Contact Electrification at Dielectric Polymer Interfaces: On Bond Scission, Material Transfer, and Electron Transfer

Triboelectric nanogenerators (TENGs) are revolutionizing mechanical‐to‐electrical energy harvesting. TENGs harvest energy through the polymer–polymer contact electrification (PCE) mechanism, driven by nanoscale processes at the contact interface. Currently, when discussing PCE there are two distinct schools of thought on which nanoscale interactions drive charging at the contact interface; 1) electron transfer, where orbital overlap leads to charge tunneling between polymers; or 2) mass (material) transfer, where polymer chain entanglement and intermolecular bonding leads to heterolytic bond scission. Here, a combination of in silico and benchtop experiments is used to elucidate the relative role of electron and mass transfer in PCE. In silico experiments show that covalent bond scission in a polymethylmethacrylate/polytetrafluoroethylene system occurs at 348 kcal mol−1, prior to electron cloud overlap, where the highest occupied molecular orbital and lowest unoccupied molecular orbital of the system remain separated by 163 kcal mol−1. Benchtop experiments show PCE‐generated charges cannot be simply discharged via electrical grounding, indicating the formation of bound surface charge from mass transfer. The calculations and contact‐electrification tests provide strong evidence to support mass transfer being the leading mechanism driving PCE.


Introduction
Polymer-polymer contact electrification (PCE) has an important role in energy-related applications to power autonomous DOI: 10.1002/admi.202300562microdevices and enable portable electronics. [1]The triboelectric nanogenerator (TENG) has been developed in numerous forms, exploiting PCE for harvesting mechanical energy from movement, [2] water drops, [3] ambient noise, [4] wind, [5] and water waves. [6]ENGs typically consist of a polymer surface that undergoes contact-frictionseparation with another material (either polymer, ceramic, metal, or fluid) and electrodes that convert surface charges generated at the interface into an electrical output.This electrical output of a TENG is dependent on the magnitude of contact surface charge on polymer, and the velocity of relative motion between contacting surfaces (according to Maxwell's equations).Here, we focus on surface charge generation on the polymer, which arises due to nanoscale interactions at the contact interface.PCE is critically important to understand to develop materials with higher surface charge generation and energy harvesting performance in TENGs.Currently, there are three mechanisms proposed for driving PCE: i) electron transfer; ii) ion transfer, and iii) material transfer.Electron transfer has been proposed to occur by overlapping electron orbitals from two distinct polymers under contact. [7,8]on transfer is considered between water adsorbate layers on two polymer surfaces due to different adsorption affinity. [9]In mass transfer, heterolytic bond cleavage occurs during contact separation, and the resultant charged fragments of macromolecules are transferred. [10]Among these mechanisms, the ion transfer mechanism seems to be reliable in case of water/polymer contact electrification, but not critical. [9,11]The surface charge formation on polymers has been measured in the complete absence of water. [12]hus, the mechanistic discussion for PCE remains between the electron transfer and mass transfer mechanisms.
Electron transfer has been proposed and championed by theoretical [7,13,14] and experimental studies. [15]In these theoretical studies, electron transfer has been conceptually assumed to occur by overlapping electron clouds.The transfer of an electron would then leave the donating polymer, and an extra electron would move into the accepting polymer.However, the energy for overcoming the repulsion (as described in the Lennard Jones Potential) to achieve this electron cloud overlap is not considered in the theoretical models. [7,14]In contrast, in the mass transfer mechanism bond cleavage occurs due to intramolecular bond formation and chain entanglement at the polymer interface, and the charge is produced by heterolytic charge transfer between broken fragments. [16]The electron and mass transfer mechanisms have similar end results, but significantly different driving forces, and thus a resolution is required to appropriate polymers for TENGs. [17]erein, we provide a simple comparison of the energy required for electron transfer or bond scission to occur between dielectric polymers, polytetrafluoroethylene (PTFE), and polymethylmethacrylate (PMMA).These simulations are complemented by simple experiments showing the inability to discharge contact-charged dielectric polymers by direct electrical grounding.

Contacting Oligomers
To study the viability of the electron transfer model, we have used density functional theory calculations to model the energy required to overcome the electron cloud repulsion force and compared it to the energy required to break covalent bonds within a polymer backbone.A combination of the PTFE and PMMA oligomers was used as a model system.Experimentally, PTFE tends to charge negatively, and PMMA tends to charge positively. [18]Steered molecular dynamics calculations were performed on 4-monomer oligomers of PTFE and PMMA, where the PMMA oligomer is pushed into the PTFE oligomer using moving harmonic restraints (Figure 1, see Supporting Information for full calculation details).
The potential energy of the system, E pot , increased by a total of 348 kcal mol −1 Figure 1a until immediately decreasing once a C─C backbone bond of the PTFE oligomer fractured.This bond cleavage occurs at a minimum F to H distance of 1.55 Å Figure 1b, which is beyond the distance required for electron cloud overlap.The correlation between repulsive force and F to H distance is clear and follows a Lennard Jones potential model Figure 1c.Further, analysis of the Molecular Orbital PACkage (MOPAC) [19] highest occupied molecular orbital (HOMO)/lowest unoccupied molecular orbital (LUMO) of the PTFE-PMMA system showed minimal change in the band-gap at the closest point prior to fracture (1.55 Å H F separation), with a further 163 kcal mol −1 being needed to promote an electron between orbitals Figure 1.The position of the HOMO and LUMO bands are also interesting, showing the HOMO is present on the PMMA molecule, but significantly spatially separated from the PTFE molecule.Two equivalent energy LUMOs are present, both within the PMMA and the PTFE (Figure 1e,f).The spatial localization of these orbitals suggests that intramolecular charge transfer will preferentially occur before intermolecular charge transfer-that would provide a further energy barrier for the electron transfer model of PCE.

Evidence for Electron Transfer
Experimentally, evidence supporting the occurrence of electron transfer has been provided by the decline in surface charge when heating, [20] by photon emission when contacting separating, [8,15] and the reduction of metal cations on negatively charged surfaces. [21]Thermal emission was observed when inducing a triboelectric surface charge on the dielectric semiconductor (SiO 2 ) by rubbing against a polymer.The surface charge on SiO 2 was measured after heating at a certain temperature.The authors observed that the surface charge decays faster at higher temperatures (especially fast at temperatures above 300 °C) and attributed this observation to electron thermionic emission. [22]owever, the decay in surface charge upon heating can also be attributed to the thermal decomposition of transferred polymer pieces (organo-ions) onto the surface of SiO 2 .In contrast to predictions for electron transfer, polymer contact was observed to charge more strongly at elevated temperatures despite the fact that the charged species have been shown to be less stable on SiO 2 at higher temperatures. [22]A key difference here was in measurement approaches; here the polymer was heated prior to and during contact-separation whereas in prior studies heating was performed after polymer contact.We have shown that by heating glassy polymers such as polystyrene, polycarbonate, and PMMA above the glass transition temperature, an order of magnitude increase in contact surface charge density can be measured. [23]his is attributed to the polymer transition from a glassy to a rubbery state at which the surface becomes softer and more adhesive.These parameters enhance the covalent bond cleavage, material transfer, and surface charge formation as discussed below. [24]hoton emission during polymer contact-separation was measured experimentally and taken as additional evidence for electron transfer in PCE. [15]The authors provided experimental contact electrification-induced emission spectra, [15] but did not consider relevant aspects of the involved molecular orbitals or tribological surface damage from bond breakage.It is well known that photoemission can be also related to covalent bond cleavage, as demonstrated in several studies. [25]The electron transition from molecular orbitals to atomic orbitals may entail bond breakage, which would effectively support the material or ion transfer-based mechanism.
The electron transfer mechanism has been also claimed for PCE after observing metal cation reduction and nanoparticle deposition on surface of negatively charged contact-separated polymer. [21]However, it was demonstrated that the metal cations are reduced and deposited also on positively charged contactseparated polymers where the electrons were not expected to be available for reduction reactions. [26]In fact, it was proven that metal particles on polymer surfaces are deposited after reacting with mechano-radicals present on the surfaces of contactseparated polymers from homolytic bond cleavage. [21]me authors have proposed PCE mechanisms that consist of an initial formation of mechano-ions and mechano-radicals, followed by further electron transfer between the radicals and ions. [14]This mechanism also does not explain why the surface charge is forming at the same magnitude as in the first contactseparation.In addition, if this mechanism would be valid, the polarity of each surface would fluctuate during contact-separation cycle.Furthermore, these hybrid mechanisms rely on initial polymer bond cleavage and mass transfer to occur.Since, radicals are energetically favorable with respect to the respective ions, it may be concluded that the mechanism would pose no advantage against charged fragment/ion-based CE mechanisms in terms of reaction energy.Electron transfer between mechano-ions and mechano-radicals is energetically equivalent to the formation of the new mechano-ions and mechano-radicals-both can be regarded as equally viable in terms of reaction energy. [14]Moreover, in view of the recent finding radicals play a key role in stabilizing ionic fragments, [27] the electron transfer based CE can be regarded as energetically unfavorable.
Thus, all the strong supporting evidence for electron transfer driven PCE can also be explained by alternative mechanisms.

Electron Transfer Kinetics
To probe the theoretical probability of electron transfer-based PCE, the electron transfer rate was determined for the polymer/polymer (PMMA/PTFE) system and compared to a polymer/metal (Al/PTFE) system.Monomers of PMMA and PTFE (Figure 2a; Figure S1, Supporting Information) and dimers of Al were employed for the estimation.At equilibrium distance, the electron transfer rate for the PMMA/PTFE system was calculated to be extremely low-5.07× 10 −56 e s −1 (Table 1).In contrast the charge transfer rate for the Al 2 /PTFE system showed a significantly higher rate of 1.33 × 10 −10 e s −1 , an increase in charge transfer rate of over 45 orders of magnitude.
To observe if this charge transfer rate in the PMMA/PTFE system increased significantly, the molecules were brought stepwise into proximity from 3.1 to 0.9 Å (Figure 2b,c).These static simulations do not assess molecular motion, just the charge transfer when molecules and electronic orbitals are forced to overlap. Figure 2b shows the calculated charge transfer from PMMA to PTFE at given positions.Here, there are two points of interest, first at 1.55 Å there is a local negative charge transfer maxima corresponding to the point at which covalent bond scission occurred in Figure 1.Second, at 1.13 Å there is a global positive maxima corresponding to the point where electron cloud repulsion force begins to dominate (Figure 2c).This global positive maxima rapidly leads to a continuous increase in negative charge transfer as the molecules come closer together-indicating this is where charge transfer via the electron cloud model in polymer can occur.Critically, this occurs at significantly lower separation distances than the point at which heterolytic bond cleavage was shown to occur in Figure 1.Further discussion of these simulations are provided in Figure S2 (Supporting Information).
The static simulation results conclusively demonstrate that even if the covalent polymer bonds do not break when the molecules are brought together, significant charge transfer only occurs once the potential energy (E pot ) for the system is in the electron cloud repulsion region, where it becomes the dominant effect in the energy landscape-at separation distances (below 1 Å) that cannot typically occur in triboelectric testing.At larger, equilibrium, distances the charge transfer rate is on the order of 10 −56 electrons per second for a polymer-polymer system, meaning negligible electron transfer will occur over the lifetime of such a device.

Evidence for Mass Transfer
In the mass transfer mechanism, heterolytic and homolytic covalent bond scission occurs. [28,29]It is known that mechano-radicals and mechano-ions are formed on fractured surfaces that has been demonstrated to occur also during PCE. [28,29]Scission of organic molecules is accompanied by either homolytic bond cleavage (leaving a radical on each fragment) or heterolytic bond cleavage (leaving an ionic charge on each fragment).While the bond cleavage and formation of ionic moieties have been well discussed in related mechanochemistry literature, [10,[30][31][32] the field of mechanical energy harvesting has failed to adapt accordingly.
Kelvin Probe force microscopy has revealed the formation of charge mosaics on both contacted-separated polymer surfaces consisting of mechano-radicals, mechano-cations, and mechanoanions. [26,33,34]Dielectric polymer surfaces do not charge uniformly positively or negatively.Instead, the surfaces develop random patterns of oppositely charged regions of nanoscopic dimensions. [28]Heterolytic and homolytic covalent bond cleavages lead to co-existence of ionic species and radicals on the contacted polymer surfaces.The net surface charge depends on the extent of the ratio between the (+) and (−) regions.
][37] Surface charge increases with decreasing the elastic modulus of the polymer. [24]rface charge has also been shown to increase by increasing the difference in hardness between contacted polymers regardless the chemical composition. [24,37]The strong surface charge can be measured even when contacting chemically identical polymers with different thermal history, surface roughness, or filler content. [24,35,37]The surface charge between identical polymers is rising with an increased discrepancy in the deformative properties of the polymer pair.Further, experimental results have shown that polymers that are intentionally designed to exhibit stronger adhesion to the contacted surface can exhibit up to three orders of magnitude higher surface charge than an equivalent polymer with four times weaker surface adhesion. [36]ofter elastomers are more prone for material transfer than the harder ones.They provide a higher contact area and density of van der Waals intermolecular bonds.When the energy of the intermolecular forces at the triboelectric interface becomes larger than the covalent bond's energy in the polymer backbone the bond scission and mass transfer occur.This relationship can be expressed by equation E adh × (n) > E cov , where E adh is the adhesive energy of the polymers repeating unit to the opposite surface, E cov is the dissociation energy of covalent bond and n is the critical number of polymers repeating units necessary for covalent bond scission to occur. [18,38]Stronger intermolecular bonding between contacted surfaces (adhesion) will enhance material transfer and bond break.For example, the minimum contact length of PMMA molecules that would be necessary for backbone scission to occur upon separation of contact states of PTFE/PMMA is only 143 Å (Section "Simulation Methods", Supporting Information).
Further, experimental measurements of transferred polymer particles have been observed by numerous groups via an array of techniques including XPS, [10,16,39] AFM, [10,16] Raman, [40] and IR spectroscopy, [40][41][42] providing strong evidence for material transfer occuring.Notably, recent work by Shi et al, [39] showed that the adhesive force during PCE directly correllated to the measured  surface charge, proving mechanical interactions and bond cleavage must contribute to the observed charging phenomena.

Discharging Contact Charged Polymers
To provide experimental evidence to support these assertions, the surface charge on Al was measured (in non-contact mode, oscillation between 0.1 and 5 mm separation) after contact with PTFE.The Al was then grounded and charge remeasured (Figure 3a-i).
Prior to grounding, the surface charge on the Al after contactseparation with PTFE was 6.60 ± 0.16 pC cm −2 (current measurement used for calculation shown Figure 3a-ii).After grounding, with an earthing wire, significant surface charge density 1.85 ± 0.08 pC cm −2 remained on Al.This charge arises as surface-bound organo-ions and organo-radicals cannot be electrically discharged via grounding.In contrast, electrons can be easily removed by grounding.To demonstrate this effect, the Al electrode was charged by induction from PTFE film without contact (Figure 3b-i).Before grounding the induction-charged Al possessed charge density of 3.75 ± 0.18 pC cm −2 , but grounding removed all charge present on Al (Figure 3b-ii).This demonstrates that the charges generated are not as mobile as electrons that are able to be removed by electrical grounding, given the nature of Al these charges must be stabilized surface ions-predicted to arise due to the material transfer mechanism.Analysis of the Aluminum plate before and after PTFE contact was performed by atomic force microscopy, which showed an increased surface RMS roughness from 70 nm for the pristine aluminum plate up to 110 nm of the aluminum plate after 10 000 contact cycles (Figure S4, Supporting Information).This increase in surface roughness presents a clear correlation between surface charging and material transfer.The presence of non-mobile charge demonstrated in Figure 3a-ii, coupled to this roughness increase is compelling in indicating the role this material transfer plays in PCE.

Conclusion
In summary, simulations conclusively show that bond cleavage occurs prior to electron transfer occurring in a PTFE/PMMA system.Further, there exists enough computational and experimental evidence of the significance of material transfer mechanisms in PCE processes.The conclusion can be critical for successful optimization and implementation of polymer based triboelectric devices including TENGs.
Osvlads Verners is currently a postdoctoral researcher at Riga Technical University.He received his Ph.D. from Pennsylvania State University in 2014.His research focus includes computational studies of triboelectric and tribovoltaic energy harvesting materials at continuum and atomistic scale, along with computational design and optimization of scalable architectures of microscale energy harvesting devices.

Figure 1 .
Figure 1.a) Potential energy and b) minimum F-H distance (represents the closest interatomic distance between the fragments in contact) versus reaction coordinate during contact simulation of PMMA/PTFE fragments (reaction coordinate corresponds to simulation step count due to linear displacement constraints being imposed); c) Potential energy and d) HOMO/LUMO energies (of the PMMA/PTFE system) versus minimum F-H distance, estimated for trajectory up to backbone dissociation; and spatial distribution of HOMO and LUMO of PTFE/PMMA in the e) relaxed, and f) compressed contact states (─O, ─H, ─F, ─C).

Figure 2 .
Figure 2. Simulation of electron transfer between PTFE and PMMA monomers.a) PTFE and PMMA monomers.b) Calculated charge transfer as these monomers are forcibly brought closer together.c) The potential energy of the system with separation showing electron cloud repulsion becomes dominant (enabling electron transfer) when the potential flips from a negative to a positive value at 1.13 Å.This 1.13 Å corresponds to the point where charge transfer begins in (b).The dotted purple line shows the energy of heterolytic bond cleavage at 1.55 Å from Figure 1 (in these calculations bond cleavage does not occur as they are static simulations).

Figure 3 .
Figure 3. Current measurements during oscillation of PTFE a) contact-charged and b) induction-charged Al electrode before and after grounding.Oscillation was between 0.1 and 5 mm separation.

Table 1 .
Partial charge transfer rates and reaction energies in minimum energy contact state (sign of charge transfer corresponds to the first reactant).