Ionic and adhesive triboelectric elastomer for normalizing charge polarity and density of contact electrification with general material

Abstract

Contact electrification (CE) is widely applied in microscale energy harvesting and active sensing. However, the CE results vary in charge density and polarity between different materials, while the preexisted charges on the surface make the CE process more unpredictable. Here, an adhesive triboelectric elastomer (ATE) is synthesized by combining the ionic liquids (ILs) and pressure-sensitive adhesive (PSA) to generate normalized charge polarity and density by CE. The ILs neutralize the preexisted charges on the targeted surface. Then, the adhesion groups of PSA bond to targeted surface and trigger the heterolytic cleavage, leaving negatively charged fragments on the targeted surface. This ATE is capable of generating stable charge density (−50 to −70 microcoulombs per square meter) on almost all the commonly used triboelectric materials, irrespective of the preexisted charges on the surface. The stable and normalized CE effect benefits all the triboelectric sensors and enables special functions, and this ATE design also deepens the understanding of CE and interface science.

A triboelectric elastomer can generate stable charge density with fixed polarity on different surfaces.

INTRODUCTION

Contact electrification (CE) is a complex physical process that occurs widely in nature, while the study of CE involves the scientific backgrounds from multiple fields, including interface science, tribology, and catalysis (1–5). Because the vibration of interface molecules at the micro-nano scale can induce charge transfer (6), the adsorption (7), binding (8), and reaction processes (9, 10) occurring at the interface are also directly affected by the CE process. The nature of the “static charge” produced in CE has been intensively studied for many years, and for now, electrons, ions, and material fragments are considered as the major charge carriers during CE. Correspondingly, mechanisms of electron transfer (11), ion adsorption (12), and material transfer (13, 14) have been proposed to explain the generation of charges in CE. Interaction between different charge transfer mechanisms leads to the regional charge density in the nanoscale (15, 16) and also determines the final unified CE result in macroscopic scale. However, further investigations of CE effect are still intensively conducted by many researchers, as previously unknown phenomena related to CE continue to emerge. For instance, the variation of electrification polarity can be triggered by the number of contacts (17), the influence of water structural disorder to the charge generation processes at oil-water interface has been clarified by Raman spectroscopy method (18), and so on. Hence, the CE effect is modulated by the collaboration and the competition of multiple physical mechanisms, while a deeper and completed theoretical picture of this physical effect is highly necessary.

On the other hand, CE has shown substantial application value in many fields, such as electrostatic precipitator (19), electrostatic printers (20), and layer-by-layer electrostatic self-assembly (21), where they have been fully developed. Nowadays, the mainstream application of CE is related to micro–energy harvesting and active sensors, where the kinetic energy can be converted into electrical energy output and the generated electrostatic field can calibrate the force and motion parameters (22–24). In this case, the charge density and the stability of CE is main parameters for promoting its application prospective. However, CE exhibits intrinsic uncertainty, as the charge density and polarity generated by contacting with different materials often vary obviously and may even undergoes polarity reversal, increasing the difficulty of device design and seriously hindering the applications of CE. To deal with the problem of uncertainty, triboelectric sequences have been proposed and summarized to define the charging polarity and guide the selection of materials (25, 26). However, the sequences do not truly resolve the problem, as contradictions between sequences exist (27), and electrification results of two closely adjacent materials in the sequence may also exhibit repeatability issues (14, 28). One of the factors contributing to the uncertainty is the specific surface properties, where nanoscale roughness (29, 30), surface defects (31), and residual stress (32, 33) can all change the electrification results. Another factor is that random and disordered charges are often generated by the inevitable CE processes during the transportation of materials, known as preexisted charges, which can change the initial state of CE. The uncertainty of the CE is a challenge that leads to unstable output power during energy harvesting and inaccuracy of triboelectric sensors. Therefore, it is crucial to propose a method for overcoming the uncertainty of CE and normalizing the charge generation results of CE.

Here, we designed an adhesive triboelectric elastomer (ATE) using polyacrylate-based pressure-sensitive adhesive (PSA) with ionic liquids (ILs) embedded, which achieves a normalized CE result in contact with almost all the commonly used triboelectric materials, regardless of positive type or negative type or even polymer being precharged. The ILs neutralize the preexisted charges on the surface through selective ion transfer so that the targeted surface is reset to initial state. Simultaneously, the PSA chains adhere to the targeted surface, and the subsequent separation process leads to the heterolysis of chemical bonds, resulting in the regeneration of triboelectric charges. Because this charge regeneration is independent of the surface electronegativity and work function of the targeted material, both the charge polarity and density can be normalized. This work has also discussed the coexistence and competition relationship of three charge generation mechanisms of CE, indicating an approach to modulate CE performance using the mechanism combination. This ATE not only enhances the reliability of contact electrical devices but also introduces more conditional variables for further understanding the mechanisms of other physical phenomena in interface science.

RESULTS

As shown in Fig. 1A, at the contact interface of two materials, multiple CE mechanisms may occur in parallel due to the complex surface state of the material. Three mechanisms are widely acknowledged as the origin of the static charges: (i) electron transfer, (ii) ion adsorption, and (iii) material transfer. Electron transfer usually occurs on the surfaces with marked work function differences, where the mechanical force facilitates the electron transfer to lower energy states upon the contact. Ion adsorption, including chemical or physical adsorption, generates charges through ions coordinating with surfaces or binding to unsaturated bonds due to the acid-base coordination or electrostatic adsorption. Material transfer is accompanied by cleavage of the chemical bonds on the molecular chains, which may occur at any position on the polymer chain, leaving molecular fragments with charges on the contact surfaces. Competitive interference between electron transfer and ion adsorption may happen, where ions and electrons usually compete for occupying the same transfer site or coupling on the contact interface (7, 34–37). Material transfer occurs rather independently, since the material transfer process brings different material fragments and active sites to the contact interface. Meanwhile, the result of electrification is dependent on the difference in functional groups on either side of the fracture site, making the material transfer effect to operate in parallel with the other two mechanisms. Hence, using the combination of ion transfer and material adhesion, we proposed this dual function of charge neutralization and controllable charge regenerating as the material design strategy for our work.

Fig. 1. Three charge transfer mechanisms, and the source and overcoming method of preexisted charge of CE.

(A) The processes of electron transfer, ion adsorption, and material transfer occurred at the contact interface of two materials. E_vac, vacuum energy level; HOMO, highest occupied molecular orbital; LUMO, lowest unoccupied molecular orbital; LA, Louis acid; LB, Louis base. (B) The source and disadvantage of the preexisted charge. (C) The design concept of ATE. Temp., temperature. (D) The voltage curve of ion liquid contact with material with preexisted charges. (E) The potential of fluorinated ethylene-propylene (FEP) film contact with several adhesive before and after. The error bars represent the SD based on 5 data points. CA, cyanoacrylate; PVAC, polyvinyl acetate; VHB, 3M very hign bond acrylic foam tape; PSA, pressure ssensitive adhesive; PVA, polyvinyl alcohol; PVP, polyvinyl pyrrolidone. (F) The induced voltage of FEP and polyamide (PA) after contact with commercial PSA and tackified adhesive PSA.

As mentioned above, the preexisted charge is one of the causes for the electrification uncertainty (Fig. 1B). Through material design, such as the addition of conductive polymers and ILs, it is possible to erase the preexisted charge on the surface of the material at each contact, ensuring that the charge density on the surface returns to initial state at each electrification. Meanwhile, with the fixed polarity and fixed quantity of charges generating on this reset surface, which can be achieved by designing specific transfer substances, the normalization and controllability of the CE result can be realized. Therefore, on the basis of the concept shown in Fig. 1C, we have raised out an adhesion material combining ILs and adhesive groups to achieve the normalized CE result. As presented in Fig. 1D, fluorinated ethylene-propylene (FEP) film contacts to a sponge soaked in 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide (EMITFSI), where the original voltage signal drops rapidly to nearly zero and remains at this value after separating, indicating the neutralization effect of ILs happening on the interface. Then, the adhesive surface generates triboelectric charges during CE through the material transfer and the heterolysis effect of chemical bonds, which is independent of the molecular electronegativity and work function of the targeted material. The heterolysis of chemical bond can be more predictable and highly tunable, leading to the regeneration of triboelectric charges with normalized polarity and density. In this case, the charge generated by electron transfer is neutralized by ions indiscriminately, and the targeted surface is reset to the initial state, while the charge regeneration through material transfer is facilitated by strong adhesion on the targeted surface. The two mechanisms demonstrate high compatibility and normalized effectiveness, inspiring the direction of material design for achieving control of CE result.

We considered adhesive material as the main substrate of the material to make the material transfer process to be more controllable. Figure 1E shows the surface electrical potential before and after the contact between several common adhesive cured films and FEP surface, where the largest potential change exhibits on FEP surface before and after the contact with the PSA. It is believed that more material transfer occurs in cured PSA film, and thus, more charges are generated. Moreover, with the adhesion group added to increase the adhesion of PSA, the CE results of FEP (generally considered as negative material) and polyamide-66 (PA66; generally considered as positive material), are not only consistent in polarity (−) but also similar in voltage output scale (as shown in Fig. 1F), which further suggests the feasibility of obtaining normalized CE results through the material transfer mechanism.

We used the condensation polymerization of different acrylate monomers to enhance the stability of PSA CE result. The proportion of soft monomers, hard monomers, and functional monomers affects the hardness and adhesion of PSA. Four types of PSAs are prepared by respectively adding different monomers according to Materials and Methods and the steps in fig. S1 (38): (i) acrylic acid (AA) and 2-ethylhexyl acrylate (2-EHA), as PSA-I; (ii) AA, 2-EHA, and 3-methacrylamidopropyl dopamine (DMA), as PSA-II; (iii) AA, 2-EHA, and butyl acrylate (BA), as PSA-III; and (iv) AA, 2-EHA, BA, and 2-EHA, as PSA-IV. Here, AA represents the hard monomer, 2-EHA and BA are the soft monomers, and DMA provides adhesion group as the tackifier (39–41). Figure 2 (A and B) shows the molecular structures and precuring photos of the four types of PSA. The chemical structures of four types of PSA are verified by ¹H nuclear magnetic resonance (NMR) spectra (fig. S2), and detailed analysis is given by note S1. The results demonstrate that the chemical structures of obtained polymers are consistent with the design. Differential scanning calorimetry (DSC) is used to measure the glass transition temperature T_g, and as shown in Fig. 2C, the addition of the DMA monomer increases the glass transition temperature, resulting in the frozen molecular chains at room temperature, while the addition of BA can lower the T_g, allowing PSA to maintain a viscoelastic state. Figure 2D shows the adhesion forces of four types of cured PSA film on the FEP tested by the method shown in fig. S3, indicating that the addition of BA monomer slightly enhances the adhesion force of PSA by softening the PSA (Fig. 2E), while the addition of DMA evidently enhances the adhesion force through the hydrogen bonding, coordination bonding, or redox and Michael addition introduced by catechol group (42), as illustrated in Fig. 2F. Besides, multiple bonding methods allow the PSA with DMA added to be adapted to various surfaces, improving universality of ATE system.

Fig. 2. The synthesis of PSAs and mechanism of making polarity unified charges after CE.

(A) Chemical structure of four PSAs polymerized of different monomers. (B) Photographs of four PSAs before completely cured. (C) DSC thermogram of four PSAs. a.u., arbitrary units. (D) Force-versus-extension curves and adhesion of four PSAs. (E) The effect of glass transition temperature T_g on the hardness of PSAs. (F) Illustration of the proposed binding mechanism of PSAs to targeted materials. (G) Induced voltage of FEP/PA66 film after contacting to four PSAs once measured in noncontact mode. The error bars represent the SD based on 5 data points. (H) Schematics of charge generation by C─C bond heterolytic cleaving. Top parts are the electron cloud distribution, and bottom parts are electron cloud-potential well models. (I) X-ray photoelectron spectroscopy (XPS) measurement of PSA-IV and FEP film before and after contacting (marked as BC and AC) with PSA-IV once. (J) XPS measurement of PSA-IV and PA66 film before and after contacting with PSA-IV once.

Figure 2G shows the CE effects of the four types of cured PSA film. The induced voltage signals in noncontact mode are tested when the FEP/PA66 film surface is 2 mm away from the PSA surface, and the detailed test method is shown in fig. S4. This voltage signals in noncontact mode reveal the change of charge polarity and charge density after the contact. The surfaces of FEP and PA66 films are thoroughly clean before the test, ensuring no preexisted charge residue. After the first contact, the induced voltage of PSA-III with BA adding increases compared to PSA-I, while the charge polarity of FEP and PA66 generated by contact is different, indicating that the CE is leading by electron transfer and CE result highly relies on the intrinsic surface properties of FEP and PA66. The addition of DMA can also increase the induced voltage slightly, and PSA-II and PSA-IV with DMA added generate the charge with the same polarity (−) to the targeted surface, no matter the targeted material is triboelectric positive type or negative type. These normalized charges come from the heterolysis of chemical bond, such as C─C bond and C─N bond, which is possible to occur in any position in PSAs. Considering the side chain with catechol group adhesive to the surface of targeted materials, this side chain is more susceptible to being stretched by stress. Three positions in the amide group with high possibility to break are marked as a (C─C), b (C─N), and c (N─C) in fig. S5, and density function theory is performed to calculate the bond-dissociation energy E_d, as displayed in table S1. The bond-dissociation energies E_d of three bonds are comparable, while both the breakage of a bond and b bond lead to the negative fragments left on the targeted surface. Thus, the electrification result statistically exhibits negative on targeted surface at the macroscopic level. The distribution of electron cloud is also performed as illustrated in fig. S6. Around the C─C bond marked by the green arrow, the blue area represents the negative potential region distributed around the carbonyl, while the red area represents the positive potential region distributed around the C atom connected with the carbonyl. The difference in electrostatic potential between the two sides of C─C bond increases after the side chain adhesion to the targeted materials. In this case, the shared electrons are expected to remain on the carbonyl side when C─C bond breaks, making the fragments negative. Similarly, the breakage of C─N bond also leads to the residue of negative fragment because of the strong electron-withdrawing ability. This bias of breaking the bond is manifested macroscopically as generating negative charges on the surface of the targeted materials, which is consistent with the experimental results. Thus, as demonstrated in Fig. 2H and fig. S7, the negative charges are generated by the cleavage of the C─C bond or C─N bond locating at the amide group, where the breakage can result in the shared electrons remaining on the side of the electron-withdrawing group (carbonyl or N atom). The x-ray photoelectron spectroscopy (XPS) is carried out to investigate the surface chemical composition of the FEP and PA66 after once contact with the cured PSA-IV film as shown in Fig. 2 (I and J). On the surface of FEP, the peak of N 1s is observed after contact [see line FEP-AC (after contact) in Fig. 2I], while the N element is not a constituent element of FEP, indicating that fragments containing the amide group from cured PSA-IV film left on the surface of FEP. In the C 1s spectra of cured PSA-IV film and PA66 before and after contact, the peak of C─O bond is observed both in the PSA-IV film and the surface of PA66 contact with PSA-IV film once, indicating that fragments containing phenolic hydroxyl group from cured PSA-IV film remain on the surface of PA66. These results demonstrate the occurrence of material transfer, and by modifying the location where the bond heterolyzes, the CE results become normalized. Thus, the PSA-IV is chosen as the main component of ATE to achieve the function of normalized charge regeneration.

By embedding ILs into PSA, the uncertainty of CE caused by the preexisted charge is further eliminated. EMITFSI, as molecular formula shown in fig. S8, exhibits high ionic conductivity and behaves as a compatible binary system of PSA and EMITFSI (43, 44). The working principle of ILs is illustrated in Fig. 3A: (1) Before contact, the disordered preexisted charges distribute on the material surface. (2) During contact, charge transfer occurs, causing further deviation from the electrically neutral on surface. (3) The cations or anions (depending on the preexisted charges) migrate to the surface due to electrostatic adsorption, forming an ion screen layer. (4) After separation, the ions with opposite polarity remain on the material surface to neutralize the surface charge, thereby achieving the effect of preexisted charges neutralization. Sponge filled with EMITFSI respectively contacting to FEP, PA66, and polyimide (PI) films, which have been electrified by contacting with aluminum (Al) film in advance, is used to detect the neutralization effect. Figure 3B shows the induced noncontact voltage signal before and after contact. For three different materials with positive or negative charges, the voltage decrease can reach more than 95%. Movie S1 visually demonstrates the effect of ion neutralization.

Fig. 3. Effect of ILs on charge neutralization.

(A) Mechanism of ILs neutralizing the two types of preexisted charge on targeted material. (B) Induced voltage for three film samples before and after contacting with ILs once. The error bars represent the SD based on 5 data points. (C) Electrical conductivity of PSA-IV with different IL contents. (D) Force-versus-extension curves of PSA-IV with different IL contents. (E) X-ray diffraction (XRD) measurement of PSA-IV, ILs, and 30 wt % content of EMITSFI mixed PSA-IV (30cIL-PSA-IV). (F) Distribution of EMI⁺ and TFSI⁻ on PA66 (a and b) and FEP (c and d) surface. (G) Illustration of ATE combined with PSA-IV and IL.

Before the curing of PSA-IV, EMITFSI is added at the quality ratio of 10, 20, 30, 40, and 50%, respectively. Figure 3C shows the conductivity of the mixed solutions before the curing measured with two-point method (fig. S9), indicating that conductivity increases as the content of EMITFSI increases. As the testing time progresses, the localized distribution of EMITFSI gradually forms as the PSA-IV cures, resulting in a gradual decrease in conductivity and ultimately leading to insulation when the PSA-IV is completely cured. Figure 3D displays the adhesions of the elastomer films cured by the mixed solution to FEP film, which the addition of EMITFSI may also reduce the adhesion, as the localized EMITFSI occupies the contact area between PSA-IV and FEP. Hence, the content of the EMITFSI needs to be further confirmed to ensure that the material transfer process can occur. Figure 3E shows the x-ray diffraction (XRD) tests of EMITFSI, PSA-IV film, and 30 wt % content of EMITFSI mixed PSA-IV (30cIL-PSA-IV for short) film, revealing that the amorphous state of cured PSA-IV film is not affect by the addition of EMITFSI.

We used time-of-flight secondary ion mass spectrometry (TOF-SIMS) to investigate the adsorption of ions on the targeted surface. Figure 3F presents the precharged FEP and PA66 contact with the 30cIL-PSA-IV film, respectively, and the two-dimensional (2D) distribution of cation EMI⁺ and anion TFSI⁻. On the PA66 surface, the number of anions TFSI⁻ [Fig. 3F (b)] adsorbed is much higher than that of cation EMI⁺ [Fig. 3F (a)], which is because more cations are adsorbed to neutralize the positive charges generated in the precontact on the PA66 surface. On the contrary, the distribution of EMI⁺ [Fig. 3F (c)] and TFSI⁻ [Fig. 3F (d)] on the FEP surface indicates that more cations are adsorbed to neutralize the preexisted negative charges on FEP surface. Meanwhile, no matter the surface is absorbing anions or cations, the CE results of FEP and PA contact with ATE are both negative charges. Thus, the effect of IL in CE is not to generate charges, but to neutralize the preexisted charges. The distribution of other ions including C⁺ and C⁻ shown in fig. S10 also suggests that fragments caused by material transfer adhere to the contact surface, indicating the concurrent occurrence of ion adsorption and material transfer process. Therefore, the ATE with binary mixture system is achieved by combining the PSA- and ILs, and the coexistence form of the two components is shown in Fig. 3G, indicating that ILs are regionally embedded in the cured PSA.

Figure 4A presents the induced voltage of FEP/PA66 film at the area of 5 × 5 mm before and after contacting with ATE with different IL contents, with FEP/PA66 contact with the Al film in advanced, which further exhibits the CE effect of ATE. When the content of ILs is low, the absolute value of induced voltage increases for FEP, while for PA66, the absolute value of induced voltage decreases. The reason is that the neutralization effect is not obvious for low content of ILs, and the CE result is the sum of the preexisted charge and the newly generated charges dominated by the material transfer. As the content of ILs increases, the neutralization effects are enhanced to gradually neutralizing the preexisted charge, resulting in the more prominent charge generation by material transfer. When the content of ILs is 30 wt %, normalized induced voltage with same polarity can be generated in both FEP and PA66 films. Furthermore, as the ion content increases, the induced voltage instead decreases, which is because the excessive content of ILs hinders the formation of hydrogen bonds and other bonding forces by PSA with the surface of the contacting material, thereby inhibiting the material transfer. When the IL content reaches 50 wt %, the induced voltage of two samples both reduce to zero after contact with ATE, indicating that the material transfer is hard to occur. Last, 30cIL-PSA-IV is selected as the optimal component ratio for ATE, and its CE effect is measured by contact with several commonly used triboelectric materials for triboelectric nanogenerator (TENG) and sensor (detailed testing method is shown in fig. S11), including the positive and negative triboelectric materials with high electrification property, as well as metal and poor electrification property, as shown in Fig. 4B and more can be seen in fig. S12. Before contact, the surface of the material is cleaned to reach electrically neutral state, ensuring no preexisted charge. After contact, the surfaces of all samples generate charge density by ATE ranges from 50 to 70 μC/m². In addition, for precharged samples, similar results of contact charge density can still be generated, as shown in Fig. 4C and fig. S13. We further explored that the reason for the fluctuation of the charge density within the range was caused by the surface roughness. A detailed explanation can be found in figs. S14 to S17 and note S2. For materials with loose internal structure or a porous, dusty surface, such as paper, wood, etc., ATE will tear and damage the materials (fig. S18), where the normalized CE result cannot be produced. Nevertheless, the design concept of ATE with ILs and adhesive material still illustrates notable effect in CE charge normalization. In the future, it is expected to further improve the normalization effect of CE through the selection of more adhesive and universal materials. Figure S19 shows the detailed time-resolved charge output of ATE and Al contact with FEP (with negative preexisted charges) and PA (with positive preexisted charges). For FEP (fig. S19A), after two contact separation, both ATE and Al carry positive charges, and the charges on FEP surface remain negative. For PA (fig. S19B), after the first contact, negative charges are generated on the surface of PA contact with ATE, while charges on the surface contact with Al remain positive. Furthermore, the signal strength of negative charges on FEP and PA after their contact with ATE is consistent. The comparison further clarifies that the CE mechanism of ATE is different from which is dominated by electron transfer.

Fig. 4. Normalized CE effect and composite charge transfer mechanism of ATE.

(A) Induced voltage of FEP/PA66 film contacted with PSA-IV with different ILs contented. The error bars represent the SD based on 5 data points. (B) Charge density of triboelectric materials with different electrification property contacted with ATE once. The error bars represent the SD based on 5 data points. (C) Charge density of triboelectric materials with preexisted charges contacted with ATE once. The error bars represent the SD based on 5 data points. (D) Atomic force microscope (AFM) map (top) and Kelvin probe force microscope (KPFM) map (bottom) on FEP surface after contacting with the ATE once. (E) Thermally stimulated depolarization currents (TSDC) of FEP films charged by different methods under open-circuit mode. (F) TSDC of PA66 films charged by different methods under open-circuit mode. (G) Mechanism of preexisted charge neutralizing and normalized charge regenerating of ATE.

Figure S20 shows the charge densities on FEP contact with ATE under different normal force. Within the range of the measurement, the charge density of FEP shows slight fluctuations, and the fluctuation amplitude does not exceed 5%. The fluctuation is caused by the contact area of FEP with ATE, which is given that after each separation of the contact, the FEP would briefly adhere to the ATE, resulting in a slight change in the position for the next contact. Overall, pressure does not markedly affect the CE performance of ATE. As long as the target surface can be fully contacted, the processes of ion neutralization and bond-breaking charging will occur completely. Figure S21 shows the voltage outputs of ATE contact with FEP in the single contact-separation process, and the contact time must be at least 35 ms to obtain a relatively stable CE result, with detailed explanations indicated in note S3. Figure S22 reveals the time-dependent stability of charges generated by ATE, which the charge from a single contact cannot be retained for a long time. This is because the charge here comes from the unbalanced electron at the broken bond with high chemical activity. However, this deficiency can be remedied through repeated contact. We conducted tests on the stability of charging and the stability of charge retention within the humidity range of 30 to 80%. As shown in fig. S23, as the humidity continues to rise, the output voltage generated by the contact separation of ATE and FEP remains basically unchanged or even slightly increases. The increase in voltage output might be due to the increase in surface humidity, where the neutralizing effect of the IL is slightly diluted. Because PSA is non–water soluble, the increase in humidity has little effect on its adhesion. Therefore, the charge generated in the previous contact is not completely neutralized and contributes to the charge quantity involved in the next contact separation. After a single contact, the charge retention performance in the noncontact state is shown in fig. S24. As the humidity increases, the noncontact-induced voltage signal decreases, indicating that the charge quantity on the FEP surface reduces. The reasons for the reduction are consistent with those of other types of TENG, namely, the conductive effect of water molecules and the neutralization effect of ions within water molecules. ATE is applicable under room temperature conditions. An increase in temperature will substantially alter the fluidity of PSA, causing its cohesion to decrease and viscosity to increase, thereby preventing the bond-breaking process from occurring. As the temperature gradually rises to 80°C, ATE gradually transforms from a elastomer body to a viscoelastic fluid, as shown in fig. S25. However, this property also gives ATE the function of recoverability. The ATE that has undergone noticeable deformation was placed into an 80° oven for 10 min. After that, the ATE will return to its original state and can be used again (as shown in fig. S26). When ATE is subjected to affect deformation or long-term use resulting in a decrease in the number of surface adhesive groups and a decrease in viscosity, its function can be restored by thermal cycling.

In the result of multiple contact between FEP and ATE (fig. S27A), the voltage of each contact is stable (the fluctuation is within 10%), while that of the FEP contact with Al is unstable and gradually increases (fig. S27B). These results indicate that the process of neutralization and reelectrification occurs during each contact to ignore the effect of the preexisted charge, while the common triboelectric materials are inevitably affected by the charge accumulation effect with each contact. The loss rate of ATE at each contact is calculated by calculating the number of ions required for each complete neutralization of preexisting charges. For ATE generating a charge density of 70 μC/m² on the target surface by itself, then the number of ions required to neutralize these charges can be obtained from the formula

$$n=\frac{\mathrm{\sigma }·S}{e}$$ (1)

where n is the number of ions, σ is the charge density, S is the contact area, and e is elementary charge. Therefore, if the area is calculated as 1 cm², then the number of ions n required to completely neutralize the preexisted charge is 4.3 × 10¹⁰, which is 2.78 × 10⁻⁸ mg, while the mass of IL in 1 g of ATE is ~300 mg. Thus, the loss of ions in each contact is only one part per million. A high-speed camera captures the contact-separation process at a shooting speed of 5000 frames per second, as shown in movie S2 and fig. S28. During the separation process, it can be observed that the adhesive effect causes the FEP membrane to stick to the ATE until it is torn apart. On a macroscopic level, the fragments are invisible. The optical microscope images show surface PSA and IL of ATE after multiple impact cycles (fig. S29A; fig. S29D is the local magnified image) in comparison to original ATE (fig. S29B; fig. S29E is the local magnified image). Although the surface PSA and IL have been lost, the underlying PSA and IL are still sufficient to serve as a supplement for the next material transfer during charging. After being heated to 80°C and then cooled, the surface of ATE gradually returned to its original state (fig. S29, C and F), further demonstrating its recyclability. This stable charging process of ATE offers advantages for the long-term stable application of triboelectric devices.

We conduct atomic force microscope (AFM) and Kelvin probe force microscope (KPFM) to test the surface morphology and surface potential distribution of the FEP surface after contact with the ATE (Fig. 4D). The surface irregularities are caused by the transfer of material fragments, and negative potentials appear at the higher positions on the surface. Figure S30 shows the same tests conducted on the PA66 surface, and the results are the same with the measure on FEP. Figure S31 gives the charge of SiO₂ sample surface potential induced by tip contact with SiO₂ at different temperatures, which shows that the charge generated by electron transfer remains at the temperature of 80°C. Figure S32 gives the charge of SiO₂ sample surface potential induced by ATE contact with SiO₂ at different temperatures, and the potential is mainly concentrated at the elevated areas where are covered by ATE fragments. As the temperature rises, the fluidity of ATE increases, and the fragments gradually spread, which makes distribution of the electric potential also spread and gradually become evenly distributed on the surface. At this point, we can confirm that the source of the charge after contact with ATE is not dominated by electron transfer but is from the charged ATE fragments. Hence, the materials with different triboelectric properties, after contact with ATE, can all acquire the same polarity and similar quantity of charge.

To further understand the contribution to surface charges generated by CE mechanisms, we measure the thermally stimulated depolarization currents (TSDC) of FEP and PA66 after contact with ATE and charging by other methods in open-circuit mode (fig. S33), allowing the activation energy, ion motion, and defect state density to be studied (45, 46). The samples with single-sided gold plating are subjected to electrification treatment as listed in tables S2 and S3, and then all the samples are rapidly cooled to a low temperature of 223 K. After that, depolarization currents of the samples are measured during heating at a constant rate of 3 K/min (0.05 K/s). The TSDC of FEP and PA66 after contacting with ATE show distinct increase as illustrated in Fig. 4 (E and F). Charge electrification by ATE starts to release at around 50°C, suggesting that the relevant devices are suitable for applying at room temperature. On the basis of the Arrhenius relation, the current I is given as (47)

$$I=\frac{S_{\mathrm{A}}{P}_0}{{\mathrm{\tau }}_0}\text{exp}\left(\frac{E_{\mathrm{a}}}{k_{\mathrm{B}}T}\right)\text{exp}[-\frac{1}{{\mathrm{\tau }}_0\mathrm{\beta }}\underset{T_0}{\overset{T}{\int }}\text{exp}\left(\frac{E_{\mathrm{a}}}{k_{\mathrm{B}}T}\right)\text{d}T]$$ (2)

where β is the heating rate, S_A is the electrode area, τ₀ is the preexponential factor, E_a is the activation energy, k_B is the Boltzmann constant, and P₀ is the initial dipolar polarization. Therefore, the activation energy E_a can be estimated from the slope of the lnI – 1/T curve (fig. S34)

$$\text{ln}I=\text{ln}\left(\frac{S_{\mathrm{A}}{P}_0}{{\mathrm{\tau }}_0}\right)-\frac{E_{\mathrm{a}}}{k_{\mathrm{B}}T}$$ (3)

The activation energy E_a of charges excited in FEP/PA66 after contacting with ATE is much larger than that of the pure FEP/PA66 film and samples electrification by other method, as shown in fig. S35. The high activation energy E_a, which reaches 14.66 eV for FEP and 14.71 eV for PA66, is near the first ionization energy of C, suggesting that the charges generated by contacting with ATE come from the C–C bond cleavage while the charges generated by other methods are electrons trapped in localized defects with much lower activation energy. More details about the comparison of the activation energy E_a are explained in note S4.

The ATE with the dual function of preexisted charge neutralization and normalized charge regeneration is recognized, and the specific physical picture of ion adsorption and material transfer during the CE process is illustrated in Fig. 4G. When triboelectric material with preexisted charges contacts with the ATE surface containing localized ILs and highly adhesive catechol group, ions from ATE surface can neutralize the preexisted charges on the triboelectric material under the driving force of electrostatic interaction. Synchronously, the catechol group forms a strong bond with the binding sites existing on the triboelectric material through hydrogen bonds, coordination bonds, or nucleophilic substitution reactions. After the separation, the transferred ions remain on the surface of the triboelectric material, while the carbon chain is stretched as the triboelectric material moves away. Because of the lower bond-dissociation energy and uneven electron cloud distribution, heterolysis of C─C bond and C─N bond near the amide group exhibits a higher probability of occurrence, thereby negative charged fragments are generated. Last, ions and fragments with negative charge are left on the triboelectric surface, obtaining the normalized CE result. In this way, the uncertainty caused by the preexisted charges and properties of the material surface can be eliminated by ion adsorption and material transfer, achieving consistent CE results as shown in fig. S36. Moreover, in ATE, the selection of ILs and adhesion groups are both interchangeable, indicating adhesive approaches for further improving the performance of ATE.

Previous studies have successfully demonstrated the CE effect and the development of related materials for sensor applications (48–50). The ATE materials, as discussed above, exhibit stable, long-term charging effects, ensuring the stability and accuracy of electrostatic sensors. Whether in a packaged structure (fig. S37) or a single-electrode open sensing setup, each interaction with the ATE material results in a consistent charging process, addressing the issue of charge fluctuation in electrostatic sensors. For instance, we conduct a measurement on the dynamic noncontact sensing performance. The FEP film moved from being in contact with the ATE to a position 9 mm away and then moved back toward complete contact with ATE. As shown in fig. S38, the displacement at a farther distance and the variation of the induced voltage can be well matched. This consistency is particularly beneficial for the emerging field of noncontact sensing, especially electrostatic tracking technology. Electrostatic tracking technology uses the electric field induced by electrostatic charges to capture the movement trajectories of objects, achieving tracking effects in nonvisual states. As an auxiliary tracking technology to visual tracking, it is widely applied in fields such as virtual reality, robot control, and embodied intelligence (51–53). Electrostatic tracking relying on common triboelectric materials faces the problem of signal variations across different objects, which often requires precalibration to ensure accurate signal interpretation (54, 55) and further introduces complexity in postprocessing. Moreover, the system struggles to track unknown objects, as their signals have not been preidentified or entered into the system. On top of this, in open tracking environments, the surface conditions of objects are random, making signal recognition for electrostatic tracking even more challenging. However, by ensuring consistent electrification process, the random variations in the tracked object’s signals are eliminated, and the highly controllable electrostatic charges can be used as a marker, much similar to a Bluetooth signal transmitter, which allow the precise object tracking to be achievable.

Sensor arrays (2 by 2 and 6 by 6) based on ATE and ionic conductive elastomer are used to demonstrate the charge marking and motion tracking functions. Ionic conductive elastomer as electrode of the sensor is produced by 3D printing, and the conductivity of the electrodes is tested to reaches 10⁻³ S/m, making it suitable as a sensor electrode. Meanwhile, we also tested the sensitivity of the electrode resistance, as shown in Fig. 5A. Within a strain of 60%, the rate of change in electrode resistance shows a certain linearity, which has the potential to be integrated into a multimode sensor for pressure sensing and noncontact sensing. The array sensors are integrated onto a double-layer PI films, while 2-mm-wide gaps and 1-mm-diameter holes are reserved on the upper PI film to facilitate the cross of the strip-shaped FEP film and the 3D-printed electrode and the flexible printed circuit (FPC) is on the lower PI film as the output channels, as shown in fig. S39. FEP strips (2-mm-wide) with holes are horizontally and vertically inserted through the PI film, and the 3D-printed electrodes pass through the holes reserved in the PI film and the FEP film, respectively, connecting to the lower FPC and the upper ATE. Figure 5B and movie S3 show the normalization function on CE signal of the sensor tested using a 2 by 2 sensor array. For a FEP film with preexisted charge, (1) a small amount of induced voltage signal is detected when approaching the sensor; (2) when the FEP film contacts with the sensor, the preexisted charges are removed; (3) during separation, charges are generated through material transfer; and (4) when the FEP film is approaching the sensor again, the newly induced voltage is detected. The same process also occurred on the PA66 surface, as shown in the second part of movie S3. For a PA66 film with preexisted charge, the preexisted charge is removed after contact with the sensor, and the sensor detected the same induced charge on the same distance as the FEP surface after separation. Although the intrinsic properties of the materials are different and they carry their preexisted charges, the induced signals after contact are consistent. As a result, the materials to be tracked after the contact with the ATE are all marked with consistent and standardized markers.

Fig. 5. ATE-based sensor array for noncontact sensing.

(A) Electrical conductivity of ionic conductive elastomer under different strain. The inset is the measurement method of the conductivity. (B) Voltage of 2 by 2 ATE-based sensor array before, during, and after contact with FEP film. (C) Schematic of the ATE-based sensor array. (D) The photo of ATE-based array. (E and F) Electrostatic tracking function of ATE-based sensor array. The array is abstracted as a 6 by 6 checkerboard, and the probe positions are indicated by purple and green dots. In the visualized data, the arrows indicate the position of thee probes.

Figure 5 (C and D) and fig. S40 show the diagram and the photos of the 6 by 6 array; the whole area is 3.5 by 3.5 cm, and the width and height of each sensor unit are both within 3 mm. It can be used in a bent state, enabling the sensing array to be conveniently integrated onto complex surfaces. Figure 5 (E and F) and movie S4 show the data visualization of the electrostatic tracking signal during detecting the polytetrafluoroethylene (PTFE) and nylon probes in noncontact mode. Before and after the contact with ATE, the PTFE probe is moved to the same position [Fig. 5E(a)], and a visible signal change is observed after the contact [Fig. 5E(b)]. The same signal change is also observed for the nylon probe (Fig. 5F), revealing the same charge generation to the different probes for tracking application. Because of the preexisted charges, the detection electrodes detect different signals in the same height. If visual access is restricted and the material properties cannot be foreseen in advance, it is impossible to determine whether the difference in the signals is caused by distance or by the material itself. Therefore, it is impossible to calibrate the position and distance of the samples using a normalized signal (Fig. 5, E(b) and F(b)]. After contacting with ATE, the normalized charges produce the signals with the same height of two probes [Fig. 5, E(c) and F(c)], indicating that the detection signal is no longer affected by the material properties and the preexisted charge. Using this kind of normalized charges, the array has the potential for applications in supervision tracking without knowing the charging history of the objects in advance, as illustrated in fig. S41. The sample marked by ATE moves to the random position, and then, the sensor can track to the corresponding position by detecting the variation of the normalized voltage signal, as more details described in note S3.

DISCUSSION

In this work, to achieve a normalized and highly controllable CE process, we propose a strategy to fabricate electrification material based on the combination of the ion neutralization effect and the material adhesion effect materials. The ATE film using ILs EMITFSI and PSA as the components is synthesized according to the above material design strategy, which can maintain normalized polarity and highly controllable charge density after CE with different materials. Here, ILs are capable of neutralizing the preexisted charges on the contact surface, leading to a charge neutralization effect universally adapted to various material surface, while PSA with adhesion group addition ensures the generation of the charges with fixed polarity and almost identical charge density on the targeted surface. The heterolysis probabilities of different chemical bonds are also discussed, which allow us to further control the performance of this ATE. At the optimal component ratio of 30 wt % EMITFSI and 70 wt % PSA, all common triboelectric materials [nylon, FEP, Al, nitrile rubber (NRB), and so on] that contact the ATE generate negative charges ranging from −50 to −70 μC/m² on the surface, regardless of the existence of initially surface charges (positively or negatively precharged). For a very long time, it is believed that the CE results between different materials should always follow the rules of electrification sequences, including the induced charge polarity and the changing tendency of charge density. However, the proposed ATE can realize a CE process against this common sense, which frees the electrification process from the constraints of the electrification sequence. We also conducted an in-depth discussion on three charge transfer mechanisms of CE, which confirms the feasibility of controlling CE outcomes via this composite electrification mechanism.

This ATE material demonstrates substantial application potential in triboelectric sensors, as the normalized and highly controllable electrification process ensures a superior stability and repeatability. Furthermore, ATE can also be used for developing some special functions, such as noncontact electrostatic tracking sensors that can track the moving object from a distance using electrostatic induction. Because of the standardized charging property of ATE, it can effectively solve the problem of disordered tracking signal caused by the nonstandardized surface conditions and variable preexisted charge of objects, further reducing the difficulty of electrostatic tracking. For the demonstration, a sensor array based on the ATE is fabricated, which produces the same intensity of sensing signals (up to 0.22 V at a distance of 5 mm) with different moving materials. Hence, this ATE sensor array can rapidly identify the position information of the moving objects, while the recognition accuracy and the design complexity of system can be substantially improved compared to the devices that use traditional triboelectric materials. The normalization and determinacy of CE results are anticipated to facilitate the practical application of triboelectric devices and contribute to the advancement of interface science by transforming the uncertainty of interfacial electrification into more stable and controllable process.

MATERIALS AND METHODS

Materials

AA, BA, 2-EHA, DMA, and lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) were purchased from Shanghai Macklin Biochemical Co. Ltd. 1-methylimidazole, bromoethane, ethyl acetate, 2-methoxyethyl acrylate (MEA), isobornyl acrylate (IBA), ethyl (2,4,6-trimethylbenzoyl) phenylphosphinate (TPO-L), poly(ethylene glycol) diacrylate (PEGDA) were purchased from Shanghai Aladdin Biochemical Technology Co. Ltd. Unless specified otherwise, all materials are used as received.

Synthesis of four types of PSA

The typical synthesis of the acrylic PSAs is as follows: A total of 4 g of monomers with different molar ratios of AA, BA, and 2-EHA was dissolved in 8 g of n-butyl alcohol, followed by the addition of 40 mg of dicumyl peroxide (DCP). The specific molar ratios of the individual components of four types of PSAs are shown in table S4. Then, the solutions were degassed by purging with argon gas for 10 min before being stirred at 65°C for 24 hours. After the reaction, the products were dried at 90°C under vacuum for 12 hours.

Synthesis of EMITFSI

The synthesis of EMITFSI is as follows: 1-ethyl-3-methyl imidazolium bromide (EMIBr) was first prepared by the quaternization reaction of 1-methyl imidazole with ethyl bromide at 80°C for 24 hours in cyclohexane under refluxing condition. The crude product was purified by repeated recrystallization using a mixture of ethyl acetate and isopropyl alcohol (volume ratio: 1:1) as the solvent. Last, the anion exchange reaction from Br⁻ to TFSI⁻ was carried out by heating an equimolar mixture of EMIBr and LiTFSI in water at 70°C for 24 hours to be phase separated into hydrophobic IL and aqueous phases. The obtained EMITFSI was repeatedly washed with pure water and then dehydrated under vacuum at 120°C for 72 hours.

Synthesis of ATE

Before the PSA solutions were dried and cured, EMITFSI was added into the solutions at the weight percentages of 10, 20, 30, 40, and 50%, and the mixtures were stirred at room temperature for 10 min. Then, the mixtures were dried at 90°C under vacuum for 12 hours.

Synthesis of ionic conductive elastomer for 3D-printed electrode

The synthesis of ionic conductive elastomer is as follows: LiTFSI was dissolved in liquid binary mixtures of 35.72 ml of MEA and 14.28 ml of IBA at room temperature. Then, 0.5 ml of cross-linker PEGDA and 0.5 ml of photoinitiator TPO-L were added into the mixture solution to make the 3D printable ink. Last, a bottom-up DLP 3D printer (C02, PioCreat) was used for 3D electrode printing.

Characterization and measurements

¹H NMR spectra were collected at 25°C in dimethyl sulfoxide–d₆ using a Bruker 400M spectrometer. The DSC experiments were performed using a DSC 214 instrument (Netzsch, Germany) under nitrogen, and the glass transition temperature of the PSAs was measured from the second heating cycle at a heating rate of 10°C/min. The adhesion forces were performed on a universal material testing system (Instron E3000, UK) at a rate of 5 mm/min. The XPS (ESCALAB 250Xi, Thermo Fisher Scientific) was performed to analyze the surface element composition of the samples. The surface element composition information of the samples was characterized by a TOF-SIMS (PHI nanoTOF II, ULVAC). The XRD data were collected using a X′Pert³ Powder X-ray diffractometer with Cu Kα radiation (2θ = 1.5416 Å). The AFM and KPFM data were obtained using a Dimension Icon AFM instrument (Bruker) in the tapping mode. The TSDC were measured using a dielectric impedance spectrometer (Nanocontral Concept 90). Voltage and charge were measured by a programmable electrometer (Model 6514, Keithley). The surface potentials were measured by an electrostatic voltmeter (Model 341B, Advanced Energy Industries). The electrical resistance was measured by a 6 ¹/₂ digit digital multimeters (Model TH1963, Tonghul). A color high-speed camera (FASTCAM Mini, Photron, Japan) was installed at the experimental sample setup. The camera was controlled via the Photron FASTCAM Viewer software (PFA).

Sensor array fabrication

The designed FPC and PI films with gaps and holes were shown as fig. S21. The 3D-printed electrodes passed through the holes, then six 2-mm-wide FEP strips were successively passed through the gaps arranged horizontally, while the other six FEP strips were successively passed through the gaps arranged vertically. The reserved holes on the longitudinal and transverse FEP strips were precisely aligned with each other, allowing the electrodes to pass through. Then, the bottom of the electrode contacted to the pads precoated with printable ink on the FPC. After being exposed to ultraviolet light with a wavelength of 375 nm for 10 min, the electrodes were connected to the FPC and fixed onto the FPC. Last, the ATE was placed on the top of the electrode by dispensing.

Statistics and reproducibility

All experiments were repeated at least three times. Data were analyzed as means ± SD. Statistical analyses for multiple samples were conducted and presented through various graphical representations such as bar graph, line chart, and point plot. Relevant sample data points were directly visualized in the graphs. Data analysis and graphical representation were carried out using Origin 2021 and Excel.

Supplementary Materials

The PDF file includes:

Figs. S1 to S41

Notes S1 to S5

Tables S1 to S4

Legends for movies S1 to S4

Other Supplementary Material for this manuscript includes the following:

Movies S1 to S4