) and polyacrylamide (PAM), doped with xanthan gum and zinc chloride (ZnCl₂) through chemical covalent bonding and physical crosslinking. These hydrogels can operate continuously at −60 °C while maintaining excellent mechanical properties.²⁹ Liu et al. achieved high optical transparency and durability at elevated temperatures (65 °C) by developing a dual-network hydrogel comprising glycerol-crosslinked PVA and sodium alginate crosslinked with Ca²⁺.³⁰ Significant progress has been made in the development of hydrogels that are simultaneously resistant to both high and low temperatures. However, challenges remain. For example, Fu et al. developed a physically cross-linked hydrogel by combining PVA with silk fibroin, sodium citrate, and glycerol. This hydrogel demonstrated usability across a temperature range of −25 °C to 60 °C and exhibited exceptionally high tensile strength. However, it only achieved 200% elongation at −25 °C, and its gauge factor (GF) was relatively low.³¹ Li et al. synthesized a poly (acrylamide-acrylic acid)/chitosan/MXene hydrogel via radical polymerization. This hydrogel exhibits excellent mechanical and self-healing properties, along with reliable structural integrity across a broad temperature range of −20 °C to 80 °C.³² Shu et al. developed ICHs hydrogels by dissolving cellulose and polyvinyl alcohol in an inorganic solution, with zinc ions imparting a high conductivity of 8.16 S m⁻¹ and a usable temperature range of −60 °C to 20 °C. However, the hydrogel exhibited poor tensile strength and elongation (0.30 MPa, 130%),³³ which could limit its potential for applications in wearable electronics. In comparison to these studies, the PLXSG hydrogel offers a combination of high conductivity and excellent mechanical properties while maintaining a wide operational temperature range.

In this study, PLXSG ionic conductive hydrogels were synthesized using a freeze–thaw cycling method with a water/glycerol binary solvent. These hydrogels demonstrated remarkable mechanical strength and were capable of withstanding extreme temperature ranges (−40 °C to 70 °C). The addition of silica nanoparticles increased the crosslinking density of the hydrogel network, thereby enhancing its mechanical properties. Moreover, the formation of numerous reversible hydrogen bonds between xanthan gum and PVA macromolecular chains improved the hydrogel's tensile fracture strength. Glycerol not only enhanced the hydrogel's flexibility but also formed hydrogen bonds with water molecules, converting a significant amount of free water into bound water, thus improving water retention at high temperatures. Additionally, the hydration of lithium ions with water molecules in the hydrogel prevented ice crystal formation at low temperatures, effectively lowering the freezing point of the water within. The migration of lithium ions endowed the hydrogel with an ionic conductivity of 0.767 S cm⁻¹ and, as a flexible sensor, it exhibited a high sensitivity of 3.623 at strains exceeding 175%. The hydrogel has been successfully utilized in various applications, including human motion posture monitoring, machine learning-enhanced handwriting recognition, sensor arrays, and Morse code-based information protection.

2. Experimental

2.1 Materials

Polyvinyl alcohol (PVA; 99.0–99.5% Alcoholysis), xanthan gum (XG, 1200–1400 MPa s), anhydrous lithium chloride (LiCl, 99.0%), and silica nanoparticles (SiO₂, 20 ± 5 nm, 99.5%) were procured from Macklin (Shanghai, China). Additionally, glycerol (Gly, AR, 99.0%) was obtained from China Pharmaceutical Group Chemical Reagent Co, Ltd (Shanghai, China). These chemicals were utilized as received without further purification. Deionized water served as the solvent throughout the experimental procedures.

2.2 Preparation of PVA/LiCl/XG/SiO₂/glycerol (PLXSG) hydrogels

SiO₂ nanoparticles were initially combined with DL in varying mass fractions (1%, 5%, 10% relative to PVA mass) to create an inhomogeneous mixture. This mixture was ultrasonicated at 20% power for 20 minutes to achieve dispersion. Subsequently, glycerol was added in a 1:1 ratio with DL, as described in literature,³⁴ and magnetically stirred for 20 minutes to ensure uniformity. PVA powder (0.94 g) was then added, and the solution was stirred for 2 hours at 90 °C until the PVA was fully dissolved. Xanthan gum (XG) was incorporated into the mixture at a 95:5 mass ratio relative to PVA.³⁵ The mixture was then supplemented with 2M LiCl and stirred for an additional 30 minutes, yielding the PVA/SiO₂/XG/LiCl hydrogel solution, with water and glycerol as binary solvents. The resulting hydrogel solution was cast into PTFE molds of various geometries and allowed to equilibrate to ambient temperature. The filled molds were subjected to a freeze–thaw cycle, consisting of a 12-hour freeze at −20 °C followed by a 10-hour thaw at room temperature; this cycle was repeated three times. The final product, designated as PLXSG hydrogel (X = 0%, 1%, 5%, 10%), was thus prepared. The comprehensive synthesis protocol is depicted in Fig. S1 (ESI†). The compositions for the preparation of each type of hydrogel are detailed in Tables S6–S8 (ESI†).

2.3 Characterization

Specimen preparation entailed an initial rapid quenching in liquid nitrogen, followed by lyophilization to render them amenable for detailed analysis. The freeze-dried samples were then sputter-coated with a metallic layer to facilitate electron microscopy. Subsequent surface microstructural examination was conducted using a field emission scanning electron microscope (SEM, model Volumescope2, Thermo Scientific, USA), operating at an acceleration voltage of 5 kV. Complementing the SEM analysis, the samples were pulverized into a fine powder and thoroughly amalgamated with potassium bromide. The final step involved the characterization of the hydrogel samples’ infrared spectra, spanning a wavenumber range of 4000 to 500 cm⁻¹, utilizing a Fourier transform infrared spectrometer (FTIR, model EQUINOX-55, Bruker Optics, Germany).

2.4 Mechanical tests

Tensile properties of the hydrogel samples were quantified using a general-purpose tensile tester (model ZQ-950B, Dongguan, China). The specimens, fashioned into a dumbbell shape, measured 30 mm in length, 2 mm in width, and 1 mm in thickness. Each sample was uniaxially stretched at a constant rate of 200 mm min⁻¹ until failure. To ensure statistical reliability, a minimum of five replicate tests were conducted per sample, with the mean value reported as the definitive result. The modulus of elasticity was deduced from the linear portion of the stress–strain curve, specifically between 10% and 40% strain. Additionally, the toughness of the hydrogels was ascertained by computing the area under the stress–strain curve, expressed in MJ m⁻³.

2.5 Sensing performance test

The hydrogel's sensing capabilities were evaluated using a universal tensile tester (ZQ-950) and an LCR digital bridge (TH 2830). Before starting the test, the hydrogel sensor (30 mm × 10 mm × 1 mm) was attached to the electronic tensile testing machine. The sensor's ends were then wired to the digital bridge, which was used to determine the hydrogel sensor's resistance. The relative resistance was computed with the subsequent formula:

	(1)

where (ΔR) denotes the relative resistance change, (R) is the resistance measured during the experiment, and (R₀) is the baseline resistance without applied force.

The hydrogel's sensitivity was defined by the gauge factor (GF), calculated using the equation:

GF = (ΔR/R0)/ε	(2)

where (ΔR) is the relative resistance change, (R₀) is the initial resistance, and (ε) represents the tensile strain experienced by the hydrogel sensor.

2.6 Conductivity test

The hydrogel's conductivity was evaluated by measuring the impedance between two metal electrode plates using an LCR digital bridge at a frequency of 1 kHz. The conductivity (σ) was determined by the equation:

σ = L/(R × A)	(3)

In this formula, (L) signifies the distance between the electrodes in centimeters (cm), (R) is the hydrogel's volume resistance in ohms (Ω), and (A) is the sample's cross-sectional area in square centimeters (cm²).

2.7 High and low temperature resistance test

The resistance of hydrogels to high and low temperatures was evaluated using a constant temperature and humidity tester (408L, Corus Instruments, China) at 40% RH humidity. Thermal images of the hydrogels at various temperatures were captured using a thermal imaging camera (UTi260E, Unidux, China). The hydrogel's resistance to drying was evaluated by measuring the mass change after 24 hours in different temperature conditions (20 °C, 40 °C, 70 °C). The residual mass ratio of the hydrogel was determined using the following formula:

Weight ratio = (wd/w0) × 100%	(4)

In this context, the symbol w₀ represents the initial mass of the hydrogel prior to any moisture loss, while the symbol w_d denotes the mass of the hydrogel at a specific time point following the loss of moisture.

3. Discussion

3.1 Design and characterization of hydrogels

Fig. 1(a) and (b) demonstrate that PLXSG hydrogels develop crystalline domains within the primary PVA network following freeze–thaw cycles. Numerous hydrogen bonds form between XG and PVA macromolecules. Simultaneously, the silanol groups on SiO₂ nanoparticle surfaces and the hydroxyl groups on PVA interlink via hydrogen bonds, serving as sacrificial bonds to dissipate energy under external forces, thereby enhancing the hydrogel's flexibility. Additionally, a significant quantity of PVA molecular chains adsorb around the silica nanoparticles through these hydrogen bonds, creating dynamic physical cross-linking points. Increased silica concentration raises the network's cross-linking density, thereby bolstering mechanical strength. The presence of glycerol introduces additional hydrogen bonds within the hydrogel, imparting superior elasticity. Furthermore, the addition of lithium chloride introduces abundant free ions, conferring excellent electrical conductivity and sensitivity to the hydrogel. The combined incorporation of glycerol, lithium chloride, and silica nanoparticles during hydrogel synthesis results in a synergistic network resilient to temperature extremes. Glycerol, through its hydroxyl groups, forms hydrogen bonds with water molecules, enhancing moisture retention and heat resistance, allowing the hydrogel to remain pliable and elastic at low temperatures, thereby preventing brittleness. Moreover, lithium chloride lowers the freezing point, inhibits ice crystal growth, and enhances frost resistance. Silicon dioxide nanoparticles strengthen mechanical properties and thermal stability, creating a network with high thermal conductivity for rapid heat dissipation, thus preventing excessive heat accumulation and enhancing heat resistance. This tripartite synergy ensures the hydrogel maintains commendable physical properties and stability under extreme temperatures.

As illustrated in Fig. 1(c), the FTIR spectrograms of PL, PLX, PLXS, and PLXSG hydrogels were analyzed to investigate the direct interactions between the components. The characteristic absorption peaks of the hydrogels correspond to C–O stretching vibrations at a wavenumber of approximately 1059 cm⁻¹ and O–H stretching vibrations at approximately 3280 cm⁻¹. The changes in the intensity of the C–O and O–H peaks indicate the extent of hydrogen bonding present in the hydrogel network. The addition of xanthan gum to PL resulted in a slight change in peak position, suggesting minimal hydrogen bonding. The incorporation of silica nanoparticles led to a notable shift in both the C–O and O–H peaks. The C–O peak shifted from 1085 cm⁻¹ to 1051 cm⁻¹, while the O–H peak shifted from 3282 cm⁻¹ to 3280 cm⁻¹, indicating significant hydrogen bonding. Subsequent addition of glycerol to PLXS caused a further shift of the C–O peak to 1031 cm⁻¹ and the O–H peak to 3265 cm⁻¹, confirming extensive hydrogen bonding and the formation of a synergistic network. Furthermore, the absence of new peaks in the hydrogel components indicates that no chemical changes have occurred.

Fig. 1(d) and (e) display the surface SEM images of two PVA hydrogels magnified 5000 times. The pure PVA hydrogel surface in Fig. 1(d) appears smoother with fewer microscopic pores, suggesting a simpler, more continuous microstructure. In contrast, with the addition of silica nanoparticles, as depicted in Fig. 1(e) and Fig. S2(a), (b) (ESI†), the hydrogel surface exhibits more distinct and densely arranged pores, with the particle distribution contributing to the surface's complexity and heterogeneity. These structural alterations may influence the hydrogel's physical and chemical characteristics, potentially enhancing mechanical strength and responsiveness. Fig. S3 (ESI†) presents a scanning electron microscope image of a PLXSG hydrogel with silica nanoparticles and glycerol. Glycerol, significantly impacts the hydrogel's microstructure. The images reveal that glycerol's presence accentuates the pore structure and improves pore interconnectivity. Additionally, the incorporation of silica is evident in the SEM image of the hydrogel's cross-section, which shows an intricate microscopic pore structure (Fig. S2(c)–(e), ESI†).

3.2 Mechanical properties of hydrogels

In the realm of human motion monitoring and electronic skin applications, the superior mechanical performance of conductive hydrogels is crucial for the stable functioning of strain sensors. For hydrogel electronic devices worn on the human body, a Young's modulus close to that of human skin (∼0.9 MPa) is essential for ensuring comfortable wear. The PLXSG hydrogel, with a Young's modulus of 1.754 MPa at room temperature, closely matches this ideal value, thereby guaranteeing both stability and comfort during use.^36,37 Leveraging crystalline domains formed through freeze–thaw cycles and a synergistic hydrogen-bond network among its components, the PLXSG hydrogel exhibits excellent resilience under 100% strain and adaptability to various deformations and torsions (Fig. 2(a)). Our investigation into the effects of different components and concentrations on the mechanical properties of PLXSG (Fig. 2(b), (c) and Fig. S3–S5, ESI†) revealed that it possesses optimal mechanical characteristics, including a maximum tensile strength of 1.82 MPa and an ultimate elongation of 796.3%. In terms of Young's modulus and toughness, PLXSG also outperforms other formulations, achieving values of 1.768 kPa and 6.09 MJ m⁻³, respectively. Glycerol, acting as a ‘lubricant’ between molecular chains, not only imparts greater flexibility to PLXSG compared to PLXS but also, with its three hydroxyl groups, forms dynamic cross-linking points within the PVA network's synergistic hydrogen-bond network, further enhancing the hydrogel's mechanical strength.³⁸ The incorporation of SiO₂ nanoparticles increases the crosslinking density and the number of hydrogen bonds, significantly boosting the mechanical performance of PLXS over PLX. The addition of xanthan gum enriches the hydrogel network's hydrogen-bond connections,³⁹ providing PLX with exceptional toughness and strength. Meanwhile, the inclusion of lithium ions enhances the stretchability of the PL hydrogel compared to pure PVA hydrogel, albeit with a slight decrease in tensile strength.

In our previous discussion, we explored the impact of various material compositions on the mechanical properties of hydrogels. We now focus on the specific influence of different concentrations of silicon dioxide (SiO₂) on the material's mechanical performance. Using PLXG hydrogel without added SiO₂ nanoparticles as a control, we conducted uniaxial tensile tests on hydrogels with SiO₂ contents of 1 wt%, 5 wt%, and 10wt% (ratio of SiO₂ nanoparticle mass to PVA mass) (Fig. 2(e) and (f)). The data revealed that as SiO₂ concentration increased, both maximum elongation at break and maximum stress also increased. However, at a 10wt% addition level, the maximum elongation at break decreased by 193% compared to the 5wt% addition, while the maximum stress reached a peak of 2.61 MPa, with a Young's modulus of 4.336 kPa and the highest toughness of 8.579 MJ m⁻³.

This outcome is likely due to the increased SiO₂ content, which led to a rapid rise in the crosslinking density of the hydrogel network, thereby significantly enhancing the hydrogel's mechanical strength. However, the excessive entanglement of PVA molecular chains disrupted the previously stable synergistic hydrogen-bond network, causing a sharp decline in elongation. Although the PLXS₁₀G hydrogel with 10wt% SiO₂ exhibited the best tensile strength and highest toughness, its high Young's modulus and reduced elongation limit may constrain its effectiveness as a flexible strain sensor. Consequently, the PLXS₅G hydrogel with a 5wt% SiO₂ addition was selected as the optimal formulation, balancing good tensile strength and toughness with the best elongation limit.

To analyze the energy dissipation behavior of the PLXSG hydrogel, load–unload cyclic experiments were conducted at strains ranging from 100% to 300%, as illustrated in Fig. 2(d). The enclosed area within the hysteresis loop represents the energy dissipated by the hydrogel in counteracting deformation under external force. As the strain increased, the internal PVA molecular chains began to unravel, and hydrogen bonds broke, resulting in irreversible damage to the crosslinking points and an enlargement of the hysteresis loop area. Additionally, the overlap between successive hysteresis loops indicates that the PLXSG hydrogel exhibits partial recovery immediately after the removal of the external force.

Following the same treatment, the PLXSG hydrogel demonstrated remarkable mechanical stability after 10 cycles of stretching at both 100% and 200% strain (Fig. 2(g) and Fig. S6, ESI†). Aside from a noticeable change in the enclosed area of the loop during the first cycle, the remaining nine cycles largely overlapped. Notably, the area of the hysteresis loop during the first cycle at 100% strain was almost negligible, indicating the PLXSG hydrogel's excellent fatigue resistance and recovery within 100% strain. This behavior may be attributed to the interlinking and entanglement of PVA macromolecular chains with interpenetrating nanoparticles, as well as an efficient energy dissipation mechanism. When the hydrogel is subjected to external force, the PVA polymer chains on the surface of the SiO₂ nanoparticles can reversibly reorient, thereby dissipating energy.

3.3 High and low temperature durability of hydrogel

As demonstrated in Fig. 3(a), four different hydrogel compositions were placed in a constant temperature and humidity chamber set at 70 °C. After 24 hours, the PL and PLX hydrogels fractured under minimal tensile force. In contrast, the PLXG hydrogel remained intact even under greater tensile force, while the PLXSG hydrogel exhibited excellent resilience, withstanding substantial tensile force. This conclusively demonstrates that the addition of glycerol prevents the hydrogel from drying out at high temperatures, and the incorporation of silicon dioxide further enhances this effect. As shown in Fig. 3(b), after being frozen at −40 °C for 24 hours, PX and PXS hydrogels formed a large number of ice crystals and turned white throughout, whereas PLXG and PLXSG hydrogels maintained good transparency without any observable internal ice crystal formation. This indicates that the addition of xanthan gum and silicon dioxide does not contribute to freeze resistance, while lithium ions and glycerol endow the hydrogel with excellent freeze resistance. Infrared thermal imaging of the hydrogels during bending (Fig. 3(c), (d) and Fig. S7, S8, ESI†) further confirmed that, even at −40 °C and 70 °C, the hydrogels did not freeze or dry out, remaining bendable, stretchable, and flexible.

Further investigation was conducted into the stress–strain curves of the PLXSG hydrogel at temperatures of 70 °C, 25 °C, 0 °C, and −40 °C to assess the stability of its mechanical performance under extreme temperature conditions (Fig. 3(e)–(g)). At 70 °C, the PLXSG hydrogel exhibited a maximum tensile strain of 625.35%, a Young's modulus of 12.56 kPa, and a toughness of 7.87 MJ m⁻³. In contrast, at −40 °C, the hydrogel displayed a maximum tensile strain of 918.88%, a Young's modulus of 1.27 kPa, and a toughness of 5.69 MJ m⁻³. Although the hydrogel demonstrated comparable elongation rates and toughness at both temperatures, the Young's modulus showed a notable disparity, likely due to greater moisture loss at elevated temperatures compared to room and low temperatures. Overall, the hydrogel maintained commendable mechanical properties across different temperatures, exhibiting excellent tensile strength and a sufficient range of tensile strain, making it suitable for use in flexible sensors even under extreme conditions. The enhanced ductility of hydrogels at low temperatures may arise from several factors. Low temperatures can cause polymer chains to rearrange or relax, reducing internal stress and enabling easier chain sliding, which enhances ductility. Lower temperatures might also decrease crosslinking density, improving chain mobility. The redistribution of water molecules may create microdomains or uniformly distributed water clusters, softening the hydrogel and increasing flexibility. Glycerol reduces cohesion between polymer chains, facilitating easier sliding at low temperatures, whereas at room temperature, stronger intermolecular forces in highly crosslinked PVA hydrogels may limit this effect. Lithium ions (Li⁺) form stable hydration shells that remain fluid at low temperatures, preserving the hydrogel's flexibility and aiding chain sliding and adjustment.

Fig. S9 and S10 (ESI†) depict the relationship between strain and the rate of resistance change in PLXSG hydrogel at −40 °C and 70 °C, respectively. At −40 °C, the hydrogel demonstrated a trend of increasing resistance change rate with increasing strain, with gauge factors (GF) for the increasing strain ranges being 0.913, 1.391, and 1.457. This indicates that the hydrogel maintains good elasticity and conductivity at low temperatures. At 70 °C, despite a decrease in conductivity, the hydrogel still exhibited sensitivities with GF values of 0.565, 0.945, and 1.283 across three strain ranges, suggesting that the hydrogel retains a certain level of elasticity and stability at high temperatures. These characteristics endow the hydrogel with potential applications in extreme environments with severe temperature fluctuations.

To assess the stability of hydrogels in high-temperature environments, mass retention tests were conducted at various temperatures. The moisture retention capacity of the PLXSG and PLXS hydrogels was specifically examined by subjecting them to 20 °C, 40 °C, and 70 °C for 48 hours (Fig. 3(h)–(j)), with mass changes monitored throughout the period. The results indicated that at all tested temperatures, the mass retention rate of the glycerol-containing PLXSG hydrogel was significantly higher than that of the glycerol-free PLXS hydrogel. Although both hydrogels experienced rapid moisture loss within the first six hours, the mass change of the PLXSG hydrogel was minimal during the subsequent 42 hours, retaining 62% of its mass at 70 °C, 79% at 40 °C, and 87% at the near-room temperature of 20 °C. In contrast, the PLXS hydrogel exhibited a more rapid loss of moisture under all three temperature conditions, with nearly complete moisture loss after 48 hours. At 40 °C, the PLXS hydrogel retained only 34% of its mass, and at 70 °C, this dropped to 22%. These results demonstrate that the addition of glycerol significantly enhances the water retention of hydrogels, enabling them to maintain moisture under extreme conditions, such as high temperatures. This enhancement highlights the improved environmental adaptability of glycerol-containing hydrogels.

The long-term stability of hydrogels in complex environments is a crucial consideration for their application as flexible sensing materials.^40–43 After 7 days at 70 °C, the hydrogels exhibited an increase in maximum stress from 4.22 MPa to 5.12 MPa, accompanied by a decrease in maximum strain from 625.67% to 457.24%, likely due to further water loss. Similarly, at −40 °C, maximum stress increased from 1.24 MPa to 2.59 MPa, while strain decreased from 917.53% to 734.5%, possibly due to glycerol leaching and additional water freezing (Fig. S11, ESI†). The conductivity of PLXSG hydrogel decreased from 0.392 S cm⁻¹ to 0.326 S cm⁻¹ after 7 days at 70 °C, and from 0.563 S cm⁻¹ to 0.543 S cm⁻¹ after 7 days at −40 °C, likely due to the reduction of lithium ions as water evaporated and the increased restriction on ion mobility (Fig. S12, ESI†). Fig. S13 (ESI†) a presents images of four hydrogels with different compositions after 30 days at 20 °C. The three hydrogels without glycerol dried out, losing their shape and flexibility, while the PLXSG hydrogel retained its flexibility, transparency, and intact structure. Fig. S13b (ESI†) shows PLXSG and PLXS hydrogels after 7 days at 70 °C. The PLXS hydrogel exhibited changes in color and shape due to excessive drying, whereas the PLXSG hydrogel maintained some transparency, indicating that moisture had not been completely lost.

3.4 Stress sensitivity properties of hydrogels

The introduction of metal ions, such as lithium ions, has modified and optimized the internal network of hydrogels, thereby enhancing the synergistic effects of multiple components within the hydrogel. This modification has endowed the hydrogel with high conductivity and excellent strain sensing capabilities. The study employed a testing platform comprising a universal tensile machine, an LCR digital bridge, and a computer to evaluate the electromechanical characteristics of the hydrogel. Fig. 4(a) illustrates the correlation between the relative change in resistance and strain for the PLXSG hydrogel at room temperature, spanning a strain range of 0% to 450%. The gauge factor (GF), defined as the slope of the line obtained through linear fitting of this relationship, reached 1.152 within the 0% to 170% strain range and peaked at 3.623 within the 170% to 450% strain range. When the hydrogel is stretched within a small strain range, lithium ions can move relatively easily through the entangled and dense network of PVA molecular chains. However, when stretched to a larger strain range (>180%), the PVA chains are separated under the applied force, causing many PVA chains to become dispersed. The increased distance between chains hinders the movement of lithium ions along the PVA, leading to an increase in relative resistance^44–46

The PLXSG hydrogel demonstrated stable electromechanical sensing behavior during cyclic strain stretching tests across both small (0.3–9%) and large (25–200%) strain ranges, with no signal drift observed (Fig. 4(b) and (c)). Notably, the hydrogel strain sensor could detect resistance changes at strains as low as 0.3%, highlighting the PLXSG hydrogel's exceptional detection limit. This capability broadens the strain detection range of flexible sensors, allowing for the detection of minute motion changes. Fig. 4(d) and (e) illustrate the temporal evolution of the resistance change rate as the hydrogel is subjected to tensile strain from 0% to 100% in 10% increments. As strain increases gradually, the resistance change rate correspondingly increases and remains consistent during a set dwell time. During the return phase, the resistance change rate remains nearly identical to that of the rising phase, further confirming that the PLXSG hydrogel flexible strain sensor can reliably identify electrical signal changes at varying strain levels, demonstrating excellent signal recognition reliability. Fig. 4(f) shows that the hydrogel sensor can distinctly differentiate between different stretch-recovery rates. As the stretch-recovery rate increases from 100 mm min⁻¹ to 500 mm min⁻¹, the density of the resistance change rate curve progressively increases, indicating the PLXSG hydrogel's robust frequency recognition capability. Additionally, the response and recovery times—key performance metrics for hydrogels—were evaluated. Fig. 4(g) illustrates that when subjected to ultra-high rate stretching to 70% strain, the PLXSG hydrogel exhibited a short response time of 214 ms and a recovery time of 286 ms, indicating its excellent response-recovery performance.

Further investigation of the hydrogel flexible sensor under a significant strain of 100% revealed its resistance change rate over 500 stretch-recovery cycles. As illustrated in Fig. 4(h), the electrical signal response of the hydrogel sensor exhibited a relatively constant value throughout the uninterrupted cyclic repetition. The gradual increase in the resistance change rate with repeated cycles can be attributed to the following factors: repeated cyclic strain in hydrogels can cause irreversible breaking of hydrogen bonds within the polymer network, leading to network relaxation or reconfiguration. Alongside this, overall moisture loss within the hydrogel and potential degradation at the electrode–hydrogel interface, such as increased contact resistance, can compromise the stability of conductive pathways, resulting in a gradual increase in the resistance change rate over time.^53–56 However, this trend is not significant in PLXSG hydrogel strain sensors and does not affect their performance. The inset demonstrates that during the 30 cycles from 3122 s to 3594 s, the resistance change rate exhibited similar peak values and waveforms. As shown in Fig. 4(i) and Tables S4, S5 (ESI†), the novel PLXSG hydrogel demonstrates superior performance in mechanical properties, conductivity, available temperature range, and sensitivity compared to recently reported ion-conductive hydrogels and other conductive materials used for strain sensing. Notably, it excels in mechanical strength and conductivity compared to existing hydrogels. These outstanding combined properties suggest that PLXSG hydrogels have significant potential for applications requiring specific functional demands, particularly in the field of health monitoring.

The high conductivity of the PLXSG hydrogel (0.767 S cm⁻¹) enables it to be connected in series with a power source at temperatures as extreme as −40 °C and 70 °C, successfully illuminating an LED light, as shown in Fig. S14 and S15 (ESI†). In Fig. S16 (ESI†), the influence of LiCl concentration on the hydrogel's conductivity and tensile strength was examined. With increasing LiCl concentration, the hydrogel's conductivity increased, while its mechanical strength decreased. A LiCl concentration of 2 M allowed the hydrogel to maintain high conductivity while exhibiting adequate mechanical strength. The impact of temperature on the conductivity of the PLXSG hydrogel was also investigated. As illustrated in Fig. S17 (ESI†), at low temperatures, the hydrogel's conductivity decreased from 0.767 S cm⁻¹ to 0.563 S cm⁻¹, which is likely due to slower ion migration at reduced temperatures. At elevated temperatures, conductivity declined from 0.767 S cm⁻¹ at room temperature to 0.392 S cm⁻¹—a reduction of approximately 50%. This significant decrease is primarily attributed to water evaporation at high temperatures, which results in the loss of a substantial number of free ions, leading to a marked reduction in conductivity.

3.5 Human movement posture detection

To demonstrate its practical application, we integrated the conductive ionic hydrogel into a flexible strain sensor, which was then affixed to various parts of the human body to monitor resistance fluctuations during physical activity. As shown in Fig. 5(a), when applied to the face, the hydrogel-based strain sensor accurately captures resistance changes corresponding to the strain induced by a smile. Fig. 5(b) illustrates the sensor's ability to detect different degrees of finger bending (0°, 30°, 60°, 90°), producing a consistent electrical signal at each fixed finger posture, with a gradual increase in signal intensity proportional to the bending angle. Additionally, the sensor is highly sensitive to subtle bodily movements, such as speech articulation. When placed on a volunteer's throat, the sensor distinctly registers the minor strain caused by repeatedly uttering the word “hi,” as depicted in Fig. 5(c). Fig. 5(d)–(g) further demonstrate that the PLXSG hydrogel-based flexible strain sensor precisely tracks more vigorous and extensive bending motions, including those of the legs, nape, wrists, and elbows. These results underscore the hydrogel sensor's excellent durability and high sensitivity to both large and small movements, highlighting its practical value in human health monitoring.⁵⁷

3.6 Deep learning enhanced handwritten letter recognition

The application of deep learning technology has significantly improved the efficiency of human–computer interaction systems, particularly in data analysis and pattern recognition. Convolutional neural networks (CNNs), a prominent and effective deep learning algorithm, have shown substantial potential in this domain. By reducing model parameters and computational load, CNNs not only enhance the reliability of recognition results but also lessen the dependency on large training datasets. Writing, as a fundamental activity for recording and transmitting information, plays a crucial role in human–computer interaction systems. For the experimental setup, a flexible writing pad was designed with a layer of transparent tape covering the surface of the PLXSG hydrogel, ensuring the stability and durability of the hydrogel sensor. Fig. 6(a) presents the assembly diagram of the handwriting recognition device and the data processing flow. By analyzing the responses of the hydrogel sensor to different letters, our system can accurately distinguish each character, even under rapid writing or varying pressure conditions. The response signals of six English letters, A through F, are recorded and displayed in Fig. 6(b). Through signal comparison, our machine learning model accurately classifies and recognizes each character. As depicted in Fig. 6(c), the reduction in the loss function and the improvement in learning accuracy with increasing training epochs confirm the effectiveness of the CNN algorithm we implemented. After several cycles of optimization training, the model achieved exceptional classification performance, ultimately reaching an accuracy rate of 98.3%. The confusion matrix indicates that the model demonstrates a high degree of recognition accuracy, with nearly all letters correctly classified, even between the most easily confused pairs, where the model maintains a 99% prediction accuracy (Fig. 6(d)). These results underscore the remarkable capability of the machine learning-enhanced PLXSG hydrogel in letter recognition tasks and highlight its broad application prospects in human–computer interaction and intelligent recognition.

3.7 Sensor arrays and Morse code

To explore the potential applications of PLXSG hydrogel in multifunctional sensing and visualization, a 5 × 5 2D sensor array was designed and fabricated using PVC as a substrate. Each individual sensor unit was connected by copper foil, creating a unified pressure-sensing array (Fig. S18, ESI†). When a subject pressed the four central units (C2, C3, D2, D3) of the sensor array simultaneously, the monitoring device accurately recorded the resistance change signals in the four sensor units without significant signal interference (Fig. 6(e)). The intensity of the resistance change was represented by three different colors: darker blue pixel blocks for signals with a resistance change rate of 0% to 10%, light blue pixel blocks for rates of 10% to 20%, and white pixel blocks for rates greater than 20%. With the hydrogel sensor array operating stably and clearly recording signal changes, the array successfully visualized the assembly of different letters by building blocks placed on it, accurately recording the letters “KUST” as electronic identifiers (Fig. 6(f)). For instance, the letter “U” was formed by the pressure exerted by eight blocks, causing specific locations within the array (A1, A5, B1, B5, C1, C5, D1, D5, E2, E4) to exhibit increased resistance changes relative to other units. Additionally, the PLXSG hydrogel was employed for encryption and information transmission via Morse code, as illustrated in Fig. 6(g). This figure displays the combinations of dots and dashes representing all 26 English letters. By varying the tapping durations and pauses on the hydrogel surface, the encrypted message “HELLO” was spelled out. As shown in Fig. 6(h), distinct tapping durations resulted in varying peak widths of resistance change, with each combination of resistance change signals accurately corresponding to the respective letters. These experimental results demonstrate the potential of PLXSG hydrogel-based array sensors for applications in electronic skin, human–machine interaction, and secure communication.

4. Conclusion

This study successfully developed an innovative PLXSG conductive ionic hydrogel by integrating xanthan gum and silica nanoparticles as enhancers, combined with a binary solvent system of water and glycerol. The hydrogel demonstrated exceptional mechanical and electrical properties across a wide temperature range, from −40 °C to 70 °C. It exhibited high tensile fracture strength (1.8 MPa), high elongation at break (790%), and significant conductivity (76.752 S m⁻¹). The hydrogel also displayed high strain sensitivity, with a gauge factor (GF) reaching up to 3.623 at strains exceeding 175%, a low strain sensing limit (0.3%), and excellent elasticity. These features suggest that the PLXSG hydrogel holds significant potential for applications in human health monitoring and flexible wearable electronic devices.

The hydrogel's rapid response to electrical signals and cyclic stability enables it to monitor a wide range of human movements, from large-scale motions such as finger, wrist, arm, and knee movements to subtle changes in skin, such as those in the cheeks and throat. Furthermore, the study explored the potential applications of the PLXSG hydrogel in multifunctional sensors, including array sensors, Morse code information transmission, and machine learning-enhanced handwriting recognition. The PLXSG hydrogel was utilized to create a 5 × 5 two-dimensional sensor array, capable of precisely capturing pressure changes, demonstrating its potential in human–machine interaction and electronic skin applications. The hydrogel sensor can encrypt and transmit information in Morse code, offering a novel solution for information security. In conjunction with deep learning technology, the PLXSG hydrogel exhibited outstanding performance in handwriting recognition tasks, indicating its considerable potential in intelligent recognition and human-machine interaction.

In conclusion, the research on the PLXSG hydrogel not only paves a new path for the development of high-performance flexible sensing materials but also provides an excellent option for the design and application of future flexible electronics and smart wearable devices.

Author contributions

All authors have contributed to the writing and revision of the manuscript and given approval to the final version of the manuscript.

Data availability

The data supporting this study's findings are available from the corresponding author upon reasonable request.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (no. 52165066), Yunnan Fundamental Research Projects (grant no. 202401AT070354) and Xingdian Talent Support Program of Yunnan Province (no. 2022).

References

1. M. Sher, L. A. Shah, L. Ara, R. Ullah, M. Khan, H.-M. Yoo and J. Fu, Sens. Actuators, A, 2024, 370, 115199 CrossRef CAS .
2. M. Lin, Z. Zheng, L. Yang, M. Luo, L. Fu, B. Lin and C. Xu, Adv. Mater., 2022, 34, 2107309 CrossRef CAS PubMed .
3. A. Chen, J. Zhang, J. Zhu, Z. Yan, Q. Wu, S. Han, J. Huang and L. Guan, J. Mater. Chem. A, 2023, 11, 4977–4986 RSC .
4. X. Fan, J. Geng, Y. Wang and H. Gu, Polymer, 2022, 246, 124769 CrossRef CAS .
5. Y. Zhou, X. Fei, J. Tian, L. Xu and Y. Li, J. Colloid Interface Sci., 2022, 606, 192–203 CrossRef CAS PubMed .
6. X. Zou, X. Wang, Z. Bai, O. Yue, C. Wei, L. Xie, H. Zhang and X. Liu, Chem. Eng. J., 2023, 463, 142349 CrossRef CAS .
7. H. Wu, Q. Zhao, Y. Liang, L. Ren and L. Ren, ACS Sustainable Chem. Eng., 2022, 10, 4425–4437 CrossRef CAS .
8. X. Yao, S. Zhang, L. Qian, N. Wei, V. Nica, S. Coseri and F. Han, Adv. Funct. Mater., 2022, 32, 2204565 CrossRef CAS .
9. W. Zhao, X. Qu, Q. Xu, Y. Lu, W. Yuan, W. Wang, Q. Wang, W. Huang and X. Dong, Adv. Electron. Mater., 2020, 6, 2000267 CrossRef CAS .
10. J. Cong, Z. Fan, S. Pan, J. Tian, W. Lian, S. Li, S. Wang, D. Zheng, C. Miao, W. Ding, T. Sun and T. Luo, ACS Appl. Mater. Interfaces, 2021, 13, 34942–34953 CrossRef CAS .
11. Y. Lu, Y. Yue, Q. Ding, C. Mei, X. Xu, S. Jiang, S. He, Q. Wu, H. Xiao and J. Han, InfoMat, 2023, 5, e12409 CrossRef CAS .
12. J. Zhang, Y. Liang, Z. Deng, H. Xu, H. Zhang, B. Guo and J. Zhang, ACS Appl. Mater. Interfaces, 2023, 15, 29902–29913 CrossRef CAS PubMed .
13. K. Shen, K. Xu, M. Zhang, J. Yu, Y. Yang, X. Zhao, Q. Zhang, Y. Wu, Y. Zhang and Y. Cheng, Chem. Eng. J., 2023, 451, 138525 CrossRef CAS .
14. X. Zhao, H. Jiang, P. Sun, R. Wei, S. Jiang, J. Hu and S. Zhang, ACS Appl. Mater. Interfaces, 2023, 15, 56275–56284 CrossRef CAS PubMed .
15. S. Wu, B. Wang, D. Chen, X. Liu, H. Wang, Z. Song, D. Yu, G. Li, S. Ge and W. Liu, Sci. China Mater., 2023, 66, 1923–1933 CrossRef CAS .
16. Z. Nie, K. Peng, L. Lin, J. Yang, Z. Cheng, Q. Gan, Y. Chen and C. Feng, Chem. Eng. J., 2023, 454, 139843 CrossRef CAS .
17. Z. Liu, J. Zhang, J. Liu, Y. Long, L. Fang, Q. Wang and T. Liu, J. Mater. Chem. A, 2020, 8, 6219–6228 RSC .
18. H. Rajati, H. Alvandi, S. S. Rahmatabadi, L. Hosseinzadeh and E. Arkan, Int. J. Biol. Macromol., 2023, 226, 1426–1443 CrossRef CAS PubMed .
19. T. Li, H. Wei, Y. Zhang, T. Wan, D. Cui, S. Zhao, T. Zhang, Y. Ji, H. Algadi, Z. Guo, L. Chu and B. Cheng, Carbohydr. Polym., 2023, 309, 120678 CrossRef CAS PubMed .
20. Y. Song, H. Tan, S. Qin, Z. Liu, C. Liu, C. Shen, P. Yang and S. Li, Nano Res., 2024, 17, 3398–3406 CrossRef CAS .
21. P. Sautrot-Ba, N. Razza, L. Breloy, S. A. Andaloussi, A. Chiappone, M. Sangermano, C. Hélary, S. Belbekhouche, T. Coradin and D.-L. Versace, J. Mater. Chem. B, 2019, 7, 6526–6538 RSC .
22. K. Wang, Y. Zhang, T. Chen, L. Bai, H. Li, H. Tan, C. Liu and X. Qu, Composites, Part B, 2023, 266, 110991 CrossRef CAS .
23. Y. Huang, M. Zhang and W. Ruan, J. Mater. Chem. A, 2014, 2, 10508–10515 RSC .
24. L. Hu, Y. Wang, Q. Liu, M. Liu, F. Yang, C. Wang, P. Pan, L. Wang, L. Chen and J. Chen, Chin. Chem. Lett., 2023, 34, 108262 CrossRef CAS .
25. M. Xu, J. Zhu, J. Xie, Y. Mao and W. Hu, Small, 2024, 20, 2305448 CrossRef CAS PubMed .
26. X. Miao, Z. Li, K. Hou, Q. Gao, Y. Huang, J. Wang and S. Yang, Chem. Eng. J., 2023, 476, 146848 CrossRef CAS .
27. Z. Li, F. Yin, W. He, T. Hang, Z. Li, J. Zheng, X. Li, S. Jiang and Y. Chen, Int. J. Biol. Macromol., 2023, 230, 123117 CrossRef CAS PubMed .
28. C. You, W. Wu, W. Yuan, P. Han, Q. Zhang, X. Chen, X. Yuan, L. Liu, J. Ye, L. Fu and Y. Wu, Adv. Funct. Mater., 2023, 33, 2208206 CrossRef CAS .
29. Y. Zhou, L. Zhang, X. Lin, J. Lu, Z. Huang, P. Sun, Y. Zhang, X. Xu, Q. Li and H. Liu, Int. J. Biol. Macromol., 2023, 233, 123573 CrossRef CAS PubMed .
30. X. Liu, Z. Wu, D. Jiang, N. Guo, Y. Wang, T. Ding and L. Weng, Adv. Compos. Hybrid Mater., 2022, 5, 1712–1729 CrossRef CAS .
31. H. Fu, F. Wang, Z. Cao, L. Liu, G. Zhu, J. Yao, J. Militky and J. Wiener, React. Funct. Polym., 2023, 186, 105572 CrossRef CAS .
32. S.-N. Li, Z.-R. Yu, B.-F. Guo, K.-Y. Guo, Y. Li, L.-X. Gong, L. Zhao, J. Bae and L.-C. Tang, Nano Energy, 2021, 90, 106502 CrossRef CAS .
33. L. Shu, X.-F. Zhang, Y. Wu, Z. Wang and J. Yao, Int. J. Biol. Macromol., 2023, 240, 124438 CrossRef CAS PubMed .
34. L. Han, K. Liu, M. Wang, K. Wang, L. Fang, H. Chen, J. Zhou and X. Lu, Adv. Funct. Mater., 2018, 28, 1704195 CrossRef .
35. S. A. Bernal-Chávez, S. Alcalá-Alcalá, Y. S. Tapia-Guerrero, J. J. Magaña, M. L. Del Prado-Audelo and G. Leyva-Gómez, RSC Adv., 2022, 12, 21713–21724 RSC .
36. C. Wang, S. S. Rubakhin, M. J. Enright, J. V. Sweedler and R. G. Nuzzo, Adv. Funct. Mater., 2021, 31, 2010246 CrossRef CAS PubMed .
37. L. Yan, T. Zhou, L. Han, M. Zhu, Z. Cheng, D. Li, F. Ren, K. Wang and X. Lu, Adv. Funct. Mater., 2021, 31, 2010465 CrossRef CAS .
38. M. Guo, Y. Wu, S. Xue, Y. Xia, X. Yang, Y. Dzenis, Z. Li, W. Lei, A. T. Smith and L. Sun, J. Mater. Chem. A, 2019, 7, 25969–25977 RSC .
39. Y. Zhou, X. Wang, X. Lin, Z. Wang, Z. Huang, L. Guo, H. Xie, X. Xu and F. Dong, Int. J. Biol. Macromol., 2024, 263, 130511 CrossRef CAS PubMed .
40. T. Habib, X. Zhao, S. A. Shah, Y. Chen, W. Sun, H. An, J. L. Lutkenhaus, M. Radovic and M. J. Green, npj 2D Mater. Appl., 2019, 3, 8 CrossRef .
41. M. Mao, K.-X. Yu, C.-F. Cao, L.-X. Gong, G.-D. Zhang, L. Zhao, P. Song, J.-F. Gao and L.-C. Tang, Chem. Eng. J., 2022, 427, 131615 CrossRef CAS .
42. B.-F. Guo, Y.-J. Wang, C.-F. Cao, Z.-H. Qu, J. Song, S.-N. Li, J.-F. Gao, P. Song, G.-D. Zhang, Y.-Q. Shi and L.-C. Tang, Adv. Sci., 2024, 11, 2309392 CrossRef CAS PubMed .
43. Y.-J. Wang, B.-F. Guo, L.-D. Peng, Y. Li, C.-F. Cao, G.-D. Zhang, J.-F. Gao, P. Song, Y.-Q. Shi, K. Cao and L.-C. Tang, Adv. Nanocomp.s, 2024, 1, 217–239 CrossRef .
44. Z. Wang, J. Chen, Y. Cong, H. Zhang, T. Xu, L. Nie and J. Fu, Chem. Mater., 2018, 30, 8062–8069 CrossRef CAS .
45. C. Ma, Y. Wang, Z. Jiang, Z. Cao, H. Yu, G. Huang, Q. Wu, F. Ling, Z. Zhuang, H. Wang, J. Zheng and J. Wu, Chem. Eng. J., 2020, 399, 125697 CrossRef CAS .
46. X. Lu, Y. Zeng, Y. Yang, X. Yang, E. Wei, C. Cui, J. Xie, Y. Qin and Z. Qian, Adv. Mater. Technol., 2023, 8, 2202123 CrossRef CAS .
47. J. Ren, Y. Liu, Z. Wang, S. Chen, Y. Ma, H. Wei and S. Lü, Adv. Funct. Mater., 2022, 32, 2107404 CrossRef CAS .
48. J. Liu, X. Chen, B. Sun, H. Guo, Y. Guo, S. Zhang, R. Tao, Q. Yang and J. Tang, J. Mater. Chem. A, 2022, 10, 25564–25574 RSC .
49. Y. Ma, D. Zhang, Z. Wang, H. Zhang, H. Xia, R. Mao, H. Cai and H. Luan, ACS Appl. Mater. Interfaces, 2023, 15, 29413–29424 CrossRef CAS PubMed .
50. L. Zhao, H. Zhang, Z. Guo, X. Yu, X. Jiao, M.-H. Li and J. Hu, ACS Appl. Mater. Interfaces, 2022, 14, 51394–51403 CrossRef CAS PubMed .
51. H. Jiang, C. Ou, D. Zhang, X. Hu, Y. Ma, M. Wang, Y. Huang and L. Xiao, ACS Appl. Polym. Mater., 2023, 5, 6828–6841 CrossRef CAS .
52. R. Wang, W. Chi, F. Wan, J. Wei, H. Ping, Z. Zou, J. Xie, W. Wang and Z. Fu, ACS Appl. Mater. Interfaces, 2022, 14, 21278–21286 CrossRef CAS PubMed .
53. J. P. Gong, Soft Matter, 2010, 6, 2583–2590 RSC .
54. S. Ko, A. Chhetry, D. Kim, H. Yoon and J. Y. Park, ACS Appl. Mater. Interfaces, 2022, 14, 31363–31372 CrossRef CAS PubMed .
55. Y. Fang, Z. Bai, L. Yang, J. Wei, Y. Wang, S. Wang and J. Cui, Adv. Mater. Technol., 2023, 8, 2301012 CrossRef CAS .
56. C. Cao, T. Huang and Y. Li, Macromol. Rapid Commun., 2024, 45, 2300467 CrossRef CAS PubMed .
57. L. Zhao, T. Ke, Q. Ling, J. Liu, Z. Li and H. Gu, ACS Appl. Polym. Mater., 2021, 3, 5494–5508 CrossRef CAS .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4tc02804j

This journal is © The Royal Society of Chemistry 2024