Structural superlubricity triboelectric nanogenerator with negligible wear and high triboelectrification efficiency Authors: Chaojie Chen, Jinhui Nie, Ying Liu, Xiaojian Xiang, Xuanyu Huang, Quanshui Zheng, Yunlong Zi Published: 2026 ABSTRACT AbstractTriboelectric nanogenerators (TENGs) can efficiently convert weak mechanical energy into electricity, making them suitable for small-scale and distributed power supply demanded by the Internet of Things. However, the inevitable friction and wear not only cause high energy dissipation and low triboelectrification efficiency but also severely limit the device’s lifetime and reliability. Here, a wear-free structural superlubricity TENG (SSL-TENG) is designed using the graphite-SiO2 pair. By achieving an atomic-level friction interface via micromachining, the device operates in a stable SSL state with a near-zero coefficient of friction (0.0034). As compared to mosaic charge distribution in previous studies, the intimate contact at the SSL interface suppresses air breakdown, enabling a unipolar charge distribution, which yields a high charge density of 0.47 mC/m2 and a 106-fold enhancement in triboelectrification efficiency. Simultaneously, the SSL-TENG demonstrates stable output performance for over 1.1×105 cycles without wear. This work provides a fundamental strategy to eliminate interfacial friction and the air-breakdown limit, paving the way for ultra-reliable and high-output energy harvesting.Subject terms: Devices for energy harvesting, Electronic properties and materials Introduction The rapid development of the Internet of Things and artificial intelligence has driven the general trend of intelligence in all walks of life1–3. Sensors, as critical components of these systems, play an irreplaceable role in collecting environmental data, enabling AI-driven applications in smart cities4,5, autonomous vehicles6,7, and industrial automation8. However, as sensors become increasingly miniaturized and their number proliferates, sustainably powering them over an extended period without frequent maintenance or battery replacement poses a significant challenge. To address this issue, energy harvesting technologies have emerged as a promising alternative, offering sustainable and maintenance-free power sources9,10. Triboelectric nanogenerators (TENGs), which operate based on the coupling effect of triboelectrification and electrostatic induction11, have gained considerable attention for applications in wearable devices, self-powered sensors, and other fields12–14. However, solid-solid friction is inevitable during repeated mechanical contact and separation cycles to generate charges for TENGs, which causes significant energy loss and decreases triboelectrification efficiency (TE) and energy conversion efficiency (ECE). Furthermore, the continuous contact-separation or sliding motions inevitably lead to material wear and degradation, thereby diminishing electrical output performance and limiting the life span15,16. Many efforts have been made to decrease friction and wear for improving TE and durability, among which common strategies include soft contact17,18, rolling friction19,20, and adding lubricant21,22. To further decrease the coefficient of friction (COF), specific material pairings and lubricants were applied to achieve ultra-low COF values of below 0.01. Specifically, Zhang et al.23 developed a tribovoltaic-effect device using the steel ball and hydrogenated diamond-like carbon (DLC), achieving a low COF of 0.006. Bu et al.24 reported a sliding TENG with the same material combination achieving COF of 0.0084. Although these approaches make notable progress in reducing COF, devices still suffer from material wear. On the other hand, previous works have primarily emphasized electrical output and introduced figure-of-merits (FOM)25 for electrical output performance evaluation while ignoring mechanical input due to the lack of related data about friction. To address this limitation, Xia et al. first proposed the concept of TE, which quantifies the charge generation property of triboelectric pairs by simultaneously considering both input driving force and output charge amount15. They also demonstrated that the amorphous metallic glass enhances the TE by more than twice compared with conventional Cu-polymer pairs, highlighting the potential of TE as a key index for TENG optimization. Simultaneously, the ECE is another parameter for the overall energy conversion process, while existing reports regarding ECE were usually related to non-contact sliding TENGs or liquid-solid TENGs, which cannot represent most common situations that involve in-contact solid-solid frictions. In this work, we employed structural superlubricity (SSL) to minimize energy dissipation in TENG, thereby developing a wear-free SSL-TENG based on graphite microflake (Gra MF) and SiO2. By using micromachining technology, we fabricated the silicon oxide layer with atomic-level roughness, facilitating a stable SSL state. The friction behavior of SSL-TENG was investigated using lateral force microscopy (LFM) and Raman spectroscopy, and a low COF of 0.0034 was observed with no detectable wear. Crucially, the atomic-scale contact at the SSL interface suppresses air breakdown, enabling a unipolar charged distribution. Consequently, the device achieves a charge density of 0.47 mC/m2 and an energy density of 1 × 103 J/m3, with greatly promoted TE and ECE which are 106 and 104 times higher than those of the macroscale control group, respectively. Remarkably, the SSL-TENG maintains stable operation for over 105 cycles without material wear, proving its exceptional robustness. This work represents the realization of wear-free SSL-TENG with high output and ultra-long working lifetime, offering an effective strategy for next-generation power solutions. Results Design and fabrication of SSL-TENG The surface roughness induces high friction and insufficient contact intimacy, which limits the triboelectric charge generation. To quantitatively evaluate the triboelectrification efficiency of SSL-TENG, we employ a standard proposed before15, as presented in Eq. 1: Where κ is the triboelectrification efficiency, QSC is the charge transfer (unit: mC), A is the contact area (unit: m), F is the driving force (unit: N). κ characterizes the mechanical-to-electrical conversion efficiency at the sliding interface by normalizing the generated charge density against the lateral driving force. For the energy conversion process, the input mechanical energy (Ein) is first converted into electrostatic energy and stored within the TENG, and then the electrostatic energy is converted into electrical energy (Eout) that can power electronics. Therefore, the energy conversion efficiency (η) is defined as: where V is the voltage, Q is the charge transfer between electrodes, and x is the displacement. Moreover, Ein can also be expressed as: Edis is the dissipated energy that includes heat losses, deformation losses arising from elastic and plastic deformation, air breakdown, and other losses such as acoustic radiation. In contact mode TENG, the friction-induced energy dissipation and material wear are significant, making friction reduction essential for developing high-performance TENGs. SSL is a state in which two solid interfaces are in direct contact and sliding relative to each other, with extremely low friction and zero wear26, which was achieved in 201227 with subversive application prospects in the field of MEMS devices and micro-robots28–30. Specifically, when two surfaces are in incommensurate contact26, in which atomic periodicities are mismatched due to either intrinsic lattice constants or rotational twists, lateral atomic forces cancel each other and results in near-zero friction, as shown in Fig. 1a. Therefore, we adopted SSL technology to decrease the interfacial friction and energy dissipation (Edis) for developing high-TE TENGs. As shown in Fig. 1b, the SSL-TENG comprises a silicon oxide layer, a Gra MF slider and two chromium (Cr) electrodes. The cross-sectional view of the device is shown in Fig. 1c. We first fabricated the Gra MF slider by patterning the highly oriented pyrolytic graphite (HOPG) substrate and then transferred Gra MF with self-retracting motion property31 into the stator by a tungsten probe (Supplementary Figs. 1, 2). To eliminate the edge friction arising from edge-pinning effect32–35, where the friction stress in edges could be 1000 times higher than that in non-edge regions36, we employed a post thermal annealing process. This treatment increases the tensile stress and bends the Gra MF, thereby eliminating edge-surface contact (Supplementary Fig. 3)36,37. The root-mean-square (sq) roughness of Gra MF slider is 80 pm (Fig. 1d). Next, the SSL-TENG stator was obtained by depositing Cr and SiO2 on the glass substrate and polishing the SiO2 layer by the chemical-mechanical polishing (CMP) technique to realize an atomically flat surface (Fig. 1e), which exhibits a sq roughness of 229 pm. Detailed fabrication process can be found in Supplementary Fig. 4 and methods. It should be noted that the height differences between the gap area and the electrode area, as well as the surface roughness induced by deposition, were eliminated through the CMP process (Supplementary Fig. 5), which enables the SSL state, thereby reducing friction force and wear. The assembled SSL-TENG includes a Gra MF slider and SiO2 stator. To control the reciprocating motion of SSL-TENG, a tungsten probe was used to press the Gra MF slider, and a piezoelectric stage was used to move the stator forward and backward. Design, operation and characterization of SSL-TENG. a Physical mechanism of SSL. b Schematic diagram of SSL-TENG structure. The tungsten probe presses the graphite flake for reciprocating motion, generating an induced current in the external circuit. c Optical microscope image (up, scale bar 20 μm) shows the top view and scanning electron microscope (SEM) image shows the cross-section view (down, scale bar 500 nm). d Surface morphology of Gra MF (sq roughness = 80 pm). Scale bar, 1 μm. e Surface morphology of SiO2 after CMP (sq roughness = 229 pm). Scale bar, 5 μm. Friction property of SSL-TENG The friction and wear behavior of Gra MF-SiO2 interface was characterized using the LFM (Supplementary Note 1) and Raman spectrum. As shown in Fig. 2a, the lateral force increases first and then decreases as the normal force continuously increases. Notably, a negative COF is observed when the normal force exceeds 40 μN, which is attributed to the edge warping effect38. Under the extreme mechanical loading (60 μN), the lateral force decreases significantly from 0.60 μN to 0.27 μN after 256 cycles (Fig. 2b), showing the robustness of the interface under high pressure. The calculated COF is 0.0034, which verifies the SSL state (COF < 0.01) in Gra MF and SiO2 contact. The friction behavior of SSL-TENG. a The lateral force of Gra MF on SiO2 varies with normal force. Each point was tested 30 times, as marked as mean ± SD. b The evolution of lateral force in 256 cycles under high normal force (58 μN). c COF comparison between SSL-TENG and macro-TENGs. Each group was tested 7 times. d The lateral force of Gra MF-SiO2 pair during 2.4×105 cycles. The loading normal force is 58 μN. e Optical image of the observation area. The yellow rectangle is the AFM scanning area, and the white rectangle is the sliding area. Scale bar, 10 μm. Surface morphology of the sliding area f before and g after sliding. h Raman spectrum of the Gra MF interface after 2.4×105 cycles, and points 1-9 represent the test points. Raman spectrum of the sliding region i before and j after the long cycle experiment. Scale bars in (f-j) are 4 μm. The sliding displacement in LFM measurement is 6 μm, with a sliding frequency of 2 Hz. Source data are provided as a Source Data file. For comparison, common triboelectric materials, including perfluoroalkoxy (PFA), fluorinated ethylene propylene (FEP), polyimide (PI), polytetrafluoroethylene (PTFE), and ultra-thin glass (UTG), are paired with macroscale Gra slider (5×5mm2 to prepare macroscale TENGs (macro-TENGs) as a control group. They are termed as PFA-Gra, FEP-Gra, PI-Gra, PTFE-Gra, and UTG-Gra, respectively. The COFs of macro-TENGs are over 20 times higher than that of SSL-TENG, while the SSL-TENG reduces the COF by 98% compared with UTG-Gra (Fig. 2c). Notably, the Gra MF slider exhibited remarkable mechanical endurance, maintaining stable sliding performance on the SiO2 substrate for over 2.4×105 cycles (Fig. 2d), which validates the exceptional long-term durability of the SSL-based interface. The sliding region in Fig. 2e was investigated by atomic force microscopy (AFM) and Raman spectrum. Compared with the initial surface morphology of SiO2 (Fig. 2f), no sliding trace of the graphite slider and abrasive chips is observed (Fig. 2g). The Raman spectrum was used to characterize the slid Gra MF surface, and no observable defect-related D peaks (1350 cm-1) were found at the interface, indicating that the Gra MF slider remained structurally intact after cycling friction (Fig. 2h). Raman spectrum mapping in the sliding area (Fig. 2i, j) confirms the absence of both D peaks (1350 cm-1) and G peaks (1580 cm-1), verifying that no graphite debris was transferred to the substrate. Therefore, the above results demonstrate the wearless SSL contact state between Gra MF and SiO2. By contrast, the macro-TENGs show obvious material wear with friction traces on Gra slider surface and Gra debris on polymer surfaces (Supplementary Figs. 6, 7). The low friction and wear-free properties of the SSL-TENG are highly reproducible across multiple samples, confirming that these properties are intrinsic to the SSL interface (Supplementary Figs. 8, 9). Overall, it is demonstrated that the implementation of SSL technology provides a robust and versatile strategy to effectively eliminate material wear and frictional energy dissipation in TENG systems. Output performance of SSL-TENG An SSL-TENG and several macro-TENGs in sliding freestanding triboelectric-layer mode were prepared for comparison. During the reciprocating motion of the Gra MF slider, triboelectric charges are generated at the sliding interface, subsequently inducing an alternating current in the external circuit (Fig. 3a). To measure the electrical output of the SSL-TENG, the device is connected to an external circuit using wire-bonding (Supplementary Fig. 10). A key metric used for performance evaluation is the intrinsic contact electrification charge density (ICECD), which is solely material-contributed charge density without an excitation circuit39. As presented in Fig. 3b, the SSL-TENG achieves the highest ICECD of 0.47 mC/m2, which is approximately 47 times higher than the macroscale UTG-Gra pair and 8 times higher than the PFA-Gra pair. Hence, the SSL-TENG exhibits a current density of 2.90 mA/m2 which is much higher than that of UTG-Gra (0.05 mA/m2) and PFA-Gra (0.20 mA/m2) pairs (Fig. 3c). While the open-circuit voltage of macro-TENGs is in the range from 2 V to 30 V (Supplementary Fig. 11), the SSL-TENG operates at a low voltage of 0.3 mV (Supplementary Fig. 12), yielding a peak power density of 6 W/m3 (Supplementary Fig. 13). Compared with previous high-performance devices with inorganic materials, the SSL-TENG also presents the highest ICECD (Fig. 3d), which is attributed to the high contact efficiency inherent to the SSL state. More importantly, the SSL-TENG demonstrates unprecedented cycling stability, maintaining consistent performance over 1.1×105 cycles without degradation (Fig. 3e). By contrast, the unannealed Gra MF slider with edge-surface contact is hindered by edge friction arising from the edge-pinning effect, exhibiting a limited lifespan of only about 104 cycles (Supplementary Fig. 14). The electrical output performance of SSL-TENG. a Working mechanism of sliding-mode TENG. b Charge density and c current density of macro-TENGs and SL-TENG. d Charge density comparison between this work and previously reported inorganic48,49 and organic50 materials; the mean value ± SD of SSL-TENG was obtained via 19 tests. e The current output over 1.1×105 cycles. The sliding distance and frequency were 20 μm and 2.1 Hz, respectively. The loading normal force was 100 μN. Source data are provided as a Source Data file. Charge transfer in SSL-TENG The charge density measured in TENGs is a net result of the dynamic equilibrium between charge generation and dissipation. Specifically, as the slider moves, charges are continuously generated at the interface via CE, while simultaneously being limited by dissipation pathways including air breakdown and dielectric breakdown. The final balanced charge density (σTENG) is governed by the minimum value among these competing processes, as expressed by: In this model, σCE represents the charge generation due to the CE. The terms σAB and σDB denote the discharge thresholds for air and dielectric breakdown, respectively. The interface of conventional macro-TENGs is characterized by significant topographical fluctuations. For example, the macroscopic graphite slider exhibits surface height variations of about 5 μm, while common polymers show fluctuations in the range of 3 μm (Supplementary Fig. 15). As a result, the high degree of surface unevenness and roughness inevitably creates air gaps at the sliding interface, resulting in air breakdown (Fig. 4a), which leads to the low charge density of macro-TENGs. Wang et al. demonstrated that operating devices in a vacuum condition to eliminate air can result in a 7-fold enhancement in charge density as compared to the ambient condition. Therefore, preventing the air breakdown is an effective way to boost the performance of TENGs. Furthermore, the surface charge distribution on macroscale surfaces often exhibits a “mosaic” pattern where localized positive and negative charges coexist40. For example, the surface potential of PI shows adjacent negative and positive regions (Fig. 4b). This bipolar distribution leads to significant charge compensation between adjacent regions of opposite polarity, effectively neutralizing a substantial portion of the generated charges. Consequently, the net charge density measured in macroscopic experiments is significantly lower than that due to the intrinsic material limit. Suppressing air breakdown and charge compensation via the SSL interface. a Schematic diagram of macroscale rough contact where interfacial air gaps allow for charge loss via air breakdown. b Surface potential map showing a mosaic charging pattern. c Schematic diagram of the SSL interface achieving atomic contact, which eliminates interfacial air gaps and suppresses air breakdown. d The homogeneous unipolar charging at the SSL interface, with the sliding region is dominated exclusively by negative charges. Scale bar, 4 μm. By contrast, the SSL-TENG operates in an atomically intimate contact regime that fundamentally bypasses the limitation imposed by air breakdown (Fig. 4c). The maintenance of a consistent, unipolar negative charge region (Fig. 4d) serves as definitive evidence of the suppression of air breakdown, effectively maximizing the charge output. Performance evaluation of SSL-TENG To evaluate the charge generation process and energy conversion process of SSL-TENG, the input mechanical force and output electricity were both characterized. More specifically, the LFM and tribometer were used to measure the lateral force as well as the displacement of SSL-TENG and macro-TENGs, respectively. Simultaneously, we recorded the charge transfer and open-circuit voltage to evaluate the electrical output of TENGs using cycles for maximized energy output (CMEO)25. The recorded F−x and V−Q plots are presented in Supplementary Fig. 16 and Fig. 17, respectively. Therefore, the relationship between input driving force and electrical output is investigated. As summarized in Fig. 5a, the κ of SSL-TENG is about 1.9×106, which is six orders of magnitude higher than that of conventional macro-TENGs. Furthermore, the η of macro-TENGs is in the range of 10−4−10−2, demonstrating that the dissipated energy via friction remains several orders of magnitude greater than the harvested electrical energy. By contrast, the η of SSL-TENG is about 3.11 %, which is 2×104 times higher than that of the UTG-Gra group and 500 times higher than that of the PFA-Gra/PTFE-Gra group, respectively. The significant efficiency enhancement is attributed to the synergistic effect of maximized electrical output and minimized frictional energy loss, both of which are enabled by the SSL interface. Furthermore, the energy density of SSL-TENG is about 1×103J/m3, which is comparable to polymer-based TENGs41. Performance evaluation of SSL-TENG. a Comparison of triboelectrification and energy conversion efficiencies between the SSL-TENG and conventional macro-TENGs (the error bars were obtained via 30 tests for each data point, representing SD value). b Performance benchmarks of cycle life and output retention, comparing SSL-TENG with previous in-contact sliding TENGs. c Assembly of a large-scale SSL-TENG array and its potential application in wearable electronics, intelligent robotics, and IoT-based sensor systems. Source data are provided as a Source Data file. Material wear severely limits the lifespan and performance of in-contact sliding TENGs. As shown in Fig. 5b, traditional rigid-rigid contact interfaces exhibit significantly lowered output retention (typically below 40%) under long-cycle operation42. While soft-contact strategies can mitigate wear to some extent, high retention was only achieved within a limited number of cycles43–46. Notably, even existing superlubric generators (COF < 0.01) barely maintain 100% performance for 10,000 cycles23,24. In contrast, the SSL-TENG leverages an incommensurate contact at the sliding interface to achieve lateral force cancellation, enabling an intrinsically wear-free state. This architecture facilitates 100% output retention across over 105 cycles, enabling maintenance-free power generation in extreme or inaccessible environments, such as high-vacuum and micro-electro-mechanical systems (MEMS) applications, without requiring lubrication or material-sacrifice strategies. In practical applications, ambient humidity is a critical factor that often degrades the stability of triboelectric performance, as moisture can increase the interfacial friction and dissipate surface charges. To evaluate the effect of humidity, we investigated the performance of SSL-TENG under high relative humidity (RH). Notably, even at 100% RH, when the lateral force increased, the COF consistently remained within the superlubricity regime (Supplementary Fig. 18a), demonstrating the stability of the SSL interface. Under high-load conditions, the lateral force decreased by 50% during cycling due to the refinement of the contact interface (Supplementary Fig. 18b). Although the electrical output exhibited around ~20% attenuation under 100% RH, the device maintained reliable output without failure (Supplementary Fig. 19). Furthermore, the SSL-TENG demonstrated good stability even after three weeks of storage (Supplementary Fig. 20). To fully harness the advantages of SSL-TENG, future efforts will focus on the assembly of SSL-TENG into large-scale integrated arrays. This scalable paradigm is envisioned to support low-power electronics, including distributed sensors, interactive robotics, and wearable devices (Fig. 5c), where the excellent durability of the SSL-TENG ensures maintenance-free operation in complex, real-world environments. However, several engineering challenges must be addressed before practical applications. Specifically, developing high-throughput mass transfer and precision assembly protocols for microscale Gra MF sliders is critical for scalable manufacturing. Furthermore, the implementation of synchronized driving strategies is essential to ensure uniform and efficient energy harvesting across large-area integrated systems. Discussion In conclusion, we have successfully developed a wear-free SSL-TENG with an exceptional operating lifespan by leveraging SSL technology to eliminate material wear and suppress air breakdown. The robustness of the SSL state between the Gra MF and SiO2 is verified by an ultralow COF (0.0034) and the absence of material debris after 2.4×105 cycles of operation. The atomically flat and intimate contact interface between graphite and SiO2 in SSL state enables a high ICECD of 0.47 mC/m2. Furthermore, the triboelectric efficiency and energy conversion efficiency of the SSL-TENG are enhanced by over 106 and 104 times, respectively, as compared to macroscopic UTG-Gra benchmarks, attributing to the reduced lateral input force and enhanced electrical output. Simultaneously, the SSL-TENG can work stably for more than 1.1×105 cycles without attenuation. This work provides an effective strategy for developing wear-free, high-electrical-output TENGs with ultra-long lifespans. Methods Fabrication of SSL-TENG slider The graphite micropillar array was fabricated using the method reported by Chen et al.47. First, depositing the Ti (20 nm) and Pt (200 nm) on the HOPG. Then, spinning coating photoresist on the metal surface and patterning it via laser direct writing to obtain the overlay array. Next, the metal layer was removed by ion beam etching, and the HOPG layer was removed by reactive ion etching. Finally, the graphite micropillar array was obtained by removing the photoresist. The processing map is presented in Supplementary Fig. 1. The self-retracting graphite slider was transferred using a tungsten probe: first, pressing one graphite micropillar and pushing out the microflake with half-width; then, lifting the probe; if the microflake can return to the initial position, it has a self-retracting property; Next, pressing the micropillar again and pushing out the microflake; last, transferring it to the SiO2 surface. The full process is presented in Supplementary Fig. 2. Fabrication of SSL-TENG Stator First, depositing Cr (5 nm) on the 4-inch glass substrate to prepare the electrode. Second, SiO2 (1 μm) was deposited using plasma-enhanced chemical vapor deposition (PECVD). Then, using chemical-mechanical polishing to obtain a flat surface with a thickness of 200 nm. Next, coating the photoresist on the glass wafer and developing the pad area. Then, the Cr/Au (30 nm + 100 nm) layer was deposited on the pad array. Finally, cleaning and dicing the wafer to obtain the device. The schematic diagram of fabrication is presented in Supplementary Fig. 4.To measure the electric signal of SSL-TENG, the Au wire bonding technique was adapted (Supplementary Fig. 10). Fabrication of macro-TENGs The substrate stator was fabricated using a printed circuit board (PCB) with two electrodes (10×10mm2). The gap between electrodes is 2 mm. The PFA tape (80 μm), FEP tape (80 μm), PI tape (50 μm), and PTFE tape (50 μm) were pasted on the PCB surface to prepare the macro-TENG stator. For the UTG-Gra pair, the Cu tape was pasted on the UTG (50 μm), and subsequently mounted onto a glass substrate. The macro-TENG slider is a rectangular HOPG with the size of 5×5×1mm3. Characterization The surface morphology of SSL-TENG was measured by AFM (Oxford Instruments, Cypher ES) in contact mode. The graphite slider surface was investigated using Raman spectroscopy. A focused ion beam-scanning electron microscope was used to obtain the cross-section image of SSL-TENG. Measurement The friction and electrical measurements of SSL-TENG were conducted in an ambient condition at a temperature of 25 °C and a relative humidity of 50%. The friction, surface morphology, and surface potential experiments were measured by AFM (Cypher ES). The Kelvin probe force microscopy (KPFM) mode with a conductive probe (HQ: NSC18/Pt, 2.8 N/m) was used to measure the surface potential of PI and the sliding region in SSL-TENG. To determine the charge polarity on the PI surface, the charged film was exposed to ambient conditions for partial dissipation, as the initial surface potential exceeded the 10 V measurement range of the KPFM system. This procedure ensured that the potential was lowered into a measurable window for accurate polarity characterization. Details for normal force and lateral force calibration can be found in Supplementary Note 1. The lateral force and displacement of macro-TENGs were measured by UMT TriboLab (Bruker) by setting the normal force as 50 mN, 100 mN, 150 mN, 200 mN, and 300 mN. The current and charge output of SSL-TENG was recorded by electrometer (Keysight B2980). The output voltage of the macro-TENG was determined using a load resistance of 500 GΩ, with the values derived according to Ohm’s law (V = IR). Considering the microsize of SSL-TENG and high impedance, the electrometer cannot directly measure the true VOC of SSL-TENG. We measured the capacitance (CM) of SSL-TENG using a parameter analyzer (Keithley 4200A-SCS) and calculated the VOC by Eq. 5 For the energy density calculation, Eq. 6 was used: where the sliding distance is 15 μm, width is 10 μm, device thickness (including electrode, SiO2, and Gra MF) is about 1 μm, thus the volume is 1.5 × 10-16 m3. The Eout of SSL-TENG was obtained from Supplementary Fig. 17. The charge density was calculated using Eq. 7: Here, A represents the contact area between Gra MF and SiO2. The average charge density was calculated from Fig. 3b. Supplementary information Supplementary Information Description of Additional Supplementary Files Supplementary Movie 1 Transparent Peer Review file Source data Source Data Acknowledgements We thank the technological support from the Materials Characterization and Preparation Facility (G.Z.) (MCPF(G.Z.)) of HKUST-GZ. We thank Mr. Xiaofei Pang for assistance with the experiments. Author contributions C.C., J.N., and Y.Z. conceived the idea. C.C. and J.N. designed the experiments. C.C. carried out the measurement of devices, analyzed the results, and discussed with J.N. and Y.Z. J.N. prepared Gra MF samples. Y.L., X.X., and X.H. helped with device fabrication. C.C. wrote the paper. C.C., J.N., Q.Z., and Y.Z. discussed and edited the manuscript. J.N. and Y.Z. supervised the project. All authors reviewed the manuscript. Peer review Peer review information Nature Communications thanks Ishara Dharmasena, Huidrom Hemojit Singh and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available. Funding J.N. discloses support for the research of this work from the National Natural Science Foundation of China [grant number 12204321], the Guangdong Basic and Applied Basic Research Foundation [grant number 2022B1515120040], and the Shenzhen Science and Technology Program [grant numbers KJZD20240903102715020 and ZDSYS20230626091701002]. X.H. discloses support for the research of this work from the National Natural Science Foundation of China [grant number 62504142] and the Shenzhen Science and Technology Program [grant number KQTD20240729102211015, ZDCY20250901112203004]. Y.Z. discloses support for the research of this work from the National Natural Science Foundation of China [grant number No. 52275560], Guangdong Natural Science Funds for Distinguished Young Scholar [grant number 2023B1515020074], and Guangzhou Municipal Science and Technology Project [grant number 2025A04J4597]. Data availability Source data are provided with this paper. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Chaojie Chen, Jinhui Nie. Contributor Information Jinhui Nie, Email: niejh@Tsinghua-sz.org. Yunlong Zi, Email: ylzi@hkust-gz.edu.cn. Supplementary information The online version contains supplementary material available at 10.1038/s41467-026-74450-w. References Associated Data Supplementary Materials Supplementary Information Description of Additional Supplementary Files Supplementary Movie 1 Transparent Peer Review file Source Data Data Availability Statement Source data are provided with this paper.