A multifunctional hydrogel-based strain sensor and triboelectric nanogenerator for running monitoring and energy harvesting

With the development of information energy technology and advanced sensing technology, flexible wearable electronic devices have attracted wide attention due to their various applications, such as electronic skin,¹ flexible transistors,² triboelectric nanogenerators (TENGs),^3–5 and flexible super-capacitors.⁶ Furthermore, the development of wearable electronic products needs flexible materials with high durability, including flexible conductive materials.^7,8 Recently, flexible electrode technology has made significant achievements, such as stretchable and self-healing devices.^9,10 The core of these flexible devices is to prepare conductive components (graphene,¹¹ nano-metal wires,¹² MXene,¹³ carbon nanotubes,¹⁴ etc.) uniformly dispersed in a flexible matrix [polydimethylsiloxane (PDMS),¹⁵ polyethylene terephthalate (PET),¹⁶ hydrogel,¹⁷ silica gel,¹⁸ etc.]. Advanced flexible material design and preparation technologies have brought new ideas to the research and development of multifunctional flexible electronic generation, such as posture sensing and energy harvesting technologies.^19–21 Integrating flexible sensor components with flexible energy devices will facilitate the implementation and application of two different working principles of devices. In addition, with the improvement of household consumption levels and the emphasis on physical and mental health, more and more people have begun to pay attention to personal fitness to enhance their physical fitness, reduce the frequency of diseases, and have a better quality of life.²² Running is an aerobic fitness exercise with simple technical requirements and easy implementation and is the most common individual fitness exercise for ordinary residents.²³ Although regular running activities have health benefits for both body and mind, incorrect posture during running can bring additional risks to athletes, such as bone and joint injuries.²⁴ Thus, the use of flexible multifunctional devices for health monitoring in various sports activities, especially in human joint sensing and skin micro-change perception, has application value.

Conductive hydrogel is an advanced soft material with a water-rich and porous structure, which relays on cross-linking of polymer chains dispersed in a water medium.^24–26 Due to good biological affinity, natural smooth tissue similarity, soft texture, and other characteristics, hydrogels have great application value in tissue engineering,²⁷ flexible actuators,²⁸ electronic skin,²⁹ energy storage equipment,³⁰ and other fields.^31,33 However, traditional hydrogels have a single function and lack toughness, viscosity, conductivity, and other properties, which seriously affect the application of hydrogels in some specific fields. Thus, researchers have developed functional hydrogels by designing various strategies, expanding the application range of hydrogels. The performance of hydrogel depends on the network composition and crosslinking mode; hence, the physical and chemical properties of hydrogel can be adjusted as needed to meet the practical application needs. Furthermore, the resistance of the new hydrogel will change when it is deformed (stretched and compressed), which makes it have applications in strain sensors, especially high-strength hydrogels with multiple networks.³² Furthermore, the novel hydrogel can form a TENG device with a dielectric film [polytetrafluoroethylene (PTFE),³³ Kapton,³⁴ polyvinyl chloride (PVC),³⁵ etc.], which plays a role in mechanical energy harvesting. Recently, green and environmentally friendly materials, multifunctional materials, and natural inorganic materials have attracted the interest of researchers.^36,37 Due to its sustainability and large-scale production capabilities, commercial materials based on natural plant fibers are receiving increasing attention. The plant materials or modified plant materials have acted as the preparation of things. However, more green new and low-carbon materials with good triboelectric properties and efficient structural design are an essential research direction to improve electrical output, environmental adaptability, and promote industrial applications of TENGs.

Hence, we design a mixed type conductive hydrogel based on polyvinyl alcohol (PVA), cotton paper (CP), graphite oxide (GO), and MXene, named PCGM hydrogel. The PVA and CP constitute the mixed type architecture of the PCGM hydrogel, which gives the PCGM hydrogel a high strength advantage. The GO and MXene can enhance the conductivity of the PCGM hydrogel, and MXene can also promote dual network cross-linking of PVA and CP in the PCGM hydrogel. Furthermore, the PCGM hydrogel can act as the PCGM-based strain sensor and triboelectric nanogenerator (P-TENG) for running posture monitoring and mechanical energy harvesting. According to the results, when the strain is less than 50%, the sensitivity of the PCGM-based strain sensor [also known as gauge factor (GF)] is 0.16. When the strain is between 50% and 400%, the GF value can reach 1.92. When the strain is between 400% and 600%, the GF value can arrive at 0.72. Furthermore, when the pressure applied to the sensor is between 0 and 16 kPa, the pressure sensitivity of the PCGM-based strain sensor is 0.0164 kPa⁻¹. When the pressure range is 16–120 kPa, the pressure sensitivity of the PCGM-based strain sensor is 0.002 86 kPa⁻¹. Flexible strain sensors can not only monitor the movement of human joint parts (such as the knee, elbow, neck, and elbow) but also achieve the perception of skin micro-vibrations caused by facial expression changes (such as smiling, thinking, and chewing), which will contribute to the comprehensive health monitoring of runners. In addition to achieving high-sensitivity sensing, the PCGM hydrogel can also act as an efficient energy harvester combined with a PTFE film. According to the results, the P-TENG maximum output power can arrive at 135 µW (matched resistance: 30 MΩ), which provides the possibility of harvesting human exercise energy. The short-circuit current (I_sc), open-circuit voltage (V_oc), and transfer charge (Q_sc) of P-TENG can arrive at 10.36 µA, 229.85 V, and 49.24 nC, respectively.

The PVA K30 solution was bought from Angxing New Carbon Materials Co., Ltd. (Changzhou, China). The GO powders were obtained from Ruixi Biotechnology Co., Ltd. (Xi’an, China). The MXene powders were purchased from Fosman Technology Co., Ltd. (Beijing, China). The cotton paper was obtained from Qianyi Paper Shopping Mall (Yiwu, China). The LiBr was bought from Pengcai Fine Chemical Co., Ltd. (Langfang, China). In addition, some chemical auxiliary reagents, such as deionized water, were obtained from Local Chemical Center. The polytetrafluoroethylene (PTFE) film was bought from Jingdong Mall.

As we all know, cotton is a fibrous aggregation of various protons and other substances with similar cell wall structures. The cotton paper (CP) made has the advantages of strong water absorption, strong elasticity, softness, and portability, as depicted in Fig. 1(a). The molecular structure of cotton is composed of a carbon chain, hydrogen bond, oxygen bond, and branch chain, which have excellent absorption performance and can accelerate the absorption of moisture in the environment. The preparation of the CP solution follows the method alkali degumming, and this is a universal strategy for preparing fiber solutions.³⁸ Concretely, add 20 g of crushed cotton paper to 1 l of NaHCO₃ (10 g) solution, heat and boil for 2 h, and rinse repeatedly with deionized water five times. Then, dry the obtained cotton paper sample and control the experimental temperature at 55 °C. Subsequently, the dried cotton fibers were dissolved in 60 ml (93.4M) of LiBr solution and repeatedly dialyzed with deionized water to dilute the CP solution to 60 mg ml⁻¹ for subsequent film preparation.

Figures 1(b) and 1(c) depict the manufacturing process of the PCGM hydrogel. First, add 30 ml of the CP solution, 3 g of GO powders, and 2 g of MXene powders to the PVA solution and then mechanically stir to form a uniform and stable suspension. Then, the prepared suspension was poured into molds of various shapes and heated at 75 °C for 30 min to obtain the PCGM hydrogel, so as to prepare hydrogel structures with different types. Furthermore, the prepared PCGM hydrogel can be tailored to meet different environmental needs, as demonstrated in Fig. 1(d). Meanwhile, the PCGM hydrogel exhibits excellent elasticity, which can bring better durability to the PCGM hydrogel device (Fig. S1, supplementary material).

The PCGM-based strain sensor has two wires arranged at both ends for the signal output of the strain sensor, as illustrated in Fig. 1(e). In addition, this simple flexible sensor design method can reduce production costs and improve preparation efficiency. According to previous work,³⁵ the PTFE film has a strong electron obtaining ability in the triboelectric series. Additionally, the PTFE film also has good chemical stability and mechanical strength. Thus, we introduce the PTFE film and combine it with the PCGM hydrogel to form a P-TENG device for mechanical energy harvesting, as presented in Fig. 1(f). The PCGM hydrogel acts as both a conductive electrode and a positive triboelectric layer. This method of integrating two devices based on different working mechanisms can fully leverage the advantages of different devices and achieve highly integrated multi-functional flexible wearable electronics.

As depicted in Fig. 1(g), the scanning electron microscope (SEM) image of the PCGM hydrogel reveals the micro-/nano-characteristics of a rough flaky-like surface structure, which is beneficial for the generation of triboelectric charges. Moreover, according to the element analysis results on the surface of the PCGM hydrogel [Figs. 1(h)–1(j)], the uniform distribution of elements can be observed, and the composition of the PCGM hydrogel is stable. The mechanical control system provides the required external forces and motion. The ΔR/R₀ of the PCGM-based strain sensor device was measured by using a digital multimeter. The electrometer (Keithley 6514) can measure the voltage on the capacitor, and the digital oscilloscope (SDS2074X Plus) can measure the output voltage of C-TENG under different resistance.

Furthermore, the results in Fig. 1(k) show that the introduction of CP into PVA can significantly improve the mechanical properties of the PCGM hydrogel. Since the hydrogel has the role of both triboelectric layer and conductive electrode, the conductivity of the PCGM hydrogel needs to be enhanced by adding GO. Meanwhile, MXene is introduced into the PCGM hydrogel to improve the dielectric properties of the material, which can help the triboelectric charge transfer of the PCGM hydrogel. However, the addition of GO and MXene reduces the mechanical properties of the PCGM hydrogel, but it also improves the tensile ratio of the PCGM hydrogel, which is attributed to the fact that there is more water in the PCGM hydrogel. Compared with the PVA hydrogel, the Raman spectrum of the PCGM hydrogel shows an obvious characteristic peak at 1667 cm⁻¹, indicating that the content of lamellar microcrystals in cotton fiber is high, as demonstrated in Figs. 1(l) and S2 of the supplementary material. Furthermore, we also provide the chemical structure of PVA, CP, GO, and MXene in Fig. S3 of the supplementary material.

From the characterization results, the structural changes of the PCGM hydrogel during the preparation process can be illustrated in Fig. 2(a). The CP and PVA can provide the crosslinked network, which endows the PCGM hydrogel with excellent mechanical properties [Figs. 2(a1)–2(a3)]. Furthermore, the introduction of MXene and GO make the PCGM hydrogel have good conductivity. Meanwhile, a large number of hydrophilic groups on the surface of MXene can be used as a bridge to connect the CP and PVA network, which can improve the mechanical properties of the PCGM hydrogel, as present in [Figs. 2(a2)–2(a4)]. Thus, the doping of MXene promotes the crosslinking of the PVA and CP network, and meanwhile, it also can strengthen the tensile property of the PCGM hydrogel. Figures 2(b1)–2(b3) illustrate the working mechanism of the PCGM-based strain sensor in the compression and tension states. Figure 2(b1) shows the starting state of the PCGM-based strain sensor. When the PCGM-based strain sensor is compressed by force [Fig. 2(b)], the conductive channels (GO and MXene) in the PCGM hydrogel are denser, which can lead to the reduction of the resistance of the PCGM hydrogel. The decrease in resistance will increase with an increase in pressure; in other words, the greater the pressure, the smaller the resistance of the PCGM hydrogel. When the external force is withdrawn, the resistance of the PCGM hydrogel can return to its initial state, thereby achieving pressure sensing. In addition, when the PCGM-based strain sensor is stretched by an external force [Fig. 2(b3)], the conductive channel in the PCGM hydrogel becomes loose, which causes an increase in the resistance of the PCGM hydrogel. As the stretching amplitude increases, the resistance of the PCGM hydrogel also increases, resulting in a high amplitude of resistance variation. This stretching characteristic of the PCGM hydrogel can be used to sense small changes in the skin. After unloading the external force, the PCGM hydrogel can quickly return to its initial state due to its elasticity. In addition, this highly sensitive sensing feature provides a basis for human posture monitoring and facial microexpression sensing.

To evaluate the stability of the PCGM-based strain sensor under different deformation conditions, we tested the relative resistance change rate (ΔR/R₀) of the PCGM-based strain sensor under different stretching amplitudes. Figures 2(c) and 2(d) demonstrate the ΔR/R₀ of the PCGM-based strain sensor under different tensile strains. The experimental results show that when the strain of a PCGM-based strain sensor increases, ΔR/R₀ will generate corresponding enhancement and can produce response effects in a large range of strain amplitude from 2% to 600%, which indicates that the PCGM-based strain sensor can be used for monitoring small strains and large strains. In addition, according to the sensing signal, during the loading and unloading process of the force acting on the PCGM-based strain sensor, the sensing signal exhibits symmetrical characteristics, indicating that the PCGM-based strain sensor has wonderful resilience. Further long-term reliability experiments have shown that the PCGM-based strain sensor has stable cycling characteristics, as depicted in Fig. 2(e). According to the experimental results, at a stretching frequency of 1 Hz and a stretching strain of 50%, under 2000 consecutive experimental tests, the single cycle curve of ΔR/R₀ exhibits similar signal characteristics, indicating that the PCGM-based strain sensor has good stability. In addition, we investigated the effect of strain frequency on the output performance of the PCGM-based strain sensors under the same strain of 50%. According to the experimental results in Fig. 2(f), under different strain frequencies, the strain of the PCGM-based strain sensor is basically the same, indicating that the PCGM-based strain sensor has stable response characteristics. The response time and relaxation time of the PCGM-based strain sensor are about 168 and 114 ms, as demonstrated in Fig. 2(g). When the external force is 16 kPa, the ΔR/R₀ curve of the PCGM-based strain sensor remains consistent at different loading/unloading frequencies, indicating once again that the PCGM-based strain sensor has excellent sensing performance, as present in Fig. 2(h). From the ΔR/R₀-strain curve of the PCGM-based strain sensor in Fig. 2(i), when the strain is less than 50%, the sensitivity of the PCGM-based strain sensor (also known as the gauge factor, GF) is 0.16. When the strain is between 50% and 400%, the GF value can reach 1.92. When the strain is between 400% and 600%, the GF value can arrive at 0.72. Furthermore, according to the results in Fig. 2(j), when the pressure applied to the sensor is between 0 and 16 kPa, the pressure sensitivity of the PCGM-based strain sensor is 0.0164 kPa⁻¹. However, when the pressure range is 16–120 kPa, the pressure sensitivity of the PCGM-based strain sensor is 0.002 86 kPa⁻¹. Thus, the PCGM-based strain sensor has a wide pressure detection range. Figure 2(k) illustrates the ΔR/R₀ curve of a PCGM-based strain sensor under different pressures from 0.5 to 110 kPa, and the ΔR/R₀ amplitude of the PCGM-based strain sensor increases when the pressure improves.

Regular running has a positive effect on physical and mental health, which can promote health and prevent obesity, cardio cerebral vascular disease, and chronic diseases. In addition, it can improve suboptimal health. Although regular running activities have health benefits for both body and mind, it is based on the high risks (accidents and injuries) during the running process. Running injuries are progressive injuries caused by repeated minor injuries. It is generally attributed to the individual biomechanical posture characteristics, anatomical factors, and training errors of runners. Notwithstanding, running is a non-competitive sport, the injury rate remains high, and the characteristics and causes of injury are also diverse. Common areas of injury include the knee, ankle, and foot. Because most running enthusiasts are not professional runners, there is a lack of knowledge related to running, and most of their physical functions cannot meet the requirements of heavy workload exercise. Hence, abnormal running postures often occur, and abnormal running postures will lead to compensation of movement postures, thus generating or exacerbating the risk of sports injury. Flexible strain sensors can monitor the movement of different parts of the human body, such as the knee, elbow, neck, and elbow, to achieve real-time feedback on human movement posture, which can contribute to the healthy development of running, as demonstrated in Figs. 3(a1) and 3(a2). The PCGM-based strain sensor can provide real-time and accurate monitoring of the movement status of human joints, including the knee, wrist, finger, and elbow, which provides clear feedback on the degree of movement of bones and joints to help in the health assessment of running, as illustrated in Figs. 3(b)–3(d) and 3(g). Simultaneously, the PCGM-based strain sensor installed at the throat can detect information, such as swallowing and coughing, exhibiting a high sensitivity response, as presented in Figs. 3(e) and 3(f). Furthermore, the PCGM-based strain sensor installed at the chest can detect the frequency and intensity of the heartbeat [Figs. 3(h1) and 3(h2)]. For example, there is a significant difference in the frequency and amplitude of the PCGM-based strain sensor's output signal before and after exercise, achieving heart rate monitoring [Fig. 3(i)].

Due to the ability of PCGM-based strain sensors to monitor relatively small strains, it can be used for facial expression recognition. Hence, we selected three nodes on the face as the installation locations for the PCGM-based strain sensor, as illustrated in Figs. 4(a1) and 4(a2). Specifically, the PCGM-based strain sensor installed on the face can be compressed as the skin shrinks, resulting in changes in the resistance of the PCGM hydrogel [Fig. 4(b)]. For example, the PCGM-based strain sensor installed at node 1 can produce feedback signals when people are angry and thinking, indirectly monitoring their emotional state [Figs. 4(c) and 4(d)]. The PCGM-based strain sensor installed at node 2 can recognize facial expressions when a person is laughing [Fig. 4(e)]. Moreover, the PCGM-based strain sensor installed at node 3 recognizes the facial expressions of a person while smiling and chewing, and due to the different degrees of skin stretching caused by the two states, the PCGM-based strain sensor generates output signals with differences, as illustrated in Figs. 4(f) and 4(g). In addition, the PCGM-based strain sensor installed on the back of the hand can also generate feedback signals for the clenching movement [Fig. 4(h)]. A series of experiments have shown that the PCGM-based strain sensor has potential applications in monitoring human joints and facial expressions, which has important supporting value for posture assessment in running. Compared to previous work,¹⁹ the strain PCGM-based sensors are more skin compatible and have higher sensing sensitivity.

In addition to achieving high-sensitivity sensing, PCGM hydrogel can also act as an efficient energy harvester combined with the PTFE film. At length, the PCGM hydrogel operates as the positive triboelectric layer, and the PTFE film functions as the negative triboelectric layer. The prepared P-TENG follows the single-electrode working mode, and the PCGM hydrogel acts as both the conductive electrode and the positive triboelectric layer, as demonstrated in Fig. 5(a). The working mechanism of P-TENG can be exhibited in Fig. 5(b). Compared to PTFE, CP, PVA, GO, and MXene have a strong ability to lose electrons. More narrowly, as the PTFE film and the PCGM hydrogel surface contact each other, electrons can transfer from the PCGM hydrogel surface to the PTFE film surface. When the PTFE film and the PCGM hydrogel begin to separate, the current can be generated in the circuit until the separation gap reaches its maximum value. When the PTFE film and the PCGM hydrogel approach each other again, a reverse current will be generated in the circuit. In the case of continuous and repeated contact separation between the PCGM hydrogel and the PTFE film, the external circuit generates continuous alternating current. From the results in Fig. 5(c), the P-TENG maximum output power can arrive at 135 µW (matched resistance: 30 MΩ) when the maximum movement distance is 3 mm, the working frequency is 5 Hz, and the pressure is 10 N. The I_sc, V_oc, and Q_sc of P-TENG can arrive at 10.36 µA, 229.85 V, and 49.24 nC, respectively (Figs. S4–S6, supplementary material). When the external force applied to the P-TENG device increases from 4 to 9 N, the P-TENG electrical outputs also increase accordingly, as presented in Figs. 5(d)–5(f). Due to the excellent flexibility of the PCGM hydrogel, the PCGM hydrogel surface can contact with the PTFE film surface more fully under high pressure, resulting in higher output performance. At a working frequency of 10 Hz, the P-TENG can maintain continuous stability in its output voltage after 60 000 consecutive operations, as presented in Fig. 5(g). Hence, the P-TENG exhibits excellent output stability, providing a guarantee for the long-term energy harvesting of the P-TENG device. Moreover, the electrical energy generated by P-TENG can exist in commercial capacitors through the rectifier circuit with different capacitance values, demonstrating excellent charging ability, as illustrated in Fig. 5(h). From experimental results, at the same time, the smaller the capacitance, the higher the voltage reached. Small capacitance values can store a limited amount of charge, resulting in higher voltages. The separation distance will influence the potential difference between the PTFE film and the PCGM hydrogel surface. In other words, the greater the separation distance, the greater the potential difference, as shown in Fig. S7 of the supplementary material.

In conclusion, we report a mixed type PCGM conductive hydrogel based on PVA, cotton paper CP, GO, and MXene. The PCGM hydrogel can act as the PCGM-based strain sensor and triboelectric nanogenerator (P-TENG) for running posture monitoring and mechanical energy harvesting. According to the results, when the strain is less than 50%, the sensitivity of the PCGM-based strain sensor (also known as gauge factor, GF) is 0.16. When the strain is between 50% and 400%, the GF value can reach 1.92. When the strain is between 400% and 600%, the GF value can arrive at 0.72. Under the low pressure region from 0 to 16 kPa, the pressure sensitivity of the strain sensor can reach 0.0164 kPa⁻¹. Under the high pressure region from 16 to 120 kPa, the pressure sensitivity of the strain sensor can reach 0.002 86 kPa⁻¹. Moreover, the PCGM-based strain sensor can be used to monitor human posture and facial expression changes. The PCGM hydrogel and PTFE can form the P-TENG device to harvest mechanical energy. Under the mechanical frequency of 5 Hz, the P-TENG maximum output power can arrive at 135 µW with a 30 MΩ load. The I_sc, V_oc, and Q_sc of P-TENG can arrive at 10.36 µA, 229.85 V, and 49.24 nC, respectively. This research provides a feasible approach for monitoring human posture and facial expression changes.

(See the supplementary material for the elastic testing experiment of the PCGM hydrogel under 3000 cycles stretching (Fig. S1), the Raman spectra of MXene and GO (Fig. S2), the chemical structure of MXene, GO, and PVA ( Fig. S3), the short-circuit current (Isc) of the P-TENG device ( Fig. S4), the open-circuit voltage (Voc) of the P-TENG device ( Fig. S5), the transferred charge (Qsc) of the P-TENG device ( Fig. S6), and the Isc and Voc of P-TENG under different separation distances (Fig. S7).

The authors have no conflicts to disclose.

Yu Zhang: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Project administration (equal); Resources (equal); Software (equal); Supervision (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). Xiaoyan He: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Project administration (equal); Resources (equal); Software (equal); Supervision (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). Chuanming Xu: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Project administration (equal); Resources (equal); Software (equal); Supervision (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.