Engineering Polymer Interfaces: A Review toward Controlling Triboelectric Surface Charge

Contact electrification and triboelectric charging are areas of intense research. Despite their low ability to accept or donate electrons, polymer insulator based triboelectric nanogenerators have emerged as highly efficient mechanical‐to‐electrical conversion devices. Here, it is reviewed the structure–property–performance of polymer insulators in triboelectric nanogenerators and focus on tools that can be used to directly enhance charge generation, via altering a polymer's mechanical, thermal, chemical, and topographical properties. In addition to the discussion of these fundamental properties, the use of additives to locally manipulate the polymer surface structure is discussed. The link between each property and the underlying charging mechanism is discussed, in the context of both increasing surface charge and predicting the polarity of surface charge, and pathways to engineer triboelectric charging are highlighted. Key questions facing the field surrounding data reporting, the role of water, and synergy between mass, electron, and ion transfer mechanisms are highlighted with aspirational goals of a holistic model for triboelectric charging proposed.

surrounding data reporting, the role of water, and marriage between mass, electron, and ion transfer mechanisms are highlighted with aspirational goals of a holistic model for triboelectric charging proposed.

Introduction
Developing pathways and materials that efficiently convert ambient energy from the environment into electricity is critical to powering distributed sensing networks, wearable & implantable electronics, and other Internet of Things devices. [1]Small amounts of motion are everywhere in our society, from human movement, blood flow, railway vibrations, car vibrations, bicycle motion, geological activity, or even simply water flowing through pipes.This abundance has led to an explosion of research in small scale mechanical energy harvesters.][6] The goal of these mechanical energy harvesting techniques is to energy from an otherwise stable material, and convert it into useable electricity.Adding functionality where a material can also harvest heat, and/or light, can also boost the energy harvested, and reduce losses. [1]Generally, such harvesting is practical to autonomously power low-energy devices (µW-mW).To minimise energy waste, energy harvesting is most useful when the harvested energy matches the necessary energy to power a specific system.Erreur !Source du renvoi introuvable.lists the typical energy consumption of some electrical devices; [7-9] [10-12] piezoelectric or triboelectric energy harvesting could maintain devices with power consumption up to 50 mW.Figure 1.Common electrical devices and their approximate power consumption. [7-9] [10-12]iboelectric energy harvesting devices or so called triboelectric nanogenerators (TENG) devices recently have gained momentum for mechanical energy harvesting to power autonomous microdevices and portable electronics.[5] In this process, both materials gain electric charge, equal in magnitude but opposite in polarity.
TENGs can be produced from cheap, lightweight, flexible, abundantly available polymer materials.In comparison to piezoelectric, [15][16][17] and ferroelectret, [18][19] materials, TENGs do not require costly electrical poling procedures to enable energy harvesting, although significant developments are overcoming the need for electrical poling in some piezoelectric polymer systems. [15,17] any approaches focus on improving TENG devices by engineering the electrostatic induction component of triboelectricity.That is, by moving faster (i.e.spinning) or further.However, methods to improve the contactelectrification side of TENGs has been relatively limited with three main routes dominating the field: i) modification of surface morphology; ii) modification of surface chemistry; and iii) incorporation of ferroelectric materials (dipole-dipole induction).These approaches, while powerful, need to be rooted in fundamental mechanistic understandings of the contact electrification to enable truly efficient TENG design.This review will shine light on such understandings as pathways to control the magnitude and polarity of surface charge via material transfer at the contact.

Triboelectrification -A Mechanistic Background
To understand the working mechanism of TENGs and the triboelectric effect in polymers, we must first understand triboelectric charging in general. In such a process, many things happen locally on the contacting surfaces, which include materials adhesion, deformation and fracturing, generation of heat, polarization, chemical bonds breaking, and light and acoustic emissions.Of course, these phenomena can vary a lot according to the physical and chemical properties, as well as the surface geometry (roughness or patterns).Some contacting electrification phenomena are well known.For example, the Volta effect due to the contact of two metals having differing work functions and the P-N junction between semiconductors.They have also gained wide application nowadays.However, this is not the case for other materials.A general theory of triboelectrification guides how charge is transferred in different materials and environments is still a puzzling issue.This is mainly due to multiple complicated surface phenomena, and the vast variety of materials that exhibit contact electrification.To do so, one needs to be clear about the following aspects: (1) the nature of surface electrical charges, (2) the origin of the charges, (3) the charge distribution state after the separation of two contacting materials, (4) the reliable characterization of their quantity on the surfaces, (5) the interaction of the charges with the environmental parameters such as light, heat, and moisture, (6) the charge evolution (disappearing or discharging) with time.
In the past two decades, the invention of TENG, which allows converting irregular mechanical energy into weak electricity power, has promoted worldwide research about this phenomenon.Subsequently, a large number of studies have been conducted in order to bring to light some fundamental interrogations about the triboelectrification or contact electrification mechanism.The studies include theoretical ones, numerical simulations, and experimental characterizations via more and more advanced instruments such as Kelvin probe force microscopy.The triboelectric charges are generally considered to be generated from electron transfer, ion transfer, or materials species transfer.Which mechanism is the dominant one?There is still no unanimous answer about this question.
In the following parts, some presentative research works are introduced, as well as main related arguments and observations.

Electron transfer
Electron transfer is commonly considered as the dominating charge carrier mechanism in contact-electrification, especially for solid-solid contact-pairs between metals and ceramics.For example, in a recently published review work, "On the origin of contact-electrification", Wang et al. defend such conclusion based on a number of experimental results obtained using Kelvin probe force microscopy, as well theoretical analyses. [24]They proposed that the electron transfer is caused by a strong electron cloud overlap between two atoms for the lowered interatomic potential barrier by shortening the bonding length.When the interatomic distance between two atoms belonging to two materials is shorter than the normal bonding length, the electrons will transfer at the interface.They considered that a strong electron cloud overlap (or wave function overlap) between the two atoms/molecules in the repulsive region will reduce the interatomic potential barrier, and result in electron transition between the atoms/molecules.Thus, they concluded that the principal role of contact (mechanical force) in the contact electrification process is to induce strong overlap between the electron clouds (or wave function in physics, bonding in chemistry). In this assumption, an important concept is the work function, which has been defined, in solid-state physics, as the minimum thermodynamic work (i.e., energy) needed to remove an electron from a solid to a point in the vacuum immediately outside the solid surface.It is worth noting that the work function is not a characteristic of a bulk material, but rather a property of the surface of the material (structure, defects, compositions).
In a case of metal-dielectric contact such as Pt-SiO 2 , Zhou et al. characterized the triboelectric charge distribution, from the multi-friction effect on charge transfer, as well as subsequent charge diffusion on the dielectric surface, by combining contact-mode atomic force microscopy (AFM) and scanning Kevin probe microscopy (SKPM). [25]They found that SiO 2 surface can be either positively charged through triboelectric process by repeatedly scanning a Si-based AFM probe with Pt coating.The triboelectric charges accumulated from friction and eventually reached to saturated concentrations of (−150 ± 8) μC/m 2 .Moreover, they further demonstrated that the polarity and magnitude of the charge transfer could be manipulated through an applied electric field between two materials, Pt coated atomic force microscopy tip and Parylene film. [26]A variation of bias voltage on the metal tip from -5 V to +5 V could switch the delivered charges from negative to positive, but this phenomena was not observed when the bias voltage was 3 to 4 V. Based on this observation, they deduced that it was electron transfer which dominated the contact electrification.The energy released by electron transfer from dielectric A to dielectric B can also be in the form of photon emission, plasmon excitation, and/or phonon excitation.Hird et al. reported that by repeating contact between silicone and a metal-loaded epoxy in vacuum, a source of 40 keV electrons could be generated, and it was capable of exciting the characteristic emission lines of Mo and Ag with characteristic x-rays at a rate of >10 5 per contact cycle.By increasing the repetition rate of the contact cycle the viability of a device that emits 10 8 xray photons per second is suggested. [29] has also been observed that contact between two identical insulators also produce static charges in the form of particulates of different sizes or film with different curvature. [28]Further verification of the above-mentioned trends was obtained under vacuum and higher temperature (≤358 K) conditions.
Another interesting phenomenon is the influence of the temperature on the transfer of charges across an interface.Lin et al., [30] by scanning an Au tip on a SiO 2 surface, revealed that hotter solids tend to receive positive triboelectric charges, while cooler solids tend to be negatively charged, which suggests that the temperature-difference-induced charge transfer can be attributed to the thermionic-emission effect, in which the electrons are thermally excited and transfer from a hotter surface to a cooler one.Further, a thermionic-emission band-structure model is proposed to describe the electron transfer between two solids at different temperatures.
The findings also suggest that contact electrification can occur when two identical materials are contacting due to a local temperature difference arising from the nanoscale rubbing of surfaces with different curvatures/roughness.
In the study, [31] friction was viewed as energy dissipation.And the energy dissipation or frictional heat conduction at the interface between two materials results in temperature variations in the materials and an abrupt temperature drop at the interface.The generated thermoelectric charges redistribute according to the temperature profile.The thermoelectric effect is the direct conversion of temperature differences to electric voltage and vice versa via a thermocouple.The Seebeck effect is the electromotive force (emf) that develops across two points of an electrically conducting material when there is a temperature difference between them.
Finally, the release of triboelectric charges studies has also been conducted.Xu et al., [27] demonstrated the temperature dependence of the surface charge density in CE by using a TENG made of a Ti and SiO 2 pair and Ti-Al 2 O 3 pair.It was found that the triboelectric charges decreased more rapidly at higher temperatures than at lower ones.It was shown the charge transfer followed an exponential decay at high temperatures for different TENGs, which was consistent with the theory of electron thermionic emission, instead of Boltzmann distribution followed by ion transfer.Moreover, it was found that a potential barrier existed at the surface that prevented the charges generated by contact electrification from flowing back to the solid where they were escaping from the surface after the contacting.Furthermore, an electron-cloud-potential-well model was proposed based on the electron-emission-dominated charge-transfer mechanism.A similar result has also been obtained by Cheng et al. [32]

Ion transfer
Contact electrifications concerning metals or semiconductors contacting can be evidently explained by work function properties.However, this explanation for polymer dielectrics is not so convincing.Thus, there remains vigorous debate around the mechanism of contact electrification and tribocharging in literature.
For understanding better this mystery, researchers like McCarty et al [33] suggested to separate polymers into two categories: nonionic insulating polymers and ion-containing ones.Here, an ionic electret is a material that has a net electrostatic charge due to a difference in the number of cationic and anionic charges in the material.
Due to the historical lack of tools to probe precise chemical and electrical surface properties at the subnanometer scale , many experimental observations have been used to deduce the nature of contact electrification.
For example, a proton-transfer mechanism has been proposed based on the observation that there exists a correlation between Lewis acid/base behaviour and contact electrification due to various ions (protons, hydroxide ions, alkali metal cations, halide anions, etc.) transfer from Lewis acid/base sites on one surface to Lewis acid/base sites on another surface [ref].In fact, it was as early as 1902 that Knoblauch had observed that solid organic acids tended to become negatively charged and solid organic bases tended to become positively charged when these powders were shaken from a piece of filter paper. [34]In their published work in 1953, Medley had also reported similar observations with acidic and basic ion-exchange [35] and a proton-transfer mechanism could explain the contact electrification of a wide range of insulating materials. [36]Proton-transfer is, of course, a specific example of the more general mechanism of ion transfer.
Whereas the ion-transfer mechanism leads to charge separation, other competing processes lead to charge recombination, or electrical discharge, of ionic electrets.These processes include tunnelling of electrons, field emission of electrons, and dielectric breakdown of the surrounding gas.The final charge on an ionic electret following contact and macroscopic separation depends on both charging and discharging processes.It is also the most difficult point to distinguish the contribution of the different processes in such kind of research.
Another reported observation is the asymmetric partitioning of hydroxide ions between adsorbed layers of water for contact involving nonionic, insulating materials.Like in many other physical or chemical processes, adsorbed water from the atmosphere can play a complicated role in the contact electrification. [33].It was also observed that the negative zeta potentials observed for nearly all organic polymers most likely result from this adsorption of hydroxide.This observation supports the hydroxide-adsorption model of contact electrification: more hydroxide would adsorb on polystyrene under basic conditions than under acidic conditions.Forssberg and co-workers observed a correlation between the zeta potentials of several inorganic minerals and the contact electrification of these minerals. [37]In addition, adsorbed water has been reported to increase the surface (ionic) conductivity of insulating solids.This may allow local regions of charge on a surface to spread and ultimately to discharge to ground.This might explain why it is difficult to observe electrostatic phenomena in very high humidity environments, due to rapid charge dissipation on the surfaces.
In a study of the effect of humidity on contact electrification of polymers containing bound ions and mobile counterions, [38] Diaz and co-workers observed almost no contact electrification at 0 % RH, maximum contact electrification around 30 % RH, and a decrease in contact electrification above 40 % RH.They also observed that the thickness of the adsorbed water layer (measured by ellipsometry) on these ion-containing polymers increased nearly linearly with relative humidity; the average thickness of the water layer at 90 % RH was around 0.4 nm, corresponding to about 1.5 monolayers of water.They thus proposed that the contact of two surfaces, each having an adsorbed film of water, yields a "water bridge" between the surfaces.

Mass Transfer
The triboelectric effect is also considered to be related to the phenomenon of adhesion, where two materials composed of different molecules tend to stick together because of attraction.In general, adhesion forces can be caused by multiple effects, including chemical adhesion, dispersive adhesion, and diffusive adhesion.In addition to the cumulative magnitudes of these intermolecular forces, there are also certain emergent mechanical effects. [39]Material transfer is by far most commonly discussed for polymers, therefore will be covered in detail in the subsequent section.

A Note on Measurements and Data Comparisons
Triboelectricity is a combination of 1) contact electrificationcharge generation at the contact-separation interface; and 2) electrostatic inductionthe movement of this charge with respect to external electrodes.While much of this review focusses on the mechanisms of contact electrification, towards increasing the amount of charge generated on a given surface, consideration of the role electrostatic induction has on the reported current, voltage, and charge measurements must be considered.
The measured electrical signals arise as described by Expanded Maxwell's equations. [40]There are many versions of these equations, [40][41][42][43][44] however these generally consist of an electromagnetic wave term and a polarisation term (Figure 2).It is important to note, that the change in polarisation ( ) and change in electric field ( ) terms, (where ε = dielectric permittivity; E = electric field; t = time; and P = polarisation) are directly related to the speed at which the two charged surfaces are moving with respect to their electrodes.Thus, when assessing the literature to compare voltage, current, or charge values, the speed at which triboelectric testing occurs will dramatically affect the reported electrical outputs.This is challenging as data reporting within the triboelectric generator field (as is also the case in the neighbouring piezoelectric field) [16] often lacks granular experimental detail (testing speed, force, frequency) that allows accurate comparisons to be made.For this reason, when assessing the effect of various factors in polymer triboelectric devices, we are comparing improvements within a dataset, not across adjacent publications where such key experimental information isn't comparable.
Figure 2. Displacement current (J D ), which underpins the measurement of triboelectric charging in materials, consists of two general terms (left) the electromagnetic wave term; and (right) the polarisation term (reproduced with permission from Wang, [42] under CC-BY-NC-ND, 2017).

Models of Polymer Triboelectrification -Electron vs Mass Transfer
Electron transfer adequately describes the charge transfer between metals, [14,24] therefore, the same logic was initially applied to insulators, both inorganic and organic.In contrast, insulators are characterized by a filled valence band and an empty conduction band separated by a comparatively large band gap; therefore, a significant amount of energy must be supplied to allow electron transfer (Figure 3 b). However, in the case of polymers, the 1D bonding network means that the creation of defects leads to fracturing of the polymer chain (Figure 3 e), meaning this approach cannot be directly used to enable electron transfer.
Figure 3. Simplified schematic of the electronic structure between different materials during contact; a) two metals, whereby an electron from the higher energy metal can be accommodated by the lower energy metal; b) a metal and an insulator, where there are no free states for an electron to full, therefore only by tunnelling can an electron transfer to the insulator (or alternatively through thermal excitation processes); and c) between a metal and a defected insulator, where the atomic defects leave available electronic states for electron transfer to occur.d & e) show schematics of d) ceramic; and e) polymer bonding networks; left) the pristine lattice; and right) with defects induced, in the ceramic network, the lattice is maintained through the multi-atom co-ordinated bonding networks, whereas in polymers the 1D bonding network is broken potentially leading to mass transfer.
The energy cost of directly stripping an electron from a polymer via this endothermic process is close to 10 eV, when considering electron removal from the first polymer material, charge separation over distance and addition to the second polymer material. [33]3] This energy is two times higher than the homolytic dissociation energy of the C−C bond (360−430 kJ mol −1 ). [47]is first ionization energy only accounts for the addition or removal of an electron from the polymer, not the subsequent transfer onto the opposing surface.The summary of energies in this multiple-step process have been estimated to be as large as 1 MJ mol −1 , which is 3 times higher compared to the energy needed for covalent break. [47]rect electron transfer between insulators is energetically more improbable than between metals. Several attempts have been made to address the flaws of electron transfer model for insulator contact electrification.The existence of intermediate states for electrons in the bandgap, which are provided by defects has been proposed. [66]Thermoluminescence experiments have shown the existence of such states, however, quantification of these electrons had demonstrated that this effect is insufficient to explain the measured surface charge. [67][70][71] Accordingly, the contact of both surfaces allows hybridization of their electron states so that some electrons are delocalized and become available for transfer. [27]This electron delocalisation or cloud model, is particularly challenging from a polymer perspective, where the high degree of chain conformation and available conformer-states means repulsion and chain movement may occur prior to electron cloud overlap.
A subtle distinction between the electron cloud model, [24] and the mass transfer model of contact electrification is that in the electron cloud model, charge transfer occurs during contact and compressionwhich is extremely energetically unfavourable due to atomic repulsive forces.In contrast, in mass transfer, charge transfer occurs during separation, and thus electron orbital overlap isn't required, simply bond cleavage.
However, can ions be responsible for the strong triboelectrification when conventional polymers are contacted?As there are no "mobile" ions in conventional polymer insulators, the hydroxide ions (OH -) are thought to play the central role, because even hydrophobic polymers adsorb a thin layer of water on their surface in ambient conditions. [76]When polymer surfaces are contacted, a water "bridge" forms on the surfaces enabling the transfer of OH -ions to the polymer, which exhibits stronger adsorption.When materials are separated, a net charge is formed on the surface due to ion imbalance in the layers of adsorbed water.ndeed, it has been shown that polymer adhesion plays a key role in contact electrification, and these adhesive processes cause bond cleavage due to a range of factors including cross-linking, chain entanglement, network topology, and non-covalent domain interactions (Figure 4 a). [83]Simple modelling of polymers undergoing vertical separation show polymer strands forming that eventually cleave at a given strain or distance (Figure 4 b). [83]gure 4. Mechanisms of polymer adhesion; a) bonding types formed during contact with polymer surfaces; and b) coursegrained simulations of separation across a rubbery polymer film (numbers represent strain values in the z-direction) (reproduced with permission by CC-BY from Raos and Zappone, [83] ) This bond cleavage occurs due to the rearrangement of polymer chains during contact and suggests that the triboelectrification mechanism in polymers is from the transfer of the cleaved and charged polymer fragments. [84- 85] he energy of covalent bonds in polymers range from 3.7 eV to 4.5 eV depending on the presence of heteroatoms (N, Si, O, S, F, etc.) in the main chain and side moieties. [86] After homolytic scission, identical radicals are formed on the ends of each chain of the ruptured macromolecule, [39] but heterolytic scission is responsible for the charge formation, since the outcome is a pair of cation and anions (mechano-ions).As observed by Kelvin force microscopy (KFM), after contact-separation surfaces develop random patterns of positively and negatively charged nanoscale domains (Figure 5). [73]Accordingly, the net charge density measured on each contacted surface is the sum of contributions from these nanoscale domains, with the dominant mechano-ions determining the polarity.As surfaces are separated, a material transfer happens along with the charged organic ionic fragments. [73,89] Figure 5. Schematic of a) the formation of a charge mosaic upon contact separation of polymers (reproduced with permission from Baytekin et al., copyright Science AAAS 2011) [73] ; and b) the bond entanglement and cleavage that leads to the observed mosaic.
The revelation of a surface charge mosaic is critical to understanding charging between polymers.It inherently complicates our understanding of interactions between charged surfaces, as now we are not describing two opposingly charged plates oscillating near each other.Instead, we are describing micro-or nano-scale domains of charge, that interact with neighbouring opposing charges on the same surface, as well as on the opposing surface.How these charges are stabilised, and how they can be manipulated at a molecular level are key questions for the field still to address.
This means that while mass transfer is thought to be the dominant effect in polymer-polymer contact electrification, the role of ions and electron transfer between transferred mechano-radicals and mechano-ions cannot be discounted.Thus the holistic mechanism of polymer contact-electrification is complex, particularly for secondary and subsequent contact-separation cycles (Figure 6).

Mass Transfer -Increasing Surface Charge
Mass transfer and the surface charge mosaic also presents two key tools that can be used to enhance charging from a polymer-polymer contact-separation experiment; 1) the approach to increasing the amount of (heterolytic) bond breakage; and 2) the development of surfaces that have a greater affinity for either positive or negative charges.Both these tools will enable an increased charge output from triboelectric devices.

Bond Cleavage and Heterolysis
The probability of a polymer backbone bond breaking heterolytically or homolitically is related to the chemistry of the polymer, both the backbone, and their functional groups. For most other polymers with a carbon-carbon backbone there is minimal driving force for heterolytic bond cleavage.This simple effect may explain why PDMS is so often reported with much higher triboelectric surface charge values compared to other polymer systems. [92]us, pathways to increase the ionic nature of carbon-carbon backbone bonds may provide steps forward to enable higher triboelectric charge generation from non-siloxane polymers.Kamiyama et al., demonstrated a clear correlation in polymer produced from substituted co-monomers.They revealed the polarity and magnitude of the surface charge varied significantly, ranging from -120 μC g -1 to +130 μC g -1 simply based on the monomer. [58]Zhang et al., used functional groups ordered by Lewis Basicity to describe triboelectric charging from polymers. [56]They observed a correlation between the polarity observed on a polymer and its related Lewis Basicity, and proposed that this is a simple predictor of triboelectric charging. [56]These observations match those of Diaz and Felix-Navarro that revealed a correlation between the pK b of a polymer and its charge (by the relationship of log (q+1)). [93]More recently, a detailed study by Li et al., demonstrated a direct correlation between triboelectric charging in polymers and the presence of electron withdrawing side-groups on the backbone. [57]Li et al, attributed this effect to changes in the electron cloud range, [57] however, such an addition of an electron withdrawing group will also increase the ionicity of the polymer backbone, leading to increased heterolytic bond cleavage.Thus the mass transfer mechanism cannot be discounted.

Increasing Bond Cleavage
The interactions that occur due to polymer chain motion during contact have been well described for decades. [39] logically follows that increasing bond cleavage, and hence material transfer, will occur due to increased adhesion.At the polymer | polymer interface, locally concentrated physical adhesive bonds form, and if the sum of the interaction forces overcomes the strength of chemical covalent bond, bond cleavage will occur. [47,83] hly deformable polymers such as elastomers provide a fuller contact area and increase the density of intermolecular adhesion bonds during contact.Increased adhesion also increases the local friction, which directly contributes to bond scission by creating shear force during contact.Generally, friction depends on the molecular weight and crosslinking density; for example, it has been shown that crosslinked polymer surfaces exhibit several orders of magnitude lower friction than non-crosslinked. [39]The increased friction also leads to greater localised heating, which may promote heterolytic bond cleavage, however the authors note that further studies are needed to validate this point.Bond heterolysis alone will not guarantee significant surface charge, as there is no intrinsic driving force for one surface to charge positively and one negatively.While homolysis does not yield charge species, it is still beneficial to contact-electrification of polymers.It has been shown that radicals colocalize with the ions and stabilize them on the contacted polymer surfaces. [94]The addition of radical-scavenging molecules (e.g., Evitamin) rapidly decreases the previously stable charge density in a range of polymers. [95]According to theoretical studies, the radical-charge stabilization mechanism is based on the formation of intermolecular oddelectron (one or three) bond. [94]e following section will focus on how charge can be increased by control of polymer mechanical properties.

Influence of Deformative Properties
Materials that are softer tend to gain larger surface charge than harder materials.This is not only because softer polymers form tighter contact but also because soft polymers are more prone to mechanical damage and bond cleavage. [96]The softness and hardness (deformative properties) of the polymer can be expressed with the cohesion energy (E coh )the energy required to pull out the molecular species from bulk.Polymers with higher E coh are represented by higher elastic modulus. [96]gure 7. a) Youngs modulus against cohesive energy density for various small organic molecules and polymers (reproduced with permission from Roberts, Rowe, and York, [96] copyright Elsevier 1991); b) Correlation between the modulus of polymer material and surface charge from contact with ITO; and c) correlation of difference in hardness between two polymers and surface charge (b, c, reproduced with permission from Šutka et al., [46] copyright Royal Society of Chemistry, 2019) Understanding this concept enables the application of other approaches that increase polymer chain mobility, or polymer compressibility to increase charge.

Influence of Thermal Treatment
One such tool that can increase polymer chain mobility is heat.The charge from contact electrification was shown to increase by over an order of magnitude when glassy polymers such as polystyrene (PS), polycarbonate (PC) and poly(methyl methacrylate) (PMMA) underwent a phase transfer from glassy to rubbery state (Figure 8 a, b). [97]Increasing the temperature below the glass transition temperature, T G , only led to a small increase in charge.For example, heating a PC sample from 25 °C to 110 °C, the surface charge only increased from 0.006 to 0.011 nC cm -2 .However, when a polymer is heated above T G the surface charge increases drastically by more than an order of magnitude (Figure 8 b).The phase transition from a glassy to a rubbery state is accompanied with a decrease of cohesive energy and increase in the surface adhesion.Above T G , polymers behave like elastomersthey are soft and exhibit large reversible deformations., [97] copyright Royal Society of Chemistry, 2020), and c) influence of thermal history on charge from contact-separation experiments, blue = no thermal treatment; red = thermal treatment at 130 °C for 60 minutes (reproduced with permission from Šutka et al., [46] copyright RSC).
The influence of heat on changing the macromolecular ordering within polymers, thus altering the intramolecular bonds formed between contacting polymers (as described in Figure 4), was studied (Figure 8).This showed that, similar to contacting a hard and soft polymer pair, contacting polymers of the same chemical composition with different thermal history demonstrated significantly enhanced charging, attributed changes in the polymers intramolecular structure. [46]

Influence of Fillers and Chain Mobility
Another pathway to change the chain mobility with polymer films is to add rigid nanoparticles.These rigid nanoparticles serve two purposes, 1) adding heterogeneity into the deformability of polymer composites, and 2) acting as stress concentration sites during contact-separation experiments.
The addition of inert inorganic nanoparticle fillers to enhanced contact charging was demonstrated by Chen et al (Figure 9 a). [98]They demonstrated that the addition of 10% of SrTiO 3 particles led to a 5-fold increase in the power density of the polymer film, which was stable over 15000 cycles.The authors attributed the improvement to an increase in the dielectric permittivity of the sample, and did not consider the change in deformative properties of the polymer by the addition of the SrTiO 3 and addition of voids into the structure.Nevertheless, the results are compelling for observing the effects of additives in polymer triboelectric surfaces. [98]igure 9. Influence of adding rigid fillers into polymers; a) schematic of added SrTiO 3 nanoparticles and voids into a PDMS film and corresponding increase in current density (reproduced with permission from Chen et al. [98] copyright American Chemical Society, 2015); b) the effect of adding Titanium nanosheets into a PDMS matrix on charge and voltage outputs from a triboelectric nanogenerator (reproduced with permission from Sriphan et al., [99] copyright American Chemical Society 2019); and c) demonstration on the increasing charge density when contacting polymer surfaces with different amounts of rigid additives (reproduced with permission from Lapčinskis et al., [100] copyright Royal Society of Chemistry, 2021).
Sriphan et al. also studied the influence of particle addition into a PDMS material, whereby BaTiO 3 piezoelectric nanoparticles, titanium nanosheets, and silver nanoparticles were used to enhance charge and voltage outputs. [99]e addition of Ti nanosheets (Figure 9 b) or Ag nanoparticles was shown to dramatically increase the measured charge, far beyond the simple addition of the piezoelectric BaTiO 3 particles.The team attributed this increase in charge density to an increased friction at the interface by adding the fillers, which boosts the contactelectrification effect.In their devices this enhanced contact-electrification is coupled to the piezoelectric effect to lead to a synergistic increase in charge output. [99] This is further challenged by the exemplar triboelectric polymer PDMS, [92] where added fillers can adsorb cross-linking agent softening the PDMS as opposed to local stiffening in other polymer systems.Lapčinskis et al. describe the addition of non-piezoelectric TiO 2 particles into a ethylene-vinyl acetate copolymer (EVA) matrix.It was shown that by contacting two EVA polymers together with different TiO 2 loadings, the charge generated followed an approximately linear trend with the difference in TiO 2 loadings between the samples (Figure 9 c). [100]When a pristine EVA layer contacted the EVA-TiO 2 composite containing 5 vol% of TiO 2 , a 24-times higher current, 4 times higher voltage, and 13 times increase in surface charge density (from 0.007 nC cm -2 to 0.093 nC cm -2 ) was observed when compared to contacting two pristine EVA layers together.This result suggests that the change in charge build up with nanoparticle addition should arise from altered chain mobility and stress inhomogeneity effects, as described above.If dielectric permittivity was the primary driver for the increased charge generation, the measured charge should scale with the total amount of TiO 2 present, not the difference between the two polymers.The effect was observed in tests of various filler nanoparticles, including WO 3 , FeO(OH) and MnO 2 .Both mechanical experiments and finite element analysis simulations demonstrated that the triboelectrification observed is related to differences in mechanical properties and uneven deformation during mechanical contact and separation. [100]gure 10.The influence of macromolecular bonding and structure on contact electrification; a) PEBA/α-FeOOH; b) schematic of the PEBA/α-FeOOH interaction; c) hydrogen bonding regimes potentially leading to local stiffening of the PEBA matrix; d) correlation of TG, hardness, and H-bonding with surface charge; e) Nanoindentation force maps of surface hardness showing the introduction of surface heterogeneity between 0.2 and 0.5 vol% α-FeOOH corresponding to the increased surface charge in (d) (reproduced with permission from Šutka et al., [101] copyright Wiley-VCH 2022) The ordered macromolecular domains in soft elastomeric matrix on the same surface also can enhance charge formation and material transfer.For that purpose, -FeOOH was added into an elastomeric polyether block amide (PEBA) matrix (Figure 10 a). [101]The addition of the -FeOOH led to formation of hydrogen bonding regions between the ether and amide polymer blocks (Figure 10 b, c), which in turn led to formation of domains with altered degree of macromolecular ordering and hardness (Figure 10 d).The size of macromolecular domains exceeds the size of individual nanofiller filler by orders of magnitude to micro-scale.Interestingly, it was shown that the continued increase in the concentration of -FeOOH, above 1 vol% led to a decrease in the surface charge measured during triboelectric testing, with this decrease correlating to a decreased amount of hydrogen bonding between the polymer and -FeOOH.This decrease arises from increasing interactions between separated -FeOOH particles which screen the polymer--FeOOH h-bonding.Nanoindentation maps show that the change in surface charge is directly correlated to heterogeneity in surface hardness, arising from these hydrogen bonding networks (Figure 10 d). [101]Finite element method modelling revealed that the irregularity in hardness creates increased stress accumulation on the polymer surface during contact-separation.
Molecular dynamic simulations demonstrated that this stress accumulation, in turn, reduces the energy for heterolytic bond cleavage.Such a hydrogen bonding network mimics the mechanical structure of spider silk, and the observations of surface charging in these systems may suggest a mechanism of spider ballooning.
These macromolecular effects are critical to understanding the compressibility and stress concentration within polymers.However, for triboelectrification to occur via mass transfer heterolytic bond cleavage must occur. This surface adhesion greatly depends on polymer surface chemistry, such as availability of hydrogen or van der Waals bonds, solubility, surface topography (roughness), molecular weight, and the force applied during contact.
Li et al., demonstrated that a dramatic enhancement of triboelectric charge was observed in polymer TEGs by a helium ion implantation approach. [104]Ion implantation of helium ions into a range of polymers including Kapton, PET, PTFE, and FEP revealed a change in voltage outputs when contacted with an Aluminium film.
The team ascribed the observed improvements and changes in output to alternation of the dielectric permittivity of the samples, however, other factors including altered mechanical properties and surface generated electric fields may alter the probability for cation or anion mass transfer to ion irradiated surfaces.Further, the doping of the polymers with helium ions enables new electronic states around the sites of implantation that may participate in the electron transfer mechanism. [104]2.6.

Influence of surface adhesion
It is logical to consider that if mass transfer is the primary mechanism for polymer contact electrification, the force required to separate two polymer surfaces should play a key role in heterolytic bond cleavage and triboelectrification.hori et al., demonstrated that increased adhesion between polymeric pharmaceuticals was directly proportional to their triboelectrification (Figure 11 a). [107]While they demonstrated the link between this triboelectrification and adhesion across particle size, their approach of powder induced triboelectrification and adhesion measurements by mass loss means that it cannot be concluded if polymer adhesion increased triboelectrification or if triboelectric charge increased subsequent adhesion. [107]ang et al., performed careful studies on the adhesion and separation of PDMS films with different structures. [108]They showed that the rate of peeling PDMS from the stainless steel surface was directly related to the force of adhesion measured between materials.It was revealed that the current generated from peeling experiments broadly related to the adhesive force and the speed of peeling (Figure 11 b).However, their key result was that varying the force of adhesion in a single 'peel' motion, achieved by peeling PDMS off patterned substrates, increased the current generated from 18 nA to over 80 nA. [108]This was ascribed to a change in the mechanism of delamination from a smooth contact motion, to a 'jumping' motion. [108]However, this result can also be considered by increased stress localisation at the high-low adhesion interface leading to increased bond cleavage and mass transfer.
Zhang et al., [111] demonstrated that in contact between Si/polytetrafluoroethylene (PTFE) and Si/polyvinylchloride (PVC), the softer and more adhesive PVC has shown more substantial material transfer to a-Si. [111]It is proposed that on sticky polymer surfaces the energy of formed adhesive (physical) bonds between contacting surfaces exceed the energy of chemical or/and physical bonding in the bulk, thus enhancing the probability of covalent bond scission and material transfer between the two contacting surfaces.
Adhesion and particularly friction is highly dependent on the molecular weight and the nature of intramolecular bonds within a polymer material.The friction on cross-linked surfaces is known to be several orders of magnitude lower than on non-crosslinked surfaces. [39]The adhesion forces also dramatically increase with dangling bonds, as the polymer chains are free to interact and entangle on the contacted surfaces.The effect of this adhesion force from cross-links was studied by Šutka et al., [46] where the molecular weight between crosslinks, M C (decreasing the crosslinking degree of PDMS) from 2000 g mol -1 to 13000 g mol -1 , the separation stress increases from 1.1 to 3.5 N cm -2 (Figure 11 c).The increase in separation stress, was found to correlate to an increase in the surface charge density on PDMS from 0.3 to 3.3 nC cm -2 (Figure 11 c). [46]igure 11.Influence of polymer adhesion on triboelectrification; a) triboelectric mass-to-charge ratio of Theophylline particles vs their adhesion to a steel canister (reproduced with permission By CC-BY license from Ghori et al.) [107] ; b) the influence of separation speed on the adhesion force, and the generated triboelectric current for a PDMS-Steel interface (reproduced with permission from Zhang et al., [108] copyright Elsevier, 2022); c) surface charge density and separation stress dependence on the mass between cross-links, M c (reproduced with permission from Šutka et al., [46] copyright RSC); and d) the separation stress and surface charge for different polymer materials during contact-separation experiments with ITO (reproduced with permission from Lapčinskis et al., [110] copyright Wiley, 2019).
The effect of separation stress was extended into different polymer systems, and a similar correlation and trend was observed for polydimethylsiloxane (PDMS), ethylene-octene copolymer (EOC), styrene-ethylene-butylenestyrene copolymer (SEBS), and polyethylene-covinyl acetate copolymer (EVA) (Figure 11 d). [110]The separation stress of tested materials increases in the order: PHC > SEBS > PDMS > EOC > EVA.Polymers that exhibit stronger adhesion against the contacted surface also show larger surface charge.PHC shows three orders of magnitude higher surface charge (8.853 nC cm −2 ) than the EVA (0.013 nC cm −2 ), where EVA has only 10% of the adhesion of the PHC surface. [110]e combination of these works demonstrates the importance of considering adhesion between surfaces for engineering and enhancing polymer contact electrification.

Topography
The factors discussed so far, adhesion, macromolecular ordering, fillers, and deformability, have been probed by isolated testing models to elucidate their contributions.However, these factors are also interconnected (as discussed above around the role of fillers).Perhaps one of the biggest examples of the interconnected nature of these factors is the role that topography plays in polymer contact electrification.Simply put, as contactelectrification and triboelectricity is a surface effect, increasing the contacting surface area should directly increase the charge generated. In PDMS (and other elastomers), it is one of the most common tools used to engineer a higher surface charge from a triboelectric generator. [92]rious methods to increase the specific surface of polymers area have been reported such as physical modification using templates or molds, [3,[114][115] block copolymer templating, [116][117][118] electrospinning, [119] direct laser printing, [120][121] precipitation, [122] reactive ion etching, [123] or chemical etching. [124]For example, silk fibroin electrospun fibers showed a 1.5 times higher peak voltage when compared to a flat silk surface during contactseparation (Figure 12 a). [125]PDMS that was patterned to have different surface geometries, including linear, cubic, and pyramidal, showed a clear increase in voltage generated from contact electrification in the order of: flat film, line, cube, and then pyramid surface (Figure 12). [3]The maximum output voltage for pyramid-patterned PDMS film was four times higher than the flat PDMS film.Tantraviwat et al. made a highly porous PDMS layer modified with a nanograss silicon mold, with the microstructured PDMS film generating twice the voltage of the flat PDMS film, owing to the improved effective contact area. [126]igure 12.The role of topography on contact electrification.a) charge generated from electrospun silk fibres against cast silk (reproduced with permission from Kim et al., [125] copyright Wiley, 2016); b) the influence of PDMS surface topography of generated voltage (reproduced with permission from Fan et al., [3] copyright ACS, 2012); c) heterogenous surface topography of PDMS induced by a nanograss mold and its influence on contact-electrification voltage (reproduced with permission, Tantraviwat et al., [126] copyright Elsevier, 2020); d) the influence of topography on surface charge density for a range of different polymers (in contact separation between chemically identical polymers); and e) the charge density as a function of roughness difference in contact between polymer surfaces (d, e reproduced with permission from, Šutka et al, [127] copyright Royal Society of Chemistry, 2020).
Contact charging has also been observed in contact between two chemically identical polymers with different surface roughness. [127]It was demonstrated that polymers with a greater porosity or surface roughness differential result in a greater generation of surface charge (Figure 12 d, e).Films of the same polymer with different roughness on contact-separation can generate an order of magnitude larger peak voltage than contacting-separating two smooth or two rough polymer films (Figure 12 d).The surface charge at the contactseparation of identical polymer smooth-smooth, porous-smooth and porous-porous films was measured for various materials: polyethylene-octene copolymer (EOC), polystyrene (PS), ethyl cellulose (EC), polycarbonate (PC), poly(methyl methacrylate) (PMMA), styrene-ethylene-butylene-styrene copolymer (SEBS) and poly(vinylidene fluoride) (PVDF).For all tested materials, a significantly higher surface charges were measured when the smooth polymer film was contacted with the porous one (Figure 12 d).
The reason for the formation of surface charge between two identical polymers with different roughnesses is the uneven surface deformation in the lateral direction, which apparently causes the friction, as well as uneven bidirectional material transfer.The uneven deformation at the contacting interface between two identical polymers with different roughness was confirmed by developing a 2D plane finite element model for a macroscale simulation. [90]When the rough polymer is pressed against a smooth polymer, a different surface length percentage change in lateral direction for two surfaces can be observed.In contrast, when two surfaces with similar roughness are contacted, the observed deformation in lateral direction for both surfaces is almost the same.
While beyond the scope of this review, Waitukaitis et al., have developed a model that describes chemically identical material contact electrification or tribocharging by describing positive and negative nanoscale domains as donors and acceptors. [45]This is an elegant solution that gets around this issue of what charges are transferred and can provide key guidance in the development of new triboelectric systems that exploit same material tribocharging.
The previous sections have summarised how mechanical properties, including deformability, heterogeneity, additives, and topography can influence triboelectric charging and contact electrification.However, these effects have only focussed on the magnitude of charge generated.The subsequent section will focus on new understandings on why surface charges form with a given polarity, and how this can be engineered towards more effective triboelectric nanogenerators.

Mass Transfer -Controlling Polarity
In the previous section we discussed approaches to increase surface charge, by increasing adhesion, roughness, porosity, stress accumulation heterogeneity, and bond cohesion.In the next section will focus on how we can use these same phenomena to control the polarity of charge that assembles on a given surface, towards complete control over polymer triboelectrification mechanisms.

Polymer Chemistry & Bond Energy
We have discussed above how soft polymers in contact with hard polymers tends to generate the most surface charge.By measuring the charge from both polymer surfaces (Figure 13 a), from 196 different polymer combinations (14 x 14), the effect this contact has on surface polarity has been extracted. [47]Ordering each of the 14 polymers tested from the lowest E coh to the highest E coh to produce a matrix revealed a clear visual correlation between E coh and charge polarity (Figure 13 b).When contacting soft versus hard polymer, the softer polymer tended to gain negative charge, while the harder polymer tended to gain a positive charge.Statistical analysis of this data showed that knowledge of a polymer's cohesive energy density only could predict the polarity of surface charge with ∼90% accuracy. [47]igure 13.Elucidating the effect of cohesive energy density on contact electrification; a) method for extracting polarity during vertical contact separation experiments between two polymer films; b) measured surface charge density of each polymer in 196 polymer contact-pairs ordered by the polymers cohesive energy density; c) method of measuring contact electrification between a polymer and the liquid metal Galistan; d) wetting of Galistan on PTFE films after 0 to 10 immersions in Galistan; and e) charge density measured from Galistan-polymer contacts correlated to the cohesive energy density (reproduced with permission from Sherrell, et al., [47] copyright the American Chemical Society, 2021).
This understand enabled Sherrell et al., to propose the following basic guidelines: 1) If ∑F adh > E coh (polymer A) and ∑F adh < E coh (polymer B), the polymer A will be transferred to polymer B.
2) If ∑F adh < E coh (polymer A) and ∑F adh < E coh (polymer B), material transfer will not occur for either polymer.
3) If ∑F adh > E coh (polymer A) and ∑F adh > E coh (polymer B), bidirectional material transfer will occur for both polymers A and B, resulting in a complex mosaic of charged surfaces. [73]

Surface Roughness & Topography
While the chemical structure of polymer is clearly important for their contact electrification, the physical surface topography, and deformability is also critical for a holistic view.
In order to understand the role that surface roughness and topography have on surface polarity from contact electrification, Verners et al., contacted chemically identical polymers with different roughness and topography and measured the surface charge obtained on each polymer (Figure 14 a). [90]Verners' experimental results directly provided evidence that a smooth surface will obtain a negative polarity, and a rough surface a positive polarity (Figure 14 b).
Figure 14.Probing the effect of roughness of polymer surface charge; a) schematic for testing rough and smooth, chemically identical polymers; b) measured surface charge density from (a); and c) using this understanding coupled to cohesive energy denstiy to enhance surface charge three fold (reproduced with permission, Verners, et al., [90] copyright Elsevier 2022) Through finite element analysis and molecular dynamics these effects were explained by stress localisation and differences in the energy of adsorption/desorption for positively and negatively charged mechano-ions.By combining a soft and smooth polymer (PVDF) with a hard and rough polymer (PS), that is combining the polarity effects from cohesive energy density and surface roughness, Verners et al, demonstrated a threefold increase in surface charge by exploiting this synergistic effect.

Towards Engineering Surfaces for Enhanced Triboelectrification
In polymer-polymer contact the material transfer is bidirectionalboth materials are transferred to opposite surface upon contact-separation. [72]Thus, we can use the developed understandings above to generate more heterolytic bond cleavage, and control cohesive energy density and roughness to induce charge separation for enhanced triboelectric outputs.
Table 1.Summary of the effects described in section 3 and their influence on polymer charging.

Parameter Effect Magnitude References
Compressive modulus Soft polymers will charge more than hard polymers in contact with ITO or a metal Up to 20x [46]   Hard vs Soft Contact Contacting a soft polymer with a hard polymer will generate more charge than soft-soft or hard-hard Up to 10x [46]   Temperature -Glass

Heating polymers above their TG dramatically increases their generated charge
Up to 20x [97]  Nanoparticle Fillers

Increased heterogeneity of stress concentration increases the charge generation with increased fillers
Up to 10x

Cohesive Energy Density Lower (soft) cohesive energy density polmyers tend to charge negatively
Polarity: 90% correllation [47]   One of the key drawbacks to generating a large electrical output from a TEG is that the generation of charge only occurs at the contact interface.This means that the charge generated is inherently 2D, yet exists on a material with defined thickness, and therefore the majority of the contacting material does not contribute to contact electrification.One pathway to dramatically increase the generated charge is to introduce charge generation within a material, either as surface charge, or complementary piezoelectric dipoles.

Dipole Templating
One of the approaches to achieve greater TEG performance involves the use of ferroelectric effect.
Ferroelectricity is commonly observed in perovskites (BaTiO 3 , PbTiO 3 ) and some polymer materials (polyvinylidene fluoride (PVDF), polyamide-11 (PA-11)). [128]Ferroelectric materials have randomly oriented dipole moments of their domains in the bulk, leading to a polarization density P = 0.An electrical field E aligns dipoles or domains in a specific direction, thus creating a net polarization vector.An applied stress causes change in P, which induces an electric field in nearby electrodes.Such TEG devices have shown the highest output performance among all, reaching state-of-the-art power density -50 mW cm -2 . [129]though, many papers report pathways to enhance the performance of triboelectric devices based on ferroelectric polymers, the proposed mechanisms for enhancement are diverse.Polarized ferroelectric layers have been considered to have a more appropriate work function for electron transfer between polymer layers. [130- 131] owever, it was convincingly demonstrated that the enhancement is observed due to induction driven by the piezoelectric charges. [132]The piezoelectric charges created in the ferroelectric bulk during contact contributes to electrostatic induction together with surface charges and thus enhances the overall performance of triboelectric generator. [133]he performance of triboelectric generator devices can be significantly improved by contacting two polarized ferroelectric of polyvinylidene difluoride (PVDF), where the polarisation of each film is inverted (Figure 15 a, b). [132]When non-polarised PVDF was contact-separated with +/-polarised PVDF the open circuit voltage 71 V and 67 V in the case of PVDF ''-'' and PVDF ''+'' was measured.Almost the sum of these output voltages 119 V was observed when contact-separating inversely polarised PVDF, which is PVDF ''-'' versus PVDF ''+''.A voltage of ~15 V was generated when PVDF films with the same polarity were tested. [132] [132] copyright Royal Society of Chemistry, 2018); and c) the double capacitor model and corresponding contact− separation stages. At fist, ferroelectric layers contact and form a connected capacitor, and then layers are compressed which results in a negative piezoelectric response.Next, the films are unloaded and separated resulting in a positive piezoelectric response, thus disconnecting the capacitors that had formed the connected circuit (c, reproduced with permission from Lapčinskis et al., [134] copyright American Chemical Society, 2019).
A clear correlation between the piezoelectric response and output of TENG device based on ferroelectric PVDF/BaTiO 3 nanocomposite films was demonstrated also by Lapčinskis et al. [134] For example, the triboelectric generator device based on inversely polarized PVDF with an intrinsic piezoelectric output voltage 0.5 V generated a V OC value of 250 V, while the generator from the BaTiO 3 (25vol%)/PVDF nanocomposite with an intrinsic piezoelectric output voltage 3.5 V generated a V OC value of 2700 V from the same 5 cm 2 contacting area.
To explain the enhancement in triboelectric generator devices based on inversely polarised ferroelectric contacting layers, a double capacitor model has been proposed (Figure 15 c). [134]Each ferroelectric layer can be considered as an individual capacitor with the charge density σ induced on each surface plane during polarization.The total capacitance and potential difference depend on the separation gap.During contact, the ferroelectric polymer layers form a circuit of two identical capacitors in series, and the total capacitance of two layers is decreasing; however, the potential difference on electrode plates is increasing in accordance with the equation: Compression then induces piezoelectric charge formation as the internal dipoles deform.The fast separation of the layers during triboelectric testing means that polarization quickly rises and the charge density returns to initial value.
The direction of the ferroelectric dipoles also influences the distribution of mechano-ions from heterolytic bond cleavage on surface, and the resultant electrical outputs are dependent on the alignment of these dipoles with the propensity of the contact surface to charge positively or negatively through triboelectrification. [133] By matching the surface charge polarity to ferroelectric dipole orientation, Lapčinskis et al., produced up to four times higher energy and power density.A triboelectric generator device based on a polyvinylacetate/BaTiO 3 nanoparticle composite showed an energy density increase from 32.4 mJ m -2 to 132.9 mJ m -2 when comparing devices with matched and mismatched arrangements of surface charge and dipole polarity.
By combining ferroelectric phenomena with engineered surfaces, that is surface additives to increase stiffness or to alter surface roughness, ferroelectric-polymer composites can be produced and tuned to achieve exceptional voltage and current outputs.
While introducing piezoelectric materials into polymer TEGs is a promising approach, this limits material selection quite significantly.In particular, fluoropolmyers present environmental hazard that are becoming more prominent.Therefore, alternative routes that avoid environmentally harmful fluoropolymer and inorganic crystals but still increase triboelectric performance are highly desired.

Engineering Internal Triboelectric Interfaces
To achieve TENGs that outperform ferroelectric materials requires the generation of surface charges in the volume of the polymer films.However, in order for these to be effective surfaces, there needs to be nano-, micro-, and macroscale ordering of the charges produced by such contact surfaces.In order to do this, it is imperative to not only control the polarity generated at a given contact interface in volume, but also where friction and slip occur within a material.
Malnieks et al., [135] provided the first insight in how to achieve such control.They produced PDMS films with different mechanical properties by varying the cross-linking density (polymer:crosslinking agent ratio, 3:1 and 10:1).These films were produced either individually, or as bilayers, where the second PDMS layer was crosslinked on top of the first.This bilayer structure had a fixed interface, and thus no slip could occur between the two components.The bilayers could then be assembled into a laminate structure, where slip only occurred at the interfaces between bilayers (Figure 16 a, b).Ordering the assembled bilayers an order of magnitude increase in triboelectric current and voltage was observed in contact separation mode, compared to a TEG made from single layers of the PDMS components (Figure 16 c). [135]gure 16.Approaches to induce volumetric surface charge into triboelectric devices; a-c) the use of different stiffness PDMS layers; a) schematic of the tested assemblies with aligned and misaligned bilayers; and single layers; b) photo of the assembled PDMS films; and c) current and voltage outputs from testing the films via contact-separation (a-c reproduced with permission from Malnieks et al., [135] copyright Elsevier 2022); and d-e) the use of electrospinning to dramatically increase the surface area and add roughness effects into the internal surface charges; d) scanning electron micrograph of the laminate cross-section (inset: photo of the laminate film); and e) areal charge density from an individual laminate; a 30 bilayer laminate based TEG; and a piezoelectric polymer (d-e, reproduced with permission from Sutka et al., [136] by CC-BY copyright Wiley 2022).
Linarts et al., extended this approach and increased the number of contact interfaces by producing high surface area, electrospun fibre, laminates from a low cohesive energy density polymer (EVA) and a higher cohesive energy density polymer (PLA) (Figure 16 d). [136]The diameter of the EVA fibres was an order of magnitude larger than the PLA fibres such that the EVA acted as a smooth surface, and the PLA a rough surface.Thus, according to mechanistic understanding, the EVA will charge negatively and the PLA positively.In this case, the slip interface was controlled using the diameter of the electrospun fibres, when the small PLA fibres were spun onto an EVA fibre mat, the small fibres could fill the pores of the EVA and formed a non-slip interface.
However, when EVA fibres were spun on a PLA fibre mat, the EVA could not penetrate the PLA pores, and thus significantly greater slip was observed at this interface.Thus, by electrospinning the EVA/PLA layers sequentially ordered internal surfaces were formed that would provide additive charge during triboelectric testing.The authors showed that within a single electrospun laminate material, the total volumetric charge from deformation was linear with the number of triboelectric interfaces according to the equation (where T is the number of triboelectric interfaces; which is the number of bilayers in a laminate -1): Combining two electrospun laminates together to form a TEG enabled a dramatic increase in charge density from 10 nC cm -3 from two films of EVA | PLA, up to 70 nC cm -3 for two electrospun laminates which were ordered based on the inverse polarisation ordering described above for ferroelectric inclusions.When converted to areal charge density, these electrospun laminates outperformed the ferroelectric polymer PVDF (Figure 16 e). [136]Although, the authors note that inversely polarised PVDF TENGs still outperform these laminates significantly. [132,136]

Outlook for Polymer Triboelectric Energy Harvesters
The mechanisms of polymer | polymer contact electrification are still being developed.However, significant progress has been made in demonstrating the importance and dominance of the mass transfer mechanism.
Numerous approaches focussed on mechanical properties, topography, and adhesion have been proposed to enhance surface charge at a single contact interface, while approaches to generate charges in the volume of polymers have shown great promise for exceptional voltage and charge outputs.
To continue to develop the field, there are critical challenges that must be addressed, specifically around, the stability of charge, the role of water in charge generation, and pathways to predict surface charging for any given contact interface.
The following sections reviews key questions that face the field.

The Role of Water at the Interface
The ubiquity of water in air and on surfaces makes it difficult to unravel the different mechanisms affecting polymer | polymer contact electrification, and a better understanding of fluid-solid interfaces is required.Water can behave as either an acid or a base, and can therefore generate mobile charges of negative (OH-) and positive sign (H+).Recent simulations have shown that charged fragments caused by bond breakage will react in the presence of humidity to generate mobile H+ or OH-charge carriers that can drive triboelectric charging. [48]yond its ability to enhance charge creation, the presence of water (and other liquids) can directly influence the formation of charge mosaics via several mechanisms, including electrostatic discharges between separating surfaces, [51] heavily influenced by the relative humidity, and charge accumulation due to evaporation on the surface, leading to negative charge forming where liquid evaporates and positive charge forming within the remaining liquid. [50]Charge mosaics occur consistently in air in the relative humidity range 40 -80%, and it is unclear what the precise role of water is on the electrostatic discharge.
In addition to the effects of water due to surface layers and ambient humidity, triboelectric charge is also generated by liquid drops sliding on surfaces, with drop motion generating a net charge within the drop and a charge of opposite sign on the surface.The surface charge is positive or negative depending upon the properties of the surface. [137]In addition, a history effect has been observed with multiple drops sequentially sliding on a surface, where the charging of later drops is slower due to the effect of previous drops. [138]More research is required to better understand and control the charge generated by liquids sliding on surfaces, as this has direct implications for polymer | polymer contact electrification with significant amounts of water present at interfaces.

Charge Stability (and Electrostatic Discharge)
Contact-separation on a pristine polymer-polymer interface will generate charge based on the mass transfer effect.How these charges exist on the surface over time, and how they discharge is an emerging question that must be addressed.Recent work by Grzybowski et al., has demonstrated that in a 1D contact line interface the charge mosaic that is formed in part arises from discharge enabled by adsorbed water. [51]The question of how these high charge density surfaces are stabilised becomes an important question when trying to use triboelectric charges to perform chemical work.Ciampi et al. demonstrated that in order to use triboelectric charges to drive electrochemical reactions, the total magnitude of charge generated was not important, rather it was the timeframe of the stability of the charge. [139]Interestingly, their work showed that fragments generated by different polymers had different stability dependent on polarity, with PVC having stable cations and unstable anions, and PTFE having stable anions and unstable cations, based on the ionisation energy of the polymers.
One pathway to study charge stability in detail is to generate bespoke and atomically engineered surfaces that can be studied for triboelectric charging experiments.Shin et al., [140] recently demonstrated the potential of this approach by functionalising PET and PEI surfaces with various functional groups with different electronegativities.This was achieved through an oxygen plasma treatment, followed by exposure to targeted reactant molecules.While this approach engineered a TEG with 8x increased charge output, the key advantage in this approach lies in the being able to study, in detail, the link between surface chemistry and triboelectric charge stability. [140]

A Holistic Model
This review has focussed on the individual factors that influence contact electrification and triboelectric charging within polymers.However, these factors still are studied in isolation, with steps to combine even two understandings (such as roughness and cohesive energy density effect on polarity) still rudimentary.There is a multidisciplinary effort to achieve a general model of triboelectric charging, firstly in polymers, then in all materials based off a combined mass transfer, ion transfer, and electron transfer mechanism.
Conceptually, this must include (at the very least) humidity, wettability, possibility of covalent bond cleavage (and heterolytic: homolytic cleavage ratio), surface heterogeneity, electron accepting and donating capability, surface charge stability, discharging, and surface energy both in stasis and under dynamic compression.To achieve this model is a large ask, and excellent efforts from researchers have started to develop these systems [43,45,65] but significant advances are still required.If we can generate a holistic model, there is the potential to revolutionise mechanical energy harvesting and sensing devices by engineering optimal devices for a given environment.

Device Engineering -Considering the Target Application
Significant research focus in the triboelectric field has gone into engineering devices with custom movement profiles, contact separation, sliding, or rotational.However, in practice the motion from the environment can come in a variety of different forms, and thus how a TENG converts environmental motion into contactseparation in a controlled way is critical.This means, that the device form factor that gives the highest laboratory voltage, current, or charge output, may not give the best output when capturing ambient energy.For instance, rotational TENGs have been demonstrated to have exceptional outputs, in part due to their high relative speed of motion.Such systems may be highly applicable to harvesting energy from fluid flow, such as wind or waterbut not practical for capturing human or vibrational motion.In contrast sliding mode TENGs may be optimal for harvesting energy from bike tires, car tires, and energy recovery from within machinery or prosthetics.Vertical contact-separation mode TENGs are by far the most commonly studied TENGs, and these systems are ideal for vibrational energy harvesting, including acoustics, and capturing human motion.In order to exploit the fundamental understanding in polymer triboelectric charging, we must engineer devices from these systems that suit the type of motion in our environment.

Conclusions
Triboelectric charging in polymers is complex, with a myriad of factors contributing to the formation, motion, and stability of charge.We have reviewed key parameters that influence the magnitude and polarity of triboelectric charging, and summarised approaches that can be taken, from a materials perspective, to generate the maximum amount of triboelectric output.Significant work is needed in the field, to standardise reporting, understand the role of water, predict charge stability, and generate a holistic model, however there is a great opportunity to generate revolutionary materials if these factors can be properly understood.

Figure 6 .
Figure 6.Schematic of proposed complicated contact-electrification mechanisms, with primary charge formation occurring due to mechano-radical and ion formation, and subsequent adsorbed ion and electron transfer with continued contactseparation.
et al. demonstrated that by measuring contact electrification for 19 different thermoplastic polymers in contact with indium-tin oxide (ITO) coated glass, the contact charge density of polymer decreases with increasing elastic modulus (Figure7 b).If we consider what is occurring in this system, we can start to understand why softer polymers charge more.Compressive modulus effectively defines the degree of motion a surface will undergo with a given applied force, a lower modulus (in the context of contact electrification experiments) means two things; 1) there is more electrostatic induction from relative charge motion due to greater displacements and vibration; and 2) the greater degree of compression for a low modulus material means the polymer chains are individually moving more, with a higher chance of becoming entangled and hence fracturing.In the same manner, stronger charging has been observed when contacting soft polymers against hard polymers (Figure7 b), regardless of chemical composition.This occurs, as contacting a hard and soft polymer together, leads to increased stress localisation within the soft polymer, compared to a balanced stress distribution when contacting two soft or two hard polymers.This effect can increase the total charge density by up to an order of magnitude.

Figure 8 .
Figure 8.Effect of temperature; a) electrostatic charging between polymers and a glass rod at different temperatures; b) effect of measured charge density as polymers cross their T G (a,b reproduced with permission from Šutka et al.,[97] copyright Royal Society of Chemistry, 2020), and c) influence of thermal history on charge from contact-separation experiments, blue = no thermal treatment; red = thermal treatment at 130 °C for 60 minutes (reproduced with permission from Šutka et al.,[46] copyright RSC).

Figure 15 .
Figure 15.Combined triboelectric and ferroelectric energy harvesters; a) the average open circuit voltage and short circuit current peak values for the TENG device based on contact between PVDF layers with different polarity; b) schematics show of the working mechanism for the triboelectric generator device based on inversely polarised PVDF layers (a, b reproduced with permission from Šutka et al.,[132] copyright Royal Society of Chemistry, 2018); and c) the double capacitor model and corresponding contact− separation stages.At first, ferroelectric layers contact and form a connected capacitor, and then layers are compressed which results in a negative piezoelectric response.Next, the films are unloaded and separated resulting in a positive piezoelectric response, thus disconnecting the capacitors that had formed the connected circuit (c, reproduced with permission from Lapčinskis et al.,[134] copyright American Chemical Society, 2019).