Common counter-electrode materials include gold, platinum and carbon. The reference electrode is typically Ag/AgCl. The working electrode is divided into enzymatic reaction electrode, molecular selection electrode [115] and ion-selective electrode. When the sensitive membrane of the ion selective electrode contacts the solution, an electrical potential is generated at the interface that is directly related to the corresponding ion concentration. For example, the multiplexed sensor proposed by Mei et al. integrates five electrochemical sensors detecting lactate, glucose, pH, Cl⁻ and urea, of which the detection of lactate and glucose uses enzyme-modified electrodes, using current changes to detect concentration; pH, Cl⁻ and urea are sensed by electrodes modified with PANI, Ag treated with FeCl₃ and NH₄⁺ ion-selective membranes, respectively. As mentioned above, sensing materials are also gradually developing in the direction of multifunctional multi-modulus in the form of arrays. For example, the multiplexed sensor proposed by Mei et al. [109] integrates five electrochemical sensors, lactate, glucose, pH, Cl⁻ and urea, of which the detection of lactate and glucose uses enzyme-modified electrodes that rely on current changes to detect concentration; pH, Cl⁻ and urea are sensed by electrodes modified with PANI, Ag treated with FeCl₃ and NH₄⁺ ion-selective membranes, respectively. As mentioned above, sensing materials are helping to improve devices in a multifunctional and multimodal way in the form of arrays.

As materials science continues to advance, the application of nanomaterials has become increasingly sophisticated. Due to their unique properties, such as porosity, rigidity, and impedance, as well as their customizable electrical and wetting properties, nanomaterials have been widely used in the fabrication of sensing materials, substrates, and other components to enhance sensor performance. Lin et al. [120] utilized N-GODs to anchor a PANI matrix as a sensing electrode, resulting in improved sensitivity and stability while avoiding the cracking commonly associated with rigid electrodes. Furthermore, the modification of electrodes with metal nanomaterials can optimize sensing performance. For example, the modification of a Prussian blue electrode with gold nanoparticles (AuNPs) yielded a larger amperometric response [121]. In the fabrication of electrochemical electrodes, metal nanomaterials are often combined with redox graphene (rGO) [122,123], which serves as a supporting substrate to enhance sensitivity. Due to its high electron mobility and large active surface area, this composite often leads to significant improvements in the sensitivity and accuracy of sensors. Additionally, such nanocomposites generally exhibit excellent electrocatalytic activity [124,125]. Yu et al. [126] developed a CuxO-nanoflake (NFs)/Cu nanoparticles (NPs) nanocomposite as a sensing material. The sensor was enriched with more active sites, a low detection limit, high selectivity, and resistance to bending and twisting. Assaduzzaman et al. [9] developed a flexible, multifunctional patch based on laser-induced graphene (LIG) combined with hybrid nano-porous carbon (H-NPC). Sensors based on nanomaterials exhibit excellent electrochemical activity, reliability and a wide detection range.

Graphene: As a monolayer two-dimensional material, Graphene has long been of interest to researchers. Special variants of graphene, such as graphene oxide and reduced graphene oxide, are widely used in wearable sweat sensors. Their high surface-area-to-volume ratio provides a large number of reaction sites, excellent electrocatalytic performance and high electron mobility. Graphene is often used as a substrate for electrodes and is combined with other materials or modification methods to enhance the overall performance of the electrode. After modification with graphene, the sensing electrode shows high sensitivity, good selectivity and flexibility, enabling the detection of small-molecule concentrations such as glucose [127,128], lactate [9], cortisol [26] and uric acid [129]. Wang, Y et al. [130] synthesized reduced graphene oxide (RGO) in situ on cotton fabric (CF). This imparted softer physical properties to the graphene material while achieving a fast response and excellent selectivity. In addition, the electrodes can be fabricated by engraving graphene films into specific patterns using laser-engraving techniques. Wang, M et al. [43] fabricated an electrode based on laser-engraved graphene (LEG) and Molecularly Imprinted Polymers (MIP) that can be repeatedly regenerated in situ. This electrode can be used to assess amino acid intake levels [131]. LEG can likewise be modified with nanomaterials (e.g., AuNPs) for highly sensitive electrochemical detection. According to the research conducted by Tu et al. [132], the deposition of AuNPs via pulsed potential results in their uniform distribution throughout the mesoporous graphene structure. This enhances their catalytic ability and provides a large number of specific binding sites. Laser-induced graphene (LIG), on the other hand, produces graphene directly on polymers using lasers without needing a film [133]. Feng et al. [134] demonstrated the use of LIG and single-walled carbon nanotubes (SWCNT) in the development of a sweat sensor. The LIG fabrication process is both rapid, taking less than a minute, and inexpensive. Additionally, laser-scribed graphene (LSG) has also been shown to have potential for use in chemical sensing [135]. The emergence of LEG and LIG has made the large-scale fabrication of graphene devices increasingly promising.

MXene: MXene is a novel material discovered by Gogosti in 2011 [136]. It is a two-dimensional transition metal carbide and nitride with a tunable structure and can exhibit excellent electrical, optical, catalytic and mechanical properties when appropriately modified [137]. The formation of composites with MXene and other materials, such as Pt and polymers, can result in enhanced performance, making it suitable for use in wearable devices. Ti₃C₂T_x was the first synthesized MXene material due to its low formation energy and in wearable sweat-sensor studies; Ti₃C₂T_x is a commonly used MXene material [138]. Ti₃C₂T_x exhibits high electrical conductivity, excellent hydrophilicity, and a large number of active sites and terminal functional groups. These properties reduce signal interference and promote the adsorption of gas and water molecules, making Ti₃C₂T_x well-suited for use in the manufacture of sensors [139]. Sharifuzzaman et al. [140] reported a catalyst composed of MXene/fluoropolymer-derived porous TiO₂ nanomaterials. On the other hand, Magesh et al. [45] developed a carbon (Pearlman catalyst) using a MXene/palladium hydroxide-loaded composite (MXene/Pd(OH)₂/C) for electrochemical sensors, achieving good selectivity, sensitivity and stability. Li et al. [141] used MXene to develop a fully optimized glucose sensor. Pt/MXene nanomaterials were fabricated using Pt nanoparticles for glucose detection on MXene nanosheets, yielding a wide linear range under neutral conditions. The sensor substrate was a conductive hydrogel synthesized from MXene and polyvinyl alcohol (PVA), providing reliable stability to the sensor.

Hydrogel-based flexible sensors possess numerous properties such as compatibility, self-healing, and adhesion [142]. Due to their water content of up to 90%, they exhibit excellent biocompatibility, electrical conductivity, and flexibility. The use of hydrogel as a sensor substrate or electrode material can improve the sensing performance. The elastic hydrogel system results in a skin-like base that can significantly improve the wearability and mechanical deformability of the device. It has been used in applications such as tissue engineering [143,144], drug delivery carriers [145], soft robots [146,147], and wound dressings [148,149]. Wang et al. [150] used TPU film to combine hydrogel and conductive material to provide the sensor with excellent mechanical and electrical properties at the same time, verifying the ability of hydrogel to improve comfort.

Gao et al. [151] used porous hydrogel instead of liquid electrolyte on sensor electrodes, maintaining better conductivity while achieving improved swelling and biocompatibility. A recent study showed that glycerin gel can also achieve alternative functions while providing anti-bending and sweat absorbing effects [127]. As a new type of gel, noble metal aero-/hydrogels possess the properties of both aerogel materials and noble metal nanoparticles [152,153]. Their highly branched, fused nanowire structure imparts inherent self-healing and flexible properties, making them promising for flexible sensing [154,155]. Li et al. [156] designed a bis-structured Pt-Ni hydrogel that achieved the aforementioned excellent properties and remained stable for two months. Chen et al. [129] reported a wearable biosensing platform based on metal aerogels using three-dimensional porous bis-structured aerogels (N-rGO/Au DAs) consisting of gold nanowires and N-doped graphene nanosheets as working electrodes. The high surface area and porous structure, as well as the synergistic effect of Au and N-rGO aerogels, enable better catalytic and electron-transfer properties than mono aerogels, facilitating non-enzymatic uric acid sensors.

An organic electrochemical transistor (OECT) is a low-cost, highly sensitive device that responds to physical stimuli, chemical functionalization, and shape changes. The OECT is composed of an ion-permeable organic semiconductor channel (OSC) connecting the source and drain electrodes. Its operation involves the conversion of the ionic current between the electrodes into a modulation of the polymer film conductivity, allowing for the detection of electrolyte solution concentrations. Poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) is commonly used as a conductive pathway between the source and drain of the OECT due to its ability to amplify signals, ease of miniaturization, and excellent biocompatibility [157,158]. PEDOT:PSS-based OECTs exhibit high transconductance (≈2 mS) at low voltages (<0.5 V) [159], making them promising for use in biosensing applications. However, these OECTs are sensitive to ionic strength [160] and the PEDOT:PSS channel accepts all cation insertions from the analyte, resulting in a lack of selectivity [161]. The selectivity of OECTs for analytes in sweat can be improved through the use of an ion-selective membrane (ISM), which permits only specific ions to pass through. In 2014, Sessolo et al. [162] first proposed the incorporation of an ISM into an OECT for the detection of K⁺, inspiring subsequent research in the field of wearable sweat sensors. Coppedè et al. [163] later proposed an OECT using an ISM for the detection of Ca²⁺, with the device being fabricated by functionalizing textile fibers. It is worth noting that many recent studies on OECT sweat sensors have incorporated textiles as a platform. Textiles are easy to prepare and can be integrated into clothing, and the detection of target analytes is not limited to ions. Tao et al. [164] developed a fabric-based OECT lactate sensor that enables PEDOT:PSS to exhibit high transconductance in both depletion and accumulation modes. Fang et al. [165] addressed the issue of insufficient transconductance in fiber OECTs by modifying carbon nanotube (CNT) fiber grids. Qing et al. [166] further integrated fiber OECTs with thermoelectric fabrics, using a manufacturing substrate composed of cotton/poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate)/dimethylsulfoxide/(3-glycidyloxypropyl)trimethoxysilane (PDG) yarn, which is lightweight, strong, and sweat-resistant.

There are numerous reports on fiber-based wearable electronics, with many studies focusing on the application of fibers for physical sensing, such as pressure [167] and thermal [168] sensing. Instead, Wu et al. [169] focused on biochemical parameters to develop a flexible microelectronic fiber that can be woven into textile substrates. Detection of UA oxidation or Na⁺ concentration can be achieved by modifying ion-sensitive membranes or nanoparticles.

Garcia-Rey et al. [5] conducted a proof-of-concept study on alginate-based biological systems to evaluate the efficacy of the colorimetric determination of sweat glucose. Alginate beads were obtained by integrating enzyme assays, consisting of GOx, HRP, and TMB, into the alginate scaffold. The porosity of the beads enabled highly sensitive glucose detection, and a similar system has been used in previous studies to calibrate and detect lactate in sweat [170]. Furthermore, another report stated that the integration of TiO₂ nanotubes into alginate can provide enhanced hydrophilicity, further improving the sensing performance of this material [171]. The touch-based cortisol sensor reported by Tang et al. [116] embeds Prussian blue redox probes within a molecularly imprinted polymer (imprinted polypyrrole, PPy, film) network. The selective binding of cortisol to the imprinted PPy film impedes PB charge transfer, resulting in a rapid change in current.

Tang et al. [117] reported a WO₃-based pH sensor. According to the article, conventional pH sensing methods primarily utilize organic materials, which carry the risk of biological toxicity. As an alternative, WO₃ is an inorganic and non-toxic material that is sensitive to hydrogen ions. However, its exchange rate is very low. The embedding of lattice H⁺ transforms WO₃ from a monoclinic phase to a cubic phase (H_xWO₃), enhancing its H+ exchange capacity and reducing its resistance, thereby improving detection sensitivity and response speed.

The use of composite sensing materials is becoming an increasingly popular trend in the research into wearable sweat sensors. While individual materials may have specific limitations, composite materials can be specifically designed to balance the strengths and weaknesses of different materials to achieve the desired functionality. Nanomaterials are also gradually becoming significant in this field. The various materials mentioned above can be nanosized or combined with nanomaterials to produce interfaces or electrodes with a large number of active sites, a large surface area, and good mechanical properties, thereby optimizing sensor performance.

In the early years of wearable sensor research, various forms of sensors, including tattoos, wristbands, and patches, have been reported. In 2012, Professor Joseph Wang’s research team reported a tattoo-based electrochemical sensor [172]. Based on this research, many applications, including the detection of ammonium [173], lactate [75], and sodium [110], have been developed in vitro and attached to the skin using transfer technology. Tattoo sensors integrated with carbon fiber exhibit excellent immunity to disturbances such as stretching, bending, and skin abrasion, and have better electrical performance than screen-printed electrodes. In addition, these sensors can be combined with artistic tattoos, which has attracted the attention of many researchers. In practice, tattoo-based sensors are more like a new type of electrode. While research on tattoo-based sensors has seen significant improvements, their popularity in subsequent research has gradually declined. Katseli et al. [7] proposed a 3D-printed electrochemical ring for sweat glucose detection. The complex fabrication process of wearable sensors is simplified by embedding conductive plastic electrodes in a ring holder and using a dual-extrusion 3D printer and commercially available filaments in a one-step printing process. In addition to the above research, many sweat sensors are now manufactured in patch form, including readable disposable sensor patches and system-based patches.

Currently, the research focus has gradually shifted, first from single-parameter to multi-parameter detection. The human body system is complex and constantly in a dynamically changing environment. As such, it is difficult to fully express the physiological health status of the human body by interpreting a single parameter. Consequently, more research is now focused on multi-parameter detection, including metabolites and electrolytes [174], sweat analytes and sweat rate [175], providing data on multiple biomarkers for analysis.

From single-component components to system-level studies, Gao et al. [6,176] proposed a sensor array capable of multiplexed sweat composition analysis, in 2016. The array was connected to integrated circuits and flexible circuit boards to simultaneously detect sweat metabolites, such as lactic acid and glucose, and electrolyte ions, including pH, calcium, sodium, and potassium. This study aimed to propose a design scheme for wearable sweat sensors from the perspectives of system integration, signal calibration, and multiplexing. This groundbreaking work provided researchers with a more complete system-level design.

Multi-function sweat sensors broaden the information collection range of single-channel sensors, and further expansion (e.g., combining physical signal detection such as acoustics) is expected to provide a more comprehensive picture of human health information. Sempionatto et al. [177] report a multiplexed sensor capable of not only detecting multiple biomarkers in sweat, but also integrating an ultrasound transducer to detect blood pressure and heart rate. One of the issues that must be considered when multiplexing integrated sensors is to prevent signal crosstalk from affecting the reliability of the data from each sensor. A solution to the signal crosstalk problem was proposed that spatially separates components at optimal distances and prevents crosstalk between acoustic and electrochemical transducers with solid-state ultrasound and sensing hydrogel layers.

While the concept of multiplexed detection and system design research is on the rise, making sensor systems lighter is also an important issue in the development of complex sensors. Gao et al. [178] proposed a lightweight flexible sensor based on their previous research. To achieve this, the researchers aimed to make materials that fit the body lighter, such as paper-based and textile sensors.

A disadvantage of using sweat as a detection target is the small sample size available at rest. To overcome this limitation, flexible sensors with integrated sweat-stimulation capabilities are being developed. Sam et al. [179] used iontophoresis electrodes to deliver agonists to stimulate sweat gland secretion, which has gradually become the main method for sweat induction. Research into natural sweat production is also increasing due to its non-invasive nature.

Research on sweat sensors does not stop at simply completing the detection function. Researchers also expect them to have therapeutic capabilities. Lee et al. [128] used microneedles to design sweat sensors capable of transcutaneous drug delivery, adding medical value to wearable devices.

As wearable sweat sensors move closer to practical applications, portability and miniaturization are inevitable trends for clinical and everyday consumer markets [180,181]. One important method for achieving device miniaturization is to replace traditional standard discrete components with integrated circuits. Wang et al. [182] reported an ultra-small wearable sweat sensing system with a size of only 1.5 cm × 1.5 cm. Its core processing chip, the MS02 chip, integrates functional circuits such as amplification and filtering, which greatly reduces the overall system area while maintaining sensor performance.

Wearable sweat sensors can be roughly divided into two categories based on whether they have a sweat transport structure: patch and microfluidic-based devices. The former undergoes electron transfer or optical reactions directly upon contact with sweat, while the latter collects or transports sweat to electrodes or reagents through microfluidic channels for detection. Taking electrochemical detection as an example, when the detection target is relatively simple, sweat can be brought directly into contact with the electrode to achieve rapid real-time monitoring. However, when multiple analytes are involved, the lack of sweat sample becomes more noticeable, making it difficult to supply multiple electrodes for simultaneous analysis. Timed sampling with an absorbent patch or a regular sweat collector can solve this problem, but it increases the effort required [183]. In addition, changes in sweat rate over time and across body parts, combined with factors such as epidermal contaminants and sweat evaporation, can easily cause fluctuations in the results of sweat sensor patches [184]. By incorporating microfluidic channels, sweat can be collected and distributed to different sensing electrodes on demand, overcoming problems such as inadequate sweat sample volume, instability, and sweat evaporation. The effects of contamination can also be reduced with specially designed channels.

Microfluidics: Microfluidic systems have the advantages of simple structure, strong ductility, and storage and transport capabilities. By designing the size (e.g., capillary [115], width variation), structure (e.g., serpentine structure, fiber, reservoir distribution), and other properties of the microfluidic channels, they can acquire specific capabilities. Examples include designing numerous pathways for the multi-modal analysis of sweat [96] and building attractive fluid dynamics and reservoirs for rapid filling [185]. Microfluidics have long been used in wearable sweat sensors as channels for the rapid sampling, collection, and analysis of sweat. Microfluidic sensing patches can be mass-produced using techniques like Roll-to-Roll (R2R) printing [186], a low-cost rotary screen printing process. Biocompatible and stretchable materials including PDMS [90] (polydimethylsiloxane) and PET (polyethylene terephthalate) are commonly used to construct microfluidic channels, allowing sensors to be integrated anywhere on the body [187].

As mentioned above, the structure of microfluidic channels can be customized. However, a more complex microfluidic system does not necessarily result in better performance. For instance, the flowmeter-based sweat rate sensor reported by Kwon et al. [46] relies on a short, straight microchannel that is thermally coupled to sweat to accurately measure temperature differences and obtain flow values. Microfluidic channels can also be designed as three-dimensional structures, allowing the vertical integration of multiple modules for sampling, transport, sensing, and storage [188].

The microfluidic channel is a semi-open system that depends on capillary action to transport sweat during collection. As such, it is important to control the fluid flow within the channel. Shi et al. [118] proposed the addition of forward and reverse Tesla valves to the inlet and outlet to prevent inlet backflow, and reduce contact between the outlet and the environment. This design is highly suitable for low-flow-rate conditions, such as sweat monitoring.

With the ongoing advancement of materials science, new microfluidic systems have been developed to address the limitations of traditional hollow microfluidic channels, such as contamination, miniaturization, and diffusion of chemical agents. Currently, there are studies exploring the combination of nanofibers with microfluidic channel materials. Mei et al. [109] reported a nanofiber-based microfluidic analysis system (NFMAS). This system utilizes highly porous and hydrophilic polyimide/sodium dodecyl sulfate (PI/SDS) nanofiber films as a microfluidic network substrate for collecting and conducting sweat. The system features both electrochemical and colorimetric sensing modules, which are separated by nanofibers. These sensors exhibit excellent mechanical properties and resistance to contamination. In subsequent research, the same group continued to apply this nanofiber microfluidic technology to fabricate flexible sensors [189], with the goal of better improvement.

Another approach is to use paper-based microfluidics. By cutting paper into a channel shape, sweat can be transported by capillary action [190,191]. Also, paper can be folded to form three-dimensional structures, allowing the creation of more complex channels. Liang et al. [192] demonstrated a three-dimensional paper-based microfluidic electrochemical device (3D-PMED). The device is constructed by prefabricating a pattern on cellulose paper and folding it into a five-layer stacked structure containing a sweat collector, vertical channel, transverse channel, electrode layer, and sweat evaporator.

There are several common methods for collecting sweat using wearable sensing systems. One approach is to use external stimuli, such as iontophoresis, to induce the release of sweat from the skin. Another approach is to collect sweat produced naturally during high-intensity exercise or to absorb sweat from the skin under resting conditions [190].

Iontophoresis: Iontophoresis involves the use of a pair of positive and negative electrodes. The anode releases pilocarpine or other long-acting sweat-inducing drugs, such as carbachol [196], to induce sweating without significant stimulation. The anode used in iontophoresis is often positioned in the center of the measuring electrodes [197] to enable real-time sweat sampling or detection. However, due to the mixing of sweat and gel and the lack of dynamic sweat sampling, this method can result in a limited number of effective samples and reduced measurement accuracy [43]. By combining iontophoresis with microfluidics [188], induced sweat can be collected and directly transferred to the sensing electrode, resulting in a more stable and quantifiable sample. Bolat et al. [196] reported on a specialized wearable, single-step sensing platform that integrates iontophoresis with microfluidics. The iontophoresis electrode is positioned directly around the PDMS microchannel, stimulating sweat glands near the microfluidic inlet. This allows secreted sweat to precisely enter the channel and be delivered to the sensor. A similar approach has been used by Wang et al. [43] in their sweat sensing patches.

Hydrophilic/Hydrophobic Substrate: Zhang et al. [97] developed a stretchable, colorimetric sweat sensor that utilizes the large wettability gradient between a superhydrophobic substrate and superhydrophilic assays to collect sweat. Superhydrophilic assays are composed of elastomeric nanofiber textiles modified with SiO₂ nanoparticles, rather than the more commonly used filter papers. Superhydrophobic substrates are created by modifying the surface of elastomer substrates with nanoparticles. This elastic substrate, featuring a wetting gradient and favorable comfortability, is well-suited for non-invasive and efficient sweat collection.

Hydrogel for Natural Sweat Extraction: Skin-conforming hydrogels have a hydrophilic interface. The hydrogel that adheres directly to the skin has the ability to reduce Laplace pressure, which prevents sweat from forming on the skin surface. In addition, water-based patches are an effective matrix for metabolite extraction, and the proper thickness of hydrogel patches can substantially improve the sensor’s capability to collect biomarkers [76]. Lin et al. [76] developed a wearable hydrogel patch that leverages the above advantages to quickly collect natural sweat from the hands. Tang et al. [116] designed a sensor that can rapidly detect cortisol through fingertip contact, using a highly permeable, porous PVA hydrogel. Its detection results can be obtained within 3.5 min of fingertip contact. Patches utilizing PVA hydrogel also exhibit excellent resistance to staining [106].

Sweat Transport and Management: When continuously monitoring sweat, it is important to consider how to prevent the accumulation of old sweat, as residual sweat can saturate the device and affect its long-term performance. Saha et al. [190] introduced an osmotic wearable for lactate sensing in sweat (OWLSS) that is capable of continuous monitoring, utilizing a specially designed paper channel to effectively manage sweat. Adding an evaporation pad to the channel and creating a serpentine channel path to increase the distance sweat must travel before reaching the evaporation pad can help delay channel saturation. Increasing the area of the evaporation pad also ensures that sweat evaporates faster than it flows into the channel under all conditions, allowing for continuous monitoring.

Wearable sweat sensors require real-time monitoring over extended periods of time, necessitating the use of wireless signal transmission. Technologies such as Bluetooth, NFC, and RFID are commonly employed in the electronic systems of wearable sweat sensors due to their advantages of miniaturization and low power consumption. The communication and control of wearable sensors consume energy, particularly in the signal processing and wireless transmission circuits, such as those utilizing Bluetooth or radio frequency transmission [200,201,202]. However, the traditional battery power supply method hinders the continuous operation of the sensor on the skin due to its difficulty in deformation, poor battery life, and the presence of harmful substances. To address the issue of energy consumption, new battery materials suitable for use in wearable devices have been developed. In their study, Liu et al. [203] mentioned the use of super-resilient hard carbon nanofabrics (s-HCNFs) as carbonaceous anode materials. This high-capacity material exhibits excellent mechanical stability under twisting and folding, and its properties can be further enhanced by designing it as a 3D micro- or nanolayered structure. These properties make s-HCNFs well-suited for use in the fabrication of flexible wearable devices. Moreover, researchers are also applying scientific research results such as self-generation, wireless power supply, and flexible batteries to integrate them into wearable sensors.

NFC: The advantages of NFC include low power consumption, few components, and small size. Only small coils and NFC chips are required to achieve wireless signal transmission [204]. NFC chips and sensors can be magnetically coupled to form a stable system [90]. The modulated antenna can also provide wireless energy supply [35]. Cheng et al. [198] developed a sweat cortisol sensor that uses NFC for wireless communication and power. When detection is required, the NFC antenna is powered after the smartphone approaches the patch to power the circuit and begin measurement. The performance of the NFC antenna is also verified under different radii of curvature to ensure that the NFC chip can still provide stable power when the patch is bent.

Biofuels and Sweat-Activated Batteries: Sweat has the potential to provide electricity support for epidermal electronics as a sustainable bioenergy source [205,206]. Biofuel cells provide power based on the bioenergy in sweat. Yu et al. [200] report on a lactate biofuel cell based on full perspiration-powered electronic skin (PPES). The enzyme-catalyzed batteries exhibit a higher current density and stability. The electrolytes in sweat (Na⁺, K⁺, Cl⁻) can also be used for electrochemical energy storage. Chen et al. [207] designed an array of sector electrodes (using copper and zinc thin films) to create a sweat-based energy generator based on redox reactions. Manjakkal et al. [208] report a supercapacitor using PEDOT: PSS as the active electrode and current collector. This material has been used in sensitive electrodes of electrochemical sensors [6] due to its high conductivity, large voltage window, and good mechanical properties.

In response to the need for flexibility, Liu et al. [199] developed a flexible sweat-activated battery that can be directly integrated onto the skin. Soft silicone is used for fixing and loading functional materials as a package housing. Two Nylon fabric bags containing electrolyte are used to stabilize the output voltage under various deformation conditions.

Triboelectric Nanogenerator: Since the invention of triboelectric nanogenerators (TENGs), many sensor platforms incorporating TENGs have been reported [209,210,211] due to their advantages of a high output voltage, power, low cost, and ease of fabrication [212,213,214,215]. In sweat-sensing applications, methods for optimizing TENGs are also being explored. Zahed et al. [216] used a PVDF/Co₃O₄ nanofiber-based TENG and a miniature Halbach magnet array for EMG to improve its electrical energy conversion efficiency. Baro et al. [217] proposed a textile-based sweat sensor incorporating a ZnO-based unipolar friction nanogenerator (STENG). Their study experimentally verified that hydrated salts in sweat promote an increase in conduction band electrons in ZnO, allowing the integrated STENG to output higher voltages. TENGs can also be integrated with flexible supercapacitors (SCs) to form a distributed power system, further enhancing the power-management capabilities of wearable devices [218]. The construction of self-charging power systems (SCPS) for TENG-SCs requires consideration of two key factors. Firstly, the AC power output by the TENG cannot be directly fed into the SC and must be converted using a rectifier. Secondly, it is difficult to share electrodes between the TENG and the SC, and alternative solutions such as shared substrates or packages are typically used [219]. These can include fabrics [220], thin films [221], carbon cloth [222,223], hydrogels [224], and others. The integration of TENGs with energy-storage technologies has led to the development of uninterrupted power supply (UPS) systems for wearable devices [225]. These systems provide a viable solution for the continuous power supply of wearable devices.

Solar Power: Singh et al. [226] installed a solar panel on top of a sweat sensor in the form of a smart watch to collect energy. A circular solar panel with a diameter of 2.2 inches was mounted on the top of the device as an energy source to power the integrated circuit of the smart watch. This combination provides a new idea for the power supply scheme of wearable devices.

Inertial Energy Harvesters: Inertial energy harvesters can convert the kinetic energy generated by human motion into electrical energy to power devices. These harvesters contain a proof mass that is excited by human motion, with the inertial energy of the motion being converted into electrical energy through piezoelectric, electrostatic, or electromagnetic methods [227,228]. To address the power density limitation imposed by the volume of the proof mass, Cai et al. [229] proposed an alternative approach. They utilized a planetary structure with a power-generating unit consisting of a base, coils, rotor, and magnets as an eccentric mass, thereby avoiding the use of a proof mass. In walking tests, the power output of this design reached 1.84 mW at the wrist and 2.95 mW at the ankle. The output of energy harvesters is influenced by factors such as the frequency of human movement and the placement of the device. To address this, Hoareau et al. [230] proposed a method for determining the optimal placement of cantilevered piezoelectric generators (PEGs). This method uses electromechanical modeling and accelerometer data to identify the optimal power source for cantilevered PEGs. According to the scenario presented in the article, the optimal placement for the device is along the axis normal to the surface of the right hand. Additionally, extremity segments such as the feet, hands, and forearms were identified as better energy sources. Beach et al. [231] compared the amount of energy harvested from the wrist, hip, ankle, and foot. Based on their analysis of the spectral content, optimal harvester parameters, and the average power output, it was concluded that the foot is able to provide more power, utilize a larger area, and exhibit a better broadband response during walking. Additionally, the foot was found to be least affected by changes in walking speed. To further improve energy harvesting, Sandhu et al. [232] proposed combining solar and kinetic energy harvesting as both energy and activity sensors.

Thermoelectricity: Thermoelectric systems convert the temperature difference (ΔT) between two interfaces into electricity. The ΔT between the human body and the environment is generally stable, making thermoelectric generators (TEGs) a promising power for providing a stable energy supply. However, TEGs require a high temperature difference [233] or a booster [234] to provide sufficient voltage, which unfortunately reduces the flexibility of portable devices and increases the complexity of the system. To solve the issues mentioned above, fiber-based TEGs can be structurally optimized to increase the output voltage [235,236]. The all-fiber thermoelectric sensing device proposed by Qing et al. [166] integrates the fiber-based organic electrochemical transistor (FOECT) and thermoelectric fiber (TEF). Such a power supply has sufficient output and does not require energy storage devices.

Despite promising research into self-powered modules, both biofuel cells and energy harvesters still face challenges in terms of energy efficiency, cell size, and durability. As a result, the dominance of lithium batteries remains difficult to challenge. A long-term vision and sustained efforts are required to fully transition from traditional rigid batteries to flexible power sources suitable for wearable systems.