High-performance triboelectric nanogenerators based on TPU/mica nanofiber with enhanced tribo-positivity

Kelvin probe force microscopy (KPFM) measurements of single pure TPU and TPU/mica nanofibers showed mica apparently enhanced the positive electrostatic surface potential (ESP) of TPU. The mean peak ESP of pure TPU nanofiber was about 194 mV, while it was increased to 218 mV on the regions of TPU/mica nanofiber without mica nanosheets aggregation and 305 mV on the regions where mica nanosheets aggregated. The power density of the TENG with TPU/mica nanofiber reached 1458 mW/m 2 , exhibiting a 16-fold enhancement compared with the one based on pure TPU nanofiber. A thin and flexible TENG was fabricated and conformally worn on a wrist and palm for body motion monition and object gripping sorting. This work proposed high-performance triboelectric nanogenerators based on TPU/mica nanofibers and a general approach to effectively utilize rigid and fragile materials with excellent triboelectric property for making triboelectric layers.


Introduction
Renewable energy development is imminent and of great significance because of the rapid depletion of fossil fuels, severe environmental pollution and the objective of carbon neutrality.In addition, the widely distributed internet of things (IoT) with massive sensor nodes is highly expected to possess self-powered properties to mitigate the enormous energy consumption, huge cost and pollution from the replacement of depleted batteries [1,2].Natured with unique merits, including wide material choice, high efficiency, easy fabrication and various working modes, the triboelectric nanogenerator (TENG) has demonstrated great potential in wasted mechanical energy harvesting from the ambient environment [3][4][5].The killer application of the TENG in random and low-frequency energy harvesting has enabled its broad prospects from self-powered sensors [6][7][8], micronano energy sources [9][10][11], high voltage sources [12,13] to large-scale blue energy harvesting [14,15].In all application scenarios, a high output of the TNEG is always expected, which largely depends on the triboelectric polarity of triboelectric materials [16,17].Normally, the greater the distance of two triboelectric materials located in the triboelectric series, the higher the triboelectric output can be generated.
Many materials, especially inorganic non-metallic materials, have been tested to possess strong triboelectric positivity or negativity [18,19], but they are rarely practically utilized in fabricating TENGs.A critical important reason is that they are normally bulky, rigid and fragile, which largely hinder the durability of TENGs and potential applications in flexible TENGs.Nevertheless, if processed into micro/nano scales, these materials can be superior fillers to modify the triboelectric polarity of traditional polymer-based triboelectric materials [19,20].It has been successfully demonstrated that the triboelectric negativity of Polyvinylidene fluoride (PVDF) can be further enhanced by incorporating 2D smectite clay nanosheets, which feature with intrinsic strong negativity but are extremely fragile in macroscale [21].Zou et al. quantified the triboelectric series of inorganic non-metallic materials [22], among which mica was ranked the highest triboelectric surface charge density (TECD) and strongest positivity.Despite the strong positivity, mica has never been practically used as triboelectric material in TENGs due to its weak-bonded layer structure as described as basal cleavage [23,24] which makes mica too fragile to survive from continuous contact-separation or sliding motion.On the other hand, the weak-bonded layer structure also indicates that mica can be easily split and exfoliated into 2D single-layer or few-layer nanosheets, which have a large surface-to-volume area ratio and flexibility.Similar to other 2D materials, such as MXene [25,26], graphene [27] and MoS 2 [28] widely applied in TENGs to boost the output performance, 2D mica nanosheets can contribute to high effective contact area and numerous charge traps.Specifically, the metallic MXene nanosheets, due to its high electronegativity [25,29] and formation of microcapacitors [26] inside polymers, have become a hot material for TENGs.A small amount of MXene mixed into the triboelectric layers or using MXene as electrode [30] can largely enhance the output performance of the TENG.
Herein, rigid and fragile mica (KAl 2 (Si 3 Al)O 10 (OH) 2 ) bulks were split and exfoliated into 2D mica nanosheets and electrospun into thermoplastic polyurethane (TPU) nanofibers that are flexible and stretchable for high-performance TENGs.Paired with PVDF/MXene nanofibers, with the concentration of mica nanosheets increased to 7.5 wt%, the transferred charge of the TENG increased from 38.6 nC to 82.4 nC and then decreased with the further increase of mica concentration.Compared to the pristine nanofibers, the dielectric permittivity and hydrophobicity of TPU/mica nanofibers showed no considerable change.To understand the output enhancement of the TENG, the relative triboelectric polarity between TPU and TPU/mica was compared and the results showed that the triboelectric positivity of TPU/mica was enhanced by introducing mica nanosheets.In addition, Kelvin probe force microscopy (KPFM) measurements of single TPU and TPU/mica nanofibers showed that mica apparently enhanced the positive electrostatic surface potential (ESP) of TPU.The mean peak ESP of pure TPU nanofiber was measured at 194 mV, while it was increased to 218 mV on the regions of TPU/mica nanofiber without mica nanosheets aggregation and 305 mV on the regions where mica nanosheets aggregated.
With enhanced triboelectric positivity, the TPU/mica nanofibers were used to make high-performance energy harvesting TENGs and wearable self-powered sensors.For the energy harvesting TENG, the power density was enhanced to 1458 mW/m 2 from 88 mW/m 2 of the one with pure TPU.In addition, a flexible and thin sensor based on TPU/mica nanofiber TENG could be conformally worn on a wrist and hand palm for monitoring the wrist flexion and plam grasping, which have huge potential in body movement monitoring and intelligent object sorting.This work proposed TPU/mica nanofibers with enhanced tribo-positivity for high-performance TENGs and a general approach to effectively utilize rigid and fragile materials with excellent triboelectric property.

Results and discussion
Fig. 1a illustrates the schematic structure of the fabricated TENG with TPU/mica nanofibers serving as positive triboelectric material and PVDF nanofiber as negative triboelectric material.2D mica nanosheets were mixed into TPU and electrospun into nanofibers that were collected onto aluminum foil that was directly used as backside electrode.To improve the triboelectric performance of PVDF, MXene nanosheets of appropriate content were also mixed and electrospun into PVDF nanofibers [31].Figs.1b and 1c show the scanning electron microscopy (SEM) images of the electrospun TPU/mica and PVDF/MXene nanofibers, respectively, from which one can see that the mica nanosheets and MXene nanosheets were well mixed into nanofibers.The energy dispersive spectroscopy (EDS) spectra and elemental mapping results of the TPU/mica nanofibers (Fig. S1) and PVDF/MXene (Fig. S2) also indicate that 2D nanosheets were well mixed and uniformly dispersed into nanofibers.The electrospun TPU/mica nanofiber membrane demonstrates good flexibility (Fig. 1d) and stretchability (Fig. 1e).As seen in Fig. S3, a maximum stretch of almost 250 % and tensile strength of 15 MPa were achieved by the TPU/mica nanofiber with a mica concentration of 7.5 wt%.To ensure that the 2D mica nanosheets can be uniformly dispersed into TPU nanofibers during electrospinning, the multilayer mica bulks were first exfoliated into single-layer or few-layers nanosheets.Fig. S4 shows an SEM image of a cluster of mica nanosheets, indicating a good crumb and cleavage from mica bulks.Fig. S5 presents the atomic force microscopy (AFM) image of few-layer mica nanosheets, revealing that mica nanosheets have an ultrathin thickness of 2 nm.Fig. 1f shows the transmission electron microscopy (TEM) of single-layer or few-layer mica nanosheets, which shows that mica nanosheets have a lateral size of approximately 500 nm -1 µm.In addition, the high-resolution TEM (HRTEM) image in Fig. 1g reveals the typical pseudohexagonal crystal structure of mica, demonstrating that a high crystalline and structural order were maintained after etching and exfoliation.Fig. 1h shows the TEM image of a single mica/TPU nanofiber, which clearly indicates that the mica nanosheets were encased in the nanofiber, as further confirmed by TEM EDS elemental mapping (Fig. 1i).
The working principle of the TENG is illustrated in Fig. 2a.After several cycles of contact-separation motion, the TPU/mica and PVDF/ MXene nanofibers are oppositely charged due to the effect of contact electrification.The difference in the ability to lose or attract electrons results in positive charges on the surface of TPU/mica nanofibers and equal amounts of negative charges on the surface of PVDF/MXene nanofibers.When the two charged surfaces are fully in contact (state I), there is no charge flow between the two backside electrodes due to electrostatic equilibrium.A varying electrostatic field is formed when the surfaces begin to separate, which drives electrons flowing from the bottom electrode to the top electrode (state II), as a result of electrostatic induction effect, until the two surfaces are fully separated (state III).Similarly, electrons would flow back from the top electrode to the bottom electrode when the two surfaces are contacting again (state IV).The electrical potential distributions of the three different states (contacted, separating and separated) were simulated using finite element analysis with COMSOL Multiphysics, as shown in Fig. 2b.
Mica is an extremely strong triboelectric positive material and the mixture of mica nanosheets into TPU, whose triboelectric positivity is weaker than pure mica, may enhance TPU's triboelectric positivity.Therefore, optimization of the concentration of mica in the TPU/mica nanofiber was conducted to investigate the output enhancement of the TENG and understand the enhancement mechanism.Before that, the concentration of MXene nanosheets in PVDF was optimized and determined to be 4 wt% based on its optimal output (Fig. S6).It was attributed to that microcapacitors [26,31] consisting conductive MXene electrodes and PVDF dielectric layer were formed inside the PVDF, resulting in a significant increase in the dielectric constant of the PVDF/MXene nanofibers (Fig. S7) and accordingly an enhancement in the output of TENG.However, when the concentration of MXene increased to a high value, MXene nanosheets began to aggregate and the conductivity of the PVDF/MXene nanofibers increased, resulting in a decrease in the output performance of TENG.Mica nanosheets were mixed into TPU at the concentrations of 0, 2.5, 5, 7.5, 10 and 12.5 wt% and electrospun into nanofibers.Fig. 2c-e show the open-circuit voltage, short-circuit current and short-circuit transferred charge of the TENG with different mica concentrations in the TPU/mica nanofiber, respectively, under 20 N and 0.5 Hz contact-separation.As can be seen, the output performance of the TENG first continued to increase to the maximum (224 V, 3.72 μA, and 82.5 nC) at a mica concentration of 7.5 wt% and then gradually decreased with increasing the concentration of mica nanosheets.To understand the output change of the TENG over the mica concentration, the dielectric constant and dielectric dissipation factor of the TPU/mica nanofiber membranes were first investigated, as shown in Figs.2f and 2g.All TPU/mica nanofiber membranes showed a lower dielectric constant than the pure TPU nanofiber, but there is no obvious trend in the dielectric permittivity of the TPU/mica nanofiber membranes with the change in mica concentration, excluding the reason of the change in the dielectric property for the output change.In view of the output of TENGs can be easily affected by humidity and the mica is typically hydrophilic, which may impose an adverse effect on the humidity-resistant ability of nanofibers.Therefore, the hydrophobicity of the nanofiber membranes was also verified, and the water contact angle (WCA) of the TPU/mica nanofiber membranes remained almost stable (Fig. 2h).Then, the surface morphology of the TPU/mica nanofibers was examined and it was found that the surface roughness and porosity of the mica/TPU nanofiber membranes decreased when the concentration of mica nanosheets increased, especially after 7.5 wt% (Fig. 3).The reason is that when the mica concentration was increased, the weight of the solution was also increased, so the nanofiber had a larger impact force flying to the collection roller during electrospinning.Therefore, the nanofibers tightly stacked together, which reduced the porosity of the mica/TPU nanofiber membrane.As a result, the increase in the output of the TENG from adding mica can be attributed to the strong triboelectric positivity of mica nanosheets, while the decrease in the output when the concentration exceeded 7.5 wt% is because of the decrease in surface roughness and porosity of the nanofiber membrane.
To verify that the strong triboelectric positivity of mica nanosheets can contribute to the output enhancement of the TENG, the relative triboelectric polarity of the TPU/mica nanofibers was systematically compared to that of the pure TPU nanofibers.The positions of mica, TPU/mica and TPU in the triboelectric series were determined by testing their relative triboelectric polarity according to a previously reported method [16,22].As illustrated in Fig. S8, a contact-separation TENG, in which the TPU/mica (the tested material) and another counterpart material act as two triboelectric layers, was fabricated for this test under 5 N and 0.5 Hz.Fig. 4a shows the open-circuit voltage of TENGs in which TPU, TPU/mica, mica and nylon were paired with each other.First, the results proved that mica is strongly triboelectric positive since it was always positively charged when in contact with typical triboelectric positive materials, including TPU and nylon.There was almost no output between the TPU-TPU pair because of the identical electron affinity for the same material.However, after incorporating mica nanosheets, the triboelectric polarity of TPU/mica was significantly enhanced compared to that of the pure TPU as indicated by the obvious positive output between the TPU/mica-TPU pair.In addition, when TPU/mica contacted with mica and nylon, a significant lower output could be observed, compared with the pure TPU, confirming an enhanced triboelectric positivity of TPU/mica.Fig. 4b shows a summary of the relative triboelectric positions among mica, nylon, pure TPU and TPU/mica.
To further prove the above conclusions, KPFM measurements of single TPU/mica and pure TPU nanofibers were conducted, from which the electrostatic surface potential (ESP) distribution of nanofibers could be mapped to identify the effect of mica nanosheets.TPU/mica and TPU nanofibers were both first contact-electrified by the PVDF/MXene nanofibers before KPFM measurements.Figs.4c and 4d show the height and ESP distribution of a single TPU/mica nanofiber from the KPFM measurement, respectively.The same section of the nanofiber was also analyzed by EDS elemental mapping to verify the distribution of mica nanosheets along the nanofiber (Fig. 4f).The regional aggregation of silicon clearly proves the existence of mica inside the nanofiber.For a more intuitive comparison, cross-section profiles of the height and ESP of the two locations, among which one contains mica while the other does not (as marked as profiles 1 and 2, respectively), are summarized in Fig. 4e.For the location 1, region A demonstrates a strong aggregation of mica, while region B hardly shows the existence of mica.Despite the fact that the height of region A is slightly lower than that of region B, the positive ESP sharply increased in region A and then suddenly decreased in region B, which shows a ESP value similar to that of location 2 where there is no obvious aggregation of mica.The KPFM height and ESP distribution of a single pure TPU nanofiber are demonstrated in Figs.4g  and 4h, respectively.Fig. 4i shows the height and ESP profiles of the marked location.A comparison of the mean peak ESP value of 5 different pure TPU nanofibers and regions of TPU/mica without mica aggregation (TPU/mica-B) and with mica aggregation (TPU/mica-A) is shown in Fig. 4j.It is worth to note that the peak ESP was calculated by clearing the mean potential of the grounded substrate as zero.The mean ESP of the pure TPU nanofiber was thus measured at 194 mV; in comparison, the ESP of the TPU/mica nanofiber increased to 218 mV for the regions without clear mica aggregation and 305 mV for the regions where mica nanosheets were obviously aggregated.The above results clearly reveal that mica nanosheets can significantly increase the positive triboelectric surface potential of TPU/mica nanofibers due to their strong triboelectric positivity.
The enhanced triboelectric polarity of the flexible and stretchable TPU/mica nanofibers can not only have potential in energy harvesting TENGs but also flexible self-powered sensors.A typical contactseparation TENG for energy harvesting based on the TPU/mica and PVDF/MXene nanofibers was first fabricated (Fig. 5a) with a common structure as-reported previously [32,33], and its electrical characteristics were systematically studied.First, the output response under different working frequencies (0.25, 0.5, 1 and 2 Hz) was tested with the fixed load of 20 N (Fig. 5b).Theoretically, the open-circuit voltage of the TENG should remain stable with the change in working frequency, while that of the TENG with TPU/mica nanofiber demonstrated a slight increase.The reason could probably be the more severe charge dissipation at lower frequencies, and surface charges could be quickly replenished at higher frequencies.The current showed a normal linear increase with increasing working frequency, as expected.As shown in Fig. 5c and Fig. S9, when the external working load increased, the voltage, current and charge all gradually increased due to the more intimate contact between the two nanofiber membranes.The output power densities of the TENGs and with TPU/mica and pure TPU were measured through impedance matching for comparison.Fig. 5d demonstrates the voltage and current of the two TENGs under different external resistors.The voltage increased and then saturated as the external resistor increased, while the current first remained almost stable and then decreased.As a result, the calculated maximum instantaneous power density of the TENG with TPU/mica nanofibers is 1450 mW/m 2 , while that of the TENG with pure TPU nanofibers is 88 mW/m 2 (Fig. 5e).It is clear that, compared with the pure TPU nanofibers, the enhanced tribo-positivity of TPU/mica enabled a 16-fold enhancement in the instantaneous power density of the TENG.With the significantly enhanced output performance, the TENG with TPU/mica nanofibers also demonstrated a faster capacitor charging speed to store harvested energy, as shown in Fig. 5f.Fig. 5g shows the output of the TENG under 10,000 working cycles, indicating its excellent long-term stability for ambient energy harvesting to power daily electronics.As depicted in Fig. 5h-j, not only could 100 green LEDs (Video S1) be instantaneously lighted by the TENG, but electronics such as watches (Video S2) and hydrometers (Video S3) could also be easily powered by the energy stored in a capacitor charged by the TENG.
Supplementary material related to this article can be found online at doi:10.1016/j.nanoen.2023.108629.
In view of the excellent flexibility and stretchability, the TPU/mica nanofiber with enhanced tribo-positivity was expected to have huge potential in wearable TENGs for self-powered body motion monitoring and robotic sensing.Fig. 6a shows a photo of the as-fabricated flexible TENG consisting of TPU/mica and PVDF/MXene nanofibers as triboelectric layers on 50 µm thick PET substrates.Double sided foam tapes (thickness 0.5 mm) were adopted as spacers at the two ends of TENG ensure effective contact and separation between the two nanofiber layers.Figs.6b and 6c show the photos of the flexible TENGs being conformally worn on the wrist and hand palm.As illustrated in Fig. 6d, when the TENG was worn on the wrist, a flexion of the wrist can result in a bending of the TENG and corresponding contact between the two triboelectric layers.With the increase of wrist flexion degree, the effective contact area increases, delivering a higher output.shows the result of wrist flexion monitoring using the flexible TENG, showing that the flexion degree can be clearly recognized through the output amplitude of the TENG (Video S4).The flexion degree responsive output of the TENG suggests huge potential in body motion monitoring and detection for applications in virtual reality and auxiliary command control for disabled population.Worn in the center of hand palm, the bending of the TENG increases with the gathering degree of fingers, which accordingly related to the size of the object grasped.As schematically illustrated in Fig. 6f, grasping a smaller object can lead to a larger bending of the TENG and thus higher output.Fig. 6g demonstrates the change of the output of the TENG when objects of different sizes were grasped by hand (Video S5), which indicates that the size of the grasped object can be identified by the output of the TENG, implying great prospects in intelligent robotic sorting.
Supplementary material related to this article can be found online at doi:10.1016/j.nanoen.2023.108629.

Conclusion
In conclusion, we proposed high-performance TENGs based on TPU/ mica nanofibers with enhanced tribo-positivity for energy harvesting and self-powered wearable sensing.Due to the strong tribo-positivity of mica nanosheets, the output performance of the TENG consisting of TPU/mica and PVDF/MXene nanofibers demonstrated a significant enhancement with the transferred charges increased from 38.6 nC to 82.4 nC at the mica concentration of 7.5 wt%.KPFM measurements of single nanofibers show that mica apparently enhanced the positive ESP of TPU.The mean ESP of pure TPU nanofiber was measured at 194 mV, while it was increased to 218 mV on the regions of TPU/mica nanofiber without mica nanosheets aggregation and 305 mV on the regions where mica nanosheets aggregated.The power density of the TENG with TPU/ mica nanofiber reached 1458 mW/m 2 , exhibiting a 16-fold enhancement compared with the TENG with pure TPU nanofibers.Compared with reported TENGs adopted 2D or nanomaterials as fillers in triboelectic materials, the mica nanosheets show a remarkable enhancement in ESP (Table S1) and power density (Table S2).Moreover, the TPU/  c and d).KPFM measurement of (g) height and (h) surface potential distribution of a single pure TPU nanofiber.(i) Extracted height and surface potential profiles of the location as marked in (g) and (h).(j) Comparison of the mean peak ESP of pure TPU and regions of TPU/mica without mica aggregation (TPU/mica-B) and with mica aggregation (TPU/mica-A).
mica nanofiber showed a great potential in flexible and wearable selfpowered sensors.The flexible TENG could be conformally worn on the wrist and palm for body motion monition and object gripping sorting, which have huge prospects in virtual reality and auxiliary command control for disabled population, and intelligent robotic sorting.acid (HCL, 37 %), lithium chloride (LiCl), N,N-dimethylformamide (DMF, 99.9 %), acetone (99.9 %), tetrahydrofuran (THF, 99.9 %) and thermoplastic polyurethane (TPU) were all purchased from Sigma--Aldrich.Raw natural mica plates were purchased from Guangzhou Beilong Electronics Co., Ltd.

Preparation of 2D MXene and mica nanosheets
2D MXene (Ti 3 C 2 T x ) nanosheets (SEM image shown in Fig. S10) synthesized by selective etching of the Al atomic layer from the MAX phase.In detail, 8 g of LiF was added to 100 ml of HCl (9 M).Five grams of MAX powder was slowly added to the LiF/HCl mixture while magnetically stirring for 30 min.Then, the mixture was stirred at 350 rpm for 24 h at 40 • C for reaction.After etching, the mixture was repeatedly washed with Milli-Q water and centrifuged at 4500 rpm for 10 min.The sediment after centrifugation was ultrasonically redispersed in 200 ml Milli-Q water in an ice bath with Ar flow for 2 h.Then, the mixture was centrifuged at 3500 rpm for 1 h, and the turbid upper black liquid was collected, so the nonetched MAX phase and large multilayer MXene bulks could be eliminated.The supernatant was washed using Milli-Q water several times until its pH value reached 5.5, after which the MXene flakes were collected by centrifugation at 12,000 rpm, vacuum dried at room temperature overnight and stored in a vacuum desiccator for further use.The X-ray diffraction XRD spectrum (Fig. S11) of final obtained MXene nanoflakes shows a successful preparation of MXene.
2D mica nanosheets were obtained by salt-assisted ball milling exfoliation from a raw natural mica material (muscovite, KAl 2 (Si 3 Al) O 10 (OH) 2 ), the XRD spectrum of which is shown in Fig. S12.Typically, 2 g of mica plates (thickness 200 µm) were first cut into small pieces with a few millimeters, which were then put in a milling container and ball milled for 1 h with a rotation speed of 500 rpm on a ball milling machine.Then, 3.4 g lithium chloride and 2 ml Milli-Q water were added into the milled container, followed by second-round ball milling for 4 h.After that, the mixture was transferred to a beaker with 50 ml Milli-Q water and vigorously magnetically stirred for 4 h, followed by ultrasonication for 2 h.Finally, the mixture was centrifuged at 500 rpm to remove the large flakes and multilayer bulks.2D mica nanosheets with a single layer and few layers in the supernatant were finally obtained by centrifuging the dispersion at 4500 rpm and dried for further use.

Electrospin of MXene/PVDF and mica/TPU nanofibers
Dried MXene flakes were first ultrasonically dispersed in Milli-Q water to obtain well-distributed single-layer and few-layer 2D MXene nanosheets.Then, the solution was centrifuged at 10,000 rpm for 10 min, and the supernatant was discarded.The remaining sediment was further ultrasonically dispersed in the DMF:acetone mixed solution (weight ratio of 3:2).Centrifugation, solvent replacement and ultrasonication were repeated 3 times to ensure that the water was fully replaced by DMF:acetone solution.The final obtained sediment was ultrasonically dispersed in DMF:acetone solution with a weight-tovolume ratio of 10 mg/ml.To prepare the solution for electrospinning, 0, 0.75, 1.5, 2.25, 3 and 3.75 ml of the above MXene solution was added into 4.77, 4.02, 3.22, 2.52, 1.77 and 1.02 ml DMF:acetone mixtures, respectively, followed by adding 750, 742.5, 735, 727.5, 720 and 712.5 mg of PVDF (Kynar HSV 900), respectively.After that, the mixtures were stirred at 600 rpm for 2 h at 60 ℃, so 15 wt% (whole solute to solution) MXene/PVDF solutions in which the weight concentration of MXene in the whole solute was 0, 1, 2, 3, 4, and 5 wt%, respectively, were prepared for electrospinning.For electrospinning, the solution was fed at a constant flow rate of 1 ml/h using a syringe pump with a 10 ml syringe and 18 G needle.The voltage applied was 16 kV, and the nanofibers were collected on aluminum foil, which was attached onto a rotation roller with a speed of 500 rpm.The distance between the needle tip and the bottom of the rollers was 15 cm.Dried mica flakes (0, 12.5, 25, 37.5, 50, and 62.5 mg) were first directly ultrasonically dispersed in 4.5 g DMF:THF solution (weight ratio of 1:2).Then, 500, 487.5, 475, 462.5, 450, and 437.5 mg TPU were added into the above prepared mica solution and stirred at 600 rpm for 2 h at 70 ℃.Therefore, 10 wt% (whole solute to solution) mica/TPU solutions in which the weight concentration of mica in the whole solute was 0, 2.5, 5, 7.5, 10, and 12.5 wt% were prepared for electrospinning.For electrospinning, the solution was fed at a constant flow rate of 1 ml/ h using a syringe pump by a 10 ml syringe and 18 G needle.The voltage applied was 12 kV, and the nanofibers were collected on aluminum foil, which was attached onto a rotation roller with a speed of 500 rpm.The distance between the needle tip and the bottom of the rollers was 15 cm.

Fabrication of TENGs
After electrospinning for 3 h, the aluminum foil with nanofibers was peeled off from the rotation roller.Then, PVDF/MXene and TPU/mica nanofiber membranes were cut into pieces (4 ×4 cm 2 ) with aluminum foil remaining as the backside electrode.After that, nanofiber membranes were attached onto acrylic substrates using double-sided foam tape with the nanofiber sides facing outside.The two acrylic sheets were packaged by a Kapton film (thickness 300 µm) to form a contactseparation TENG, in which the nanofiber membranes faced each other with a gap of 5 mm.For the flexible and wearable TENG, nanofiber membranes (2.5 ×3 cm 2 ) with aluminum foil electrodes were attached onto flexible PET substrates (thickness: 50 µm).Double-sided foam tape (thickness 0.5 mm, width 2 mm) strips were adopted as spacers at the two ends of the device to ensure effective contact and separation.

Characterizations
X-ray diffraction (XRD) measurements were performed on a D8 Advance Bruker diffractometer with Cu Kα radiation (λ = 1.5418Å), employing 1 0.25 • divergent silt and a 0.125 • anti-scattering slit.XRD spectra were recorded in the 2θ range from 5 • to 80 • with a step of 0.02 • and a counting time of 2 s per step.Scanning electron microscopy (SEM) was performed on Lyra, Tescan.Transmission electron microscopy (TEM) was performed on an FEI Tecnai T20 operating at 200 keV.The dielectric constants of the nanofiber membranes were calculated by measuring the capacitance of a parallel plate capacitor, in which a circular nanofiber membrane (diameter 12 mm) was sandwiched between two electrodes, using a Keysight LCR meter.The dielectric dissipation factor was directly read from the LCR meter.The water contact angles of the nanofiber membranes were measured using an optical contact angle measurement instrument (Dataphysics 15EC).Kelvin probe force microscopy (KPFM) measurements were conducted on a Bruker Multimode 8 AFM system under ambient conditions.A Pt-coated silicon tip (cantilever resonance frequency ~75 kHz; spring constant 2.8 N/m) was applied in lift mode with a lift height of 40 nm to conduct the KPFM.Nanofibers were directly electrospun on aluminum foil, which was grounded during KPFM measurement.The electrical output of the MM-TENG was measured using a Keithley electrometer (model 6514).The contact-separation process of the MM-TENG was controlled with a contact-separation distance of 5 mm by a universal testing machine (MTS 810 system) with adjustable applied force and frequency.

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

Fig. 1 .
Fig. 1.(a) Schematic illustration of the structure of the TENG consisting of TPU/mica nanofibers and PVDF/MXene nanofibers.(b) SEM image of TPU/mica nanofibers.(c) SEM image of PVDF/MXene nanofibers.(d) Photo of the electrospun flexible TPU/mica nanofiber membrane.(e) Demonstration of the strechability of the TPU/mica nanofiber.(f) TEM image of single-layer and few-layer 2D mica nanosheets.(g) HRTEM image of a single-layer 2D mica nanosheet.TEM image (h) and EDS elemental mapping (i) of a single TPU/mica nanofiber.

Fig. 2 .
Fig. 2. (a) The working principle of the TENG based on contact electrification and electrostatic induction.(b) Electrical potential distribution simulation during contact and separation.(c) Open-circuit voltage, (d) short-circuit current and (e) short-circuit transferred charge of the TENG with different mica concentrations in the TPU/mica nanofiber.(f) Dielectric constant and (g) dissipation factor of the TPU/mica nanofiber membranes with different mica concentrations.(h) Water contact angles of the TPU/mica nanofiber membranes with different mica concentrations. Fig.6e

Fig. 4 .
Fig. 4. (a) Relative triboelectric polarity test result of TPU/mica and pure TPU nanofibers with typical triboelectric materials.(b) Positions of TPU/mica and TPU in the simplified triboelectric series.KPFM measurement of (c) height and (d) surface potential distribution of a single TPU/mica nanofiber.(e) Extracted height and surface potential profiles of two locations as marked in (c) and (d).(f) EDS elemental mapping of the same single TPU/mica nanofiber shown in (c and d).KPFM measurement of (g) height and (h) surface potential distribution of a single pure TPU nanofiber.(i) Extracted height and surface potential profiles of the location as marked in (g) and (h).(j) Comparison of the mean peak ESP of pure TPU and regions of TPU/mica without mica aggregation (TPU/mica-B) and with mica aggregation (TPU/mica-A).

Fig. 5 .
Fig. 5. (a) Photo of the as-fabricated TENG for energy harvesting.(b) Output voltage and current of the TENG under different working frequencies.(c) Output voltage of the TENG under different working loads.Output voltage and current (d) and instantaneous output power density (e) of the TENGs with TPU/mica and pure TPU nanofibers under different external resistors.(f) Capacitor charging ability comparison between the TENGs with TPU/mica and pure TPU nanofibers by charging capacitors of 22 μF (f) (load: 20 N; frequency: 2 Hz).(g) Long-term stability test of the TENG under 10000 working cycles.(h) Instantaneously powering 100 LEDs using the TENG.(load: 20 N; frequency: 2 Hz) (i) Powering an electronic watch using the TENG.(j) Powering an electronic hygrometer using the TENG.(load: 20 N; frequency: 2 Hz).

Fig. 6 .
Fig. 6.(a) Photo of the as-fabricated flexible TENG for wearable sensing.(b) Photo of the flexible TENG being conformally worn on the wrist.(c) Photo of the flexible TENG being conformally worn on the palm.(d) Schematic illustration of the wrist flexion monitoring using the flexible TENG.(e) The output of the TENG under different wrist flexion degrees.(f) Schematic illustration of the grasped object size detection using the flexible TENG.(g) The output of the TENG under different object size grasped.