Whole garment knitting-enabled direct current triboelectric nanogenerators

INTRODUCTION

In the era of rapid technological advancement, the advent of the Internet of Things (IoT) has significantly enhanced daily life by enabling seamless communication between devices^[1,2]. This interconnectivity has not only transformed the way we interact with technology but also revolutionized the field of wearable electronics^[3,4]. Traditionally used for clothing and warmth, fabrics have now evolved into intelligent materials capable of real-time monitoring of human movement and health metrics through embedded sensors^[5,6]. However, the development of wearable devices remains limited by power supply constraints. Conventional batteries, with their finite energy density, often require a trade-off between size and capacity, thus restricting the potential of wearable technology^[7,8]. Triboelectric nanogenerators (TENGs) have emerged as a promising alternative for sustainable power generation, offering efficient energy conversion within a compact structure^[9]. Fabric-based TENGs, in particular, have garnered attention due to their inherent softness and breathability, enabling seamless integration into clothing while maintaining wearer comfort. They can effectively harvest mechanical energy from daily activities and provide a continuous power supply for embedded sensors in wearable systems^[10]. However, conventional TENGs typically generate alternating current (AC)^[11,12], which is incompatible with most wearable electronics that rely on direct current (DC). The conversion from AC to DC introduces additional energy loss and system complexity. Therefore, direct current TENGs (DC-TENGs), which eliminate the need for rectification, represent a more efficient strategy^[13-15].

Existing DC-TENGs can be classified into three main mechanisms: air breakdown^[16-19], the tribovoltaic effect^[20-23], and mechanical rectification^[24-27]. Air breakdown-based TENGs are particularly notable due to their relatively simple structure and superior output performance. This makes them highly suitable for applications requiring high power output and stability. Furthermore, the Van de Graaff generator’s design inspired air breakdown DC-TENGs^[28]. Yang et al. developed the first belt-type DC-TENG that transfers charges via moving belts^[29]. Unlike the Van de Graaff unit for high-voltage applications, this DC-TENG employs self-generated high voltage to trigger air breakdown, thereby spontaneously equalizing interfacial potentials and delivering direct current output. Liu et al. further simplified the design by suspending electrodes to establish air-breakdown gaps^[30]. This approach eliminates belt-mediated charge transfer, directly harvesting frictional charges between two materials while consolidating charge-transfer and charge-collection functions using the same conductive material. Similarly, Chen et al. used an acrylic board with an uneven surface as the substrate and strategically placed two sets of conductive yarns in the fabric at the highest and lowest positions of the acrylic board as the frictional electrode and electrostatic breakdown electrode^[31], respectively. This innovative design enabled the development of a fabric-based DC-TENG utilizing air breakdown. However, despite these significant advancements, fabric-based DC-TENGs through air breakdown still face several challenges. These include limited practical applicability and complex post-processing procedures. For example, polymer coatings on yarns or fabrics and additional sewing steps are often required, making the fabrication process cumbersome and less practical for large-scale applications^[4,10]. Therefore, simplifying the fabrication process and enhancing wearability remain critical for advancing fabric-based DC-TENGs toward real-world wearable applications.

In addition, TENGs typically exhibit high voltage, low current, and extremely high internal impedance, leading to low energy utilization efficiency when directly driving low-impedance electronic devices. To address this challenge, effective power management systems are required to improve energy utilization in practical applications^[32-35]. The earliest power management for TENGs can be traced back to 2013, when Cheng et al. utilized the motion of TENGs to delay energy release^[36], thereby obtaining high-peak-pulse currents. This type of power management is a kind of mechanical rectification that needs to be used in combination with the motion of TENGs and switches. In 2020, Harmon et al. developed a self-powered power management system (PMS)^[37]. Compared with the previous methods, the PMS does not rely on the specific motion patterns of the TENG to control the switch^[38], nor does it require any special structural design^[39,40]; it only requires charging the front-stage capacitor to complete the switch-off operation. Cheng et al. developed two kinds of fabric-based DC-TENGs that utilize air breakdown by coating the fabric with conductive silver paste as electrodes^[41,42]. Among them, Ref.^[42] fabricated an energy management module (EMM), which, after processing through the EMM, is capable of lighting up four 30 W commercial light bulbs, demonstrating remarkable output performance and highlighting the significant potential of power management circuits in enhancing energy output.

In this study, we present a knitted fabric-based DC-TENG (KF DC-TENG) that utilizes air breakdown and exploits the unique properties of knitted structures. Silver-coated polyamide (PA) yarns serve as electrodes, while insulating PA yarns are introduced in the insulating region to prevent inter-electrode conduction. The electrode gap can be conveniently tuned by adjusting the number of conductive knitting yarns. Whole-garment knitting technology is employed to create a stable three-dimensional air-breakdown gap in a single step, eliminating the need for complex post-processing. Additionally, the structural asymmetry between interlock and plain stitches enables dual-sided DC output, improving mechanical energy utilization efficiency. To address the low energy utilization issue of TENGs, a low-cost power management circuit (PMC) costing only ~1.5 CNY is developed, which boosts the energy output by nearly 290 times and enables stable power supply for electronic watches and Light Emitting Diodes (LEDs). The KF DC-TENG exhibits excellent softness and breathability and is compatible with standard knitting computer aided design (CAD) programs, facilitating scalable industrial production. After systematic optimization, a single unit (total geometric area of 8 × 2.5 cm²) can light 744 series-connected LEDs when rubbed against polytetrafluoroethylene (PTFE) fabric, while four units interfaced with fluorinated ethylene propylene (FEP) film can light 1,392 LEDs. Notably, the device maintains stable output even after 14,000 friction cycles, demonstrating excellent durability. With the PMC, the KF DC-TENG delivers a peak pulse current of 20 mA under no-load conditions, charges a 100 μF capacitor to 20 V within 1 min, and can power an electronic watch for 8 minutes after only 17 seconds of charging. These results demonstrate the facile fabrication and high output performance of the KF DC-TENG, revealing its strong potential for wearable energy-harvesting applications.

EXPERIMENTAL

RESULTS AND DISCUSSION

The human understanding of electricity dates back to observations of lightning, and modern technologies have enabled its controlled use to improve daily life. Here, we harness electrostatic breakdown generated by fabric friction - analogous to “lightning” within textiles - to power electronic devices. However, constructing a three-dimensional breakdown gap using inherently two-dimensional fabrics remains a significant challenge in fabric-based air breakdown DC-TENGs. The structural characteristics of knitted fabrics offer an effective solution to this challenge. Fabrics knitted on a double needle bed are typically thicker than those from a single needle bed, and this thickness difference can provide the gap necessary for air breakdown, enabling efficient energy conversion in DC-TENGs. Accordingly, we employed interlock stitch (double-needle-bed) as the friction electrode and plain stitch (single-needle-bed) as the breakdown electrode [Figure 1A]. To prevent electrical conduction between the two electrodes, they were knitted with separate sets of conductive yarns, with an insulating PA plain stitch section inserted between them. For a more in-depth explanation of the design principle and methods to address the potential cross-conduction issue, see Supplementary Note 1. As each electrode is formed from a continuous set of yarns, the friction and breakdown electrodes are inherently connected within their respective circuits [Figure 1B]. The detailed machine diagram and knitting process are provided in Supplementary Figure 1 and Supplementary Note 2. All yarns used are commercially available. Silver-coated PA yarns exhibit excellent electrical conductivity: six 10 cm strands of 100D silver-coated PA show a resistance of only 9 Ω, which decreases to 5 Ω and 2 Ω after being knitted into the friction and breakdown electrodes, respectively [Supplementary Figure 2]. Owing to the interlocking loop structure of weft-knitted fabrics, the inherent gaps between loops provide high breathability. Air permeability testing revealed an average value of ~1,500 mm/s at a pressure difference of 100 Pa for a 20 cm² test area [Supplementary Figure 3]. Loop size, and thus breathability, can be tuned by adjusting the needle lowering depth during knitting, with greater lowering depths resulting in higher air permeability [Supplementary Note 3].

Figure 1. Structure and working mechanism of the KF DC-TENG. (A) Simulation diagram of KF DC-TENG, (B) actual photograph of KF DC-TENG, (C) motion cycle diagram of KF DC-TENG, (D) corona discharge, (E) charge output of single-contact and double-contact, (F) energy harvesting and optimization, (G) output comparison of TENG with and without power management circuit. TENG: Triboelectric nanogenerator; KF DC-TENG: knitted fabric-based direct current TENG; PMC: power management circuit; AC: alternating current; PA: polyamide.

The working principle of the KF DC-TENG under contact sliding is shown in Figure 1C. When the fabric contacts FEP (stage ⅰ), the breakdown electrode, supported by the friction electrode, forms a small air gap. Due to the different electronegativities of various materials, FEP, owing to its strong electronegativity, can capture electrons from the surface of the silver-coated PA. As the KF DC-TENG slides to the right side (stage ii), a high-density charge layer forms on the FEP surface after friction. When the breakdown electrode passes over this charged surface, a strong electric field develops between the FEP and the electrode. Once the electric field reaches a critical strength, charges are ejected from the FEP surface, ionizing the air to form a conductive channel. These charges are collected by the breakdown electrode and returned to the friction electrode via the external circuit. Meanwhile, the friction electrode continuously transfers electrons to the FEP surface through the triboelectric effect during the sustained contact and friction with FEP. When the KF DC-TENG slides to the left (stages iii and iv), the same process occurs. To demonstrate that air breakdown indeed occurs during fabric sliding, time-lapse photography was employed to capture the glow discharge in a completely darkened test environment, as shown in Figure 1D. Due to the unevenness of the fabric, air breakdown also occurs near the friction electrode. Additionally, the interlock stitch, which forms raised areas on both sides of the fabric, enables both sides of the KF DC-TENG to generate output through friction [Supplementary Figure 4]. When both sides are rubbed simultaneously, the charge generated by three cycles of double contact is higher than that generated by single contact [Figure 1E]. Simultaneous sliding of both sides allows both the front and back surfaces of the friction electrode to contact FEP, thereby facilitating charge transfer. Therefore, air breakdown occurs on both sides of the breakdown electrode, which leads to a higher output compared to single-sided contact. Details of the testing method, along with the measured current and voltage, are presented in Supplementary Figure 5. Notably, the DC-TENG can power electronic components without the need for rectification by a rectifier bridge [Figure 1F]. However, the inherently high impedance of the TENG limits its ability to achieve maximum output power under low load conditions. To address this, this study presents a detailed investigation of a PMC, which enhances the current output by nearly 80 times compared with the original fabric [Figure 1G].

To evaluate the fabric’s output under different conditions, we employed the testing platform shown in Figure 2A, which allowed precise control of parameters such as speed and displacement. The acceleration, velocity, and displacement of the linear motor were set to 10 m/s², 0.1 m/s, and 50 mm, respectively, while the relative humidity was maintained at 40% relative humidity (RH). Unless otherwise specified, these parameters were kept constant in all subsequent tests. To determine the optimal output structure of the KF DC-TENG, we systematically examined the influence of six factors: the number of knitted courses of the insulating plain stitch (A), the number of knitted courses of the breakdown electrode (B), the number of knitted courses of the friction electrode (C), the number of knitted wales of the fabric structure (D), the minimum number of knitting cycles (E), and the breakdown gap (F) on its output performance [Figure 2B]. To ensure a consistent number of breakdown electrodes during motion cycles, an additional breakdown electrode was incorporated into the structure. Unless otherwise specified, all samples consisted of two friction electrodes and three breakdown electrodes, fabricated using 600D silver-coated PA yarns and PA yarns. As shown in Figure 2C and D, the output decreased as A increased. Based on the fabric’s motion process [Supplementary Figure 6 and Supplementary Note 4], a larger A reduced the effective charge-collection area of the electrodes, thereby lowering the output. Notably, when A = 3, the output does not reach its maximum, because the breakdown and friction electrodes are positioned too closely. In addition, fuzz generated on the conductive yarn surface during knitting can cause short circuits between electrodes, further reducing the output. This highlights the importance of controlling the electrode spacing during knitting to prevent short circuits. As shown in Figure 2E, the current does not return to zero during fabric motion, which is attributed to surface irregularities that induce charge accumulation at sharp points on the breakdown electrodes. Consequently, multiple electrostatic breakdown events occur simultaneously during sliding, and their superposition prevents the current from returning to zero. As the value of B increases, the breakdown electrodes gain more irregular sharp tips. Although these tips do not enlarge the effective charge-collection area [Supplementary Figure 7], they form strong local electric fields with the FEP film, enabling more breakdown events to occur during fabric sliding and thereby increasing the output [Figure 2F]. As shown in Figure 2G and H, the output decreased as the value of C increased. This reduction is attributed to decreased contact pressure on the friction electrode. Under a constant normal force, increasing the electrode area reduces interfacial contact with the FEP film, thereby limiting charge transfer. Furthermore, reduced pressure enlarges interfacial gaps, promoting air breakdown and decreasing the surface charge density of FEP, which is consistent with the corona discharge observed in Figure 1D. As shown in Figure 2I and J, the output increases as D increases. As the friction electrode area expands, the sliding contact area on the FEP surface also increases, thereby enlarging the effective air breakdown region. Furthermore, as shown in Figure 2K and L, an increase in the knitting cycle number E resulted in an increase in the output. Compared with E = 1, E = 2 introduces an additional central electrode, resulting in higher output. Furthermore, a larger E corresponds to the participation of more breakdown electrodes, leading to enhanced output.

Figure 2. The Influence of Fabric Structure on the Output Performance of KF DC-TENG. (A) Test platform, (B) fabric structure, (C) current and (D) charge transfer of KF DC-TENG under different insulating plain knit courses, (E) current and (F) charge transfer of KF DC-TENG under different breakdown electrode knit courses, (G) current and (H) charge transfer of KF DC-TENG under different friction electrode knit courses, (I) current and (J) charge transfer of KF DC-TENG under different wales, (K) current and (L) charge transfer of KF DC-TENG under different knit cycles. TENG: Triboelectric nanogenerator; KF DC-TENG: knitted fabric-based direct current TENG; FEP: fluorinated ethylene propylene.

Finally, we investigated the effect of the breakdown gap on the output [Supplementary Figure 8]. In this context, “FE 600D-1” refers to a friction electrode knitted using 600D silver-coated PA yarn with E = 1, while “BE 600D-2” denotes a breakdown electrode knitted using 600D silver-coated PA yarn with E = 2. As shown in Supplementary Figure 9, fabrics knitted from yarns with different linear densities exhibit distinct thicknesses. This thickness difference between the friction electrode and the breakdown electrode serves as the breakdown gap, thereby enabling air breakdown. According to Paschen’s law formula^[43], the breakdown voltage V_b satisfies the following equation (1):

where A denotes the saturation ionization of gas, B corresponds to the excitation and ionization energies, and γ is the secondary electron emission coefficient. These three parameters are constants determined by the type of gas. p denotes the atmospheric pressure, and d represents the breakdown gap. Therefore, the magnitude of the breakdown voltage is proportional to the product of the breakdown gap and the atmospheric pressure. Since atmospheric pressure is typically constant, variations in the gap directly cause the breakdown voltage (V_b) to change within a certain range. As shown by the curves in Supplementary Figure 8A-C, as the linear density of the friction electrode yarn gradually decreases (leading to a smaller breakdown gap), the output of the KF DC-TENG correspondingly decreases. In contrast, in Supplementary Figure 8D-F, when the linear density of the yarn used for the breakdown electrode gradually decreases (resulting in a larger breakdown gap), the KF DC-TENG output continuously increases. This behavior can be explained by Paschen’s law, which predicts a non-monotonic relationship between breakdown voltage and the product p∙d, with a minimum at an intermediate value. In this experiment, as the breakdown gap increases, it progressively approaches the minimum value. This results in the breakdown voltage V_b decreasing, leading to an observed increasing trend in output.

Moreover, we have knitted an air-layer structure to serve as the friction electrode, while the breakdown electrode remains in plain stitch structure, as shown in Supplementary Figure 10A. The air-layer structure, fabricated using a double needle bed, forms a double-layer fabric and exhibits output comparable to that of the interlock structure under identical knitting parameters [Supplementary Figure 10B-D]. This result suggests that other double needle bed knitted structures, which create a thickness difference relative to single needle bed structures, can also be applied in KF DC-TENG fabrication. In addition, all voltage curves related to the tests in this section are shown in Supplementary Figure 11, and their trends are essentially consistent with the trends of current and charge changes.

Apart from the fabric structure, the testing conditions also have a significant impact on the output of the KF DC-TENG. In these tests, the fabric parameters were set as A = 4, B = 6, C = 1, and D = 60. Figure 3A and B show the effect of sliding speed on the output performance. As the sliding speed increases, the peak current rises, whereas the total accumulated charge over three cycles decreases. The reason is that as the sliding speed increases, the amount of charge collected by the breakdown electrode within the same time period also increases. Since the magnitude of the current is proportional to the amount of charge collected in the same time period, the higher the sliding speed, the higher the peak current. However, the total collected charge decreases because the linear motor must accelerate to high speed and decelerate to zero within a very short time. The rapid speed change generates strong inertia, causing slight vibrations in the testing device. This reduces the contact tightness between the KF DC-TENG and the FEP film, thereby lowering the output. Similarly, when the magnitude of the acceleration is insufficient to reach the maximum speed within a displacement distance of 50 mm, the peak current will be significantly reduced [Figure 3C]. Moreover, when the acceleration of the fabric motion increases, the time required for the fabric to reach the maximum speed is reduced. As a result, the longer the fabric remains at high speed during unidirectional motion, the smaller the total collected charge becomes [Figure 3D].

Figure 3. The Influence of Testing Conditions on the Output Performance of KF DC-TENG. (A) Current and (B) charge transfer of KF DC-TENG under different accelerations, (C) current and (D) charge transfer of KF DC-TENG under different velocities, (E) current and (F) charge transfer of KF DC-TENG under different displacement distances, (G) current and (H) charge transfer of KF DC-TENG under different weights, (I) current and (J) charge transfer of KF DC-TENG under different humidities, (K) current of KF DC-TENG after 14,000 friction cycles. TENG: Triboelectric nanogenerator; KF DC-TENG: knitted fabric-based direct current TENG.

As shown in Figure 3E and F, increasing the displacement distance increases the total collected charge while the peak current remains essentially unchanged. This is because the peak current is proportional to speed, and changing only the displacement does not affect speed. A longer displacement provides more time for air breakdown during unidirectional motion, thereby increasing the total collected charge. Besides, as previously mentioned, increasing the number of knitting courses of the friction electrode reduces the output due to decreased pressure. To verify this, weights were used to test the effect of pressure on the KF DC-TENG output [Figure 2A], and the results confirmed that increasing the applied weight enhances the contact between the fabric and FEP film, thereby improving the output performance. As presented in Figure 3G and H, both the peak current and the collected charge increased with the added weight, consistent with the prior analysis. However, when the weight reached 2 kg, the collected charge declined due to deformation of the unsupported acrylic plate, which reduced contact intimacy between the FEP film and the fabric. Furthermore, the fabric naturally maintains a moisture absorption and desorption equilibrium with the external environment and the human body during the wearing process. As shown in Figure 3I and J, the output continuously decreases with increasing humidity, as higher humidity accelerates charge dissipation in the air and thus reduces the total charge involved in air breakdown. We also evaluated the output performance after water washing. Daily laundering conditions were simulated using tap water and laundry detergent. As shown in Supplementary Figure 12, the output variation was negligible, indicating good washing resistance. The durability test indicated that after 14,000 consecutive cycles, the current output remained essentially unchanged, demonstrating excellent stability [Figure 3K]. All related voltages for these tests are provided in Supplementary Figures 13-15, showing trends consistent with current and charge measurements.

As shown in Figure 4A and Supplementary Figure 16, the power management circuit comprises three parts: the TENG that provides energy, the self-powered switch, and a buck circuit. Since the buck converter operates by continuous switching to regulate the output voltage, the self-powered switch is designed to automatically control the power supply via a rational circuit design. The self-powered switch consists of three electronic components [Supplementary Figure 17]: a high-voltage ceramic capacitor, a Zener diode, and a thyristor. The thyristor conducts only when the gate is forward biased relative to the cathode. Therefore, the operation sequence is as follows: First, the power source charges capacitor C₁, for which a low-capacitance, high-voltage ceramic capacitor is chosen. This choice not only prevents the capacitor from being damaged by high voltage during TENG charging but also allows the capacitor to reach a high voltage more rapidly. Second, when the voltage of capacitor C₁ reaches the breakdown voltage of the Zener diode D₁, the Zener diode conducts, thereby causing the gate and cathode of the thyristor to conduct. As a result, the entire thyristor conducts, discharging the energy stored in capacitor C₁ into the subsequent buck circuit. The thyristor then turns off and waits for C₁’s voltage to reach the breakdown voltage of the Zener diode again before turning on once more. The frequency of this self-powered switch is determined by the relationship between the capacitance of C₁ and the breakdown voltage of the Zener diode. Additionally, the current through resistor R in the buck circuit can be explained in three stages [Supplementary Figure 18]. Stage 1: When the self-powered switch is on, current flows from C₁ to charge capacitor C₂ and supply power to resistor R. Due to the blocking action of diode D₂, the current cannot pass through D₂ at this stage. Stage 2: After the charge in C₁ is depleted and the thyristor turns off, inductor L induces an electromotive force in the same direction as that of C₁, continuing to charge C₂ and supply power to R. The current then flows through D2 back to the inductor. Stage 3: Once the energy in L is exhausted, C₂ serves as the power source and continues to supply R. Although these three stages may appear discrete, they are in fact part of a continuous and cyclic process. When the switching frequency is sufficiently high, the current stabilizes and becomes effectively direct current. It should be noted that an electrometer cannot accurately measure transferred charge under pulsed signals [Supplementary Figure 19A and B]. However, with an inductance value of 1 μH, the charge measured by the electrometer is essentially consistent with the integral of the pulsed current. Therefore, the transferred charge is evaluated by integrating the pulsed current [Supplementary Figure 19C and D].

Figure 4. Research on Power Management Circuit. (A) Power management circuit, (B) Current and (C) Charge of the PMC under different capacitance values of C₂, (D) Discharge time of the PMC with C₂ capacitance values of 47 μF and 10,000 μF, (E) Current and (F) Charge under different inductance values of L, (G) Number of pulse currents and (H) Charge under different capacitance values of C₁, (I) current of the PMC at different loads when C₂ = 100 μF, (J) Current comparison of the PMC with C₂ capacitance values of 10 μF and 100 μF under a load of 10 kΩ, (K) Charge of the PMC under different loads, (L) Comparison of charging time for a 100 μF capacitor with and without the PMC, (M) Comparison of charging energy with and without the PMC, (N) Comparison of the output of the fabric-based TENG with other fabric-based TENGs^{[10,42,44-48]}. TENG: Triboelectric nanogenerator; KF DC-TENG: knitted fabric-based direct current TENG; PMC: power management circuit.

Figure 4B shows the current output when resistor R is replaced by an electrometer. Without the resistor, the current output appears as pulses. As the capacitance of C₂ increases, the peak current decreases, but the charge transferred by each pulse current remains essentially unchanged [Figure 4C and Supplementary Figure 20]. A larger capacitance of C₂ also results in a longer pulse duration [Figure 4D], which in turn extends the time required for the thyristor to turn off and for the voltage of C₁ to reach the breakdown voltage of the Zener diode again. Furthermore, the movement speed of the KF DC-TENG affects the charge transferred within a given time. To investigate this, the linear motor speed was varied (total displacement distance per cycle: 100 mm, 1 Hz = 0.1 m/s). As shown in Supplementary Figure 21, as the movement frequency increases, the time between two pulse currents shortens, while the charge transfer of each pulse current remains essentially unchanged. The increase in speed allows the voltage of C₁ to reach the breakdown voltage of the Zener diode more quickly, thereby enabling the PMC to generate more pulse currents in the same time. However, since the capacitance of C₁ remains unchanged, the total amount of charge released each time the switch opens stays the same [Supplementary Figure 21B]. When the fabric output is insufficient to drive the PMC, no pulse current is generated [Supplementary Figure 21C]. Moreover, the size of inductor L also affects the output, and several different inductors were used to test their impact on the output performance of the PMC as shown in Figure 4E and F, and Supplementary Figure 22. Different inductance values produced output oscillations of varying degrees. To reduce oscillations and the increased resistance brought by high inductance values [Supplementary Figure 23 and Supplementary Note 5], an inductor of 10 μH was selected. As demonstrated in Figure 4G, reducing the capacitance of C₁ increases the number of pulses generated within a given time, as a smaller capacitance allows faster charging and more rapid attainment of the Zener breakdown voltage. However, a larger capacitance stores more energy at the same voltage [Figure 4H and Supplementary Figure 24]. When C₁ = 22,000 pF, the pulse current reaches 20 mA and the transferred charge reaches 69.36 μC. To obtain a more stable and higher pulse frequency from the PMC, C₁ = 100 pF was selected for subsequent tests. After the above analysis, load tests were conducted to evaluate the PMC performance (C₁ = 100 pF, C₂ = 100 μF, breakdown voltage of D₁ is 50 V, L = 10 μH, movement speed is 0.4 m/s, A = 4, B = 6, C = 1, E = 4). As shown in Figure 4I and Supplementary Figure 25, the peak current gradually decreases as the resistance value increases due to the impeding effect of the resistor on the flow of electrons through the conductor. Nevertheless, the output remains stable without reverse current signals. At a load resistance of 10 kΩ, the peak current reaches 80 μA, demonstrating stable current output under practical operating conditions. Changing the capacitance of C₂ from 100 μF to 10 μF increases the peak current to 160 μA, as shown in Figure 4J. Although this change enhances the peak current, it also results in a highly unstable current curve. This observation is consistent with previous findings that larger capacitances of C₂ correlate with lower peak currents, which indicates that C₂ has current-stabilizing and filtering functions. Additionally, the charge variation under different loads was analyzed [Figure 4K]. Under zero load without the PMC, the peak current and voltage of the KF DC-TENG output reached 4 μA and 2,000 V, respectively. However, due to its pulse-like current output, the charge image rises in a stepwise manner [Supplementary Figure 26]. In contrast, the charge variation curve after the PMC is similar to a straight line, indicating a more stable output. The KF DC-TENG, after being connected to the PMC, can collect 25 μC of charge in just 0.35 s under a load of 10 kΩ whereas the original one needs to take 12 s to reach this under no load. This significant reduction in charging time underscores the PMC’s ability to enhance charge collection efficiency and output stability under low load conditions. To demonstrate the optimized output performance of the PMC, as shown in Figure 4L, the KF DC-TENG after the PMC can charge a 100 μF capacitor to 20 V within one minute, while without the PMC, it can only charge to 2.2 V in two minutes. The output power density is increased from the original 0.234 mW/m² to 68 mW/m², an increase of nearly 290 times [Figure 4M], which is also at a relatively high level compared with other fabric-based TENGs [Figure 4N and Supplementary Table 1]. Furthermore, the entire PMC circuit costs only about 1.5 CNY [Supplementary Table 2], making it a highly cost-effective solution. This low cost not only improves accessibility but also enables seamless integration into a wide range of everyday devices and systems, underscoring its potential for large-scale applications. It should be noted that the yarn specifications used in this section and subsequent tests are 600D for the friction electrode and 400D for the breakdown electrode.

When the TENG is employed to power electronic components, its discontinuous output during the motion process necessitates the use of a capacitor before connecting to the load. Consequently, the ability of the TENG to charge a capacitor is of critical importance. Figure 5A illustrates the circuit used to evaluate the charging performance of the TENG. As shown in Figure 5B, smaller capacitances result in faster charging rates. For example, a 10 μF capacitor can be charged to 5 V within 50 s by a single unit of the KF DC-TENG. As the number of knitted units in the KF DC-TENG increases, the charging speed further improves [Figure 5C]. Since the KF DC-TENG has a direct current output, it can directly power electronic components without any rectification equipment [Figure 5D]. When we manually rubbed a single unit of the KF DC-TENG (A = 4, B = 6, C = 1, area 8 × 2.5 cm²) on the FEP surface, it was able to light up 744 LEDs (2 V each), as shown in Figure 5E and Supplementary Movie 1. To demonstrate its performance in fabrics, a knitted plain fabric made of PTFE was used to rub against the KF DC-TENG, and it again lit 744 LEDs with a single unit of the fabric, showing its potential for application in fabrics [Figure 5F and Supplementary Movie 2]. A four-unit KF DC-TENG (A = 4, B = 6, C = 1, speed 0.2 m/s, weight 2 kg) successfully lit 1,392 LEDs [Figure 5G and Supplementary Movie 3]. Furthermore, the performance of the KF DC-TENG embedded into fabrics was also investigated. In practical applications, repeated rubbing may induce twisting and deformation, which can degrade electrical output. To address this issue, the KF DC-TENG was integrated with a spacer fabric, which effectively stabilizes the structure and enables efficient harvesting of human motion energy when embedded into fabrics, as shown in Supplementary Figure 27 and Supplementary Movie 4.

Figure 5. Actual Practical Applications. (A) Schematic diagram of the capacitor charging circuit, (B) voltage plots for charging different capacitors with KF DC-TENG when E = 1, (C) voltage plots for charging a 47 μF capacitor with KF DC-TENG at different E values, (D) circuit diagram of DC-TENG directly powering a resistor, (E) manually lighting up 744 LEDs with KF DC-TENG when E = 1, (F) lighting up 744 LEDs with KF DC-TENG by friction with PTFE fabric when E = 1, (G) lighting up 1392 LEDs with a four unit KF DC-TENG, (H) circuit diagram of DC-TENG powering electronic devices through a PMC, powering (I) a wristwatch, (J) a calculator, (K) a hygrometer, and (L) a 1.5 W commercial light bulb through the PMC. Photos in (E-G), (I-L) are taken by the authors. TENG: Triboelectric nanogenerator; KF DC-TENG: knitted fabric-based direct current TENG; PMC: power management circuit; LEDs: light emitting diodes.

After integration with the PMC [Figure 5H], a four-unit KF DC-TENG (A = 4, B = 6, C = 1, speed 0.4 m/s, stroke 100 mm, C₁ = 100 pF, unchanged in subsequent tests) only took 17 s to charge a 1,000 μF capacitor, which enabled a watch (3 V) to operate continuously for 8 min [Figure 5I and Supplementary Movie 5]. Additionally, a calculator (1.5 V) could also be powered and used, and it stopped working only when the capacitor voltage dropped below 1 V [Figure 5J and Supplementary Movie 6]. Furthermore, the energy generated by the KF DC-TENG after the PMC could power a hygrometer (1.5 V) continuously [Figure 5K and Supplementary Movie 7]. A commercial 1.5 W LED light could be continuously lit without noticeable flickering (speed 0.8 m/s), as shown in Figure 5L and Supplementary Movie 8. These applications demonstrate the excellent mechanical energy harvesting capability and promising application potential of the KF DC-TENG with the PMC.

CONCLUSIONS

In summary, this work presents a KF DC-TENG constructed from conventional interlock and plain stitch structures, aiming to realize wearable energy harvesting via air breakdown in textiles. The KF DC-TENG is designed via knitting CAD and manufactured by whole-garment knitting technology, which avoids complicated post-processing and supports direct one-step fabrication. The structural asymmetry between double-needle-bed and single-needle-bed knitting enables dual-sided energy conversion, offering a universal design strategy for knitted direct current generators. This study provides a simple, scalable, and wearable approach for fabric-based energy harvesting, with good compatibility to industrial textile manufacturing. Benefiting from its softness, breathability, and structural stability, the KF DC-TENG shows promising potential for wearable electronics. With further structural optimization and system integration, this kind of knitted direct current generator is expected to offer a practical pathway toward self-powered wearable systems and smart textiles in the future.