Zhong Lin Wang (*zhong.wang@mse.gatech.edu)

Piezoelectric nanogenerator (PENG) was first introduced by using piezoelectric nanowires for converting tiny mechanical energy into electric power.1 Research in nanogenerators has been vastly expanded in the last decade due to the invention of the triboelectric nanogenerator (TENG).2,3 As of today, the definition of nanogenerator has far exceeded its traditional meaning, and it represents a field that uses the Maxwell’s displacement current to convert mechanical energy into electric power/signal.4–6 This field is attracting a wide range of interest due to the huge advances in the internet of things, big data, sensor network, robotics, and artificial intelligence.7–13 TENGs are playing a key role in harvesting high entropy energy distributed in our living environment for effective driving of distributed electronics and systems.14–20 

The piezotronic effect is about the use of piezoelectric polarization charges at an interface for effectively tuning/controlling charge carriers across a metal–semiconductor interface, which was first introduced in 2007.21,22 The piezo-phototronic effect is about the use of piezoelectric polarization charges at a p–n junction for effectively tuning/controlling charge carriers recombination or separation at the interface, which was first introduced in 2010.23,24

From a recent SCI database search, there are 57 countries and regions, over 800 units and over 6000 scientists worldwide who are engaged in TENG research. The papers published in the public domain ever since the invention of nanogenerators in 2006 are given in Fig. 1, which unambiguously show that the research in nanogenerators is a focused field worldwide.

FIG. 1.

Statistics of publications from Web of Science by 2020. The number of publications in each year (a) and country (b) when “Nanogenerator” was used as the keyword for search in Web of Science.

FIG. 1.

Statistics of publications from Web of Science by 2020. The number of publications in each year (a) and country (b) when “Nanogenerator” was used as the keyword for search in Web of Science.

Close modal

This special issue is about some of the current progress made in nanogenerators and piezotronics. Dudem et al. systematically reviewed the theoretical progress made for TENGs and how they can be utilized for optimizing the output power. Fundamental technical advances made for improving the performance of the TENG have been reported by Li et al., Zhang and Bowen, and Basset, not only about the choice of materials but also about the power management system. Since the choice of materials for TENG is rather broad, Fan and Wu reported the use of 2D materials for nanogenerators and piezotronics, and Kim reported the use of metamaterials for energy harvesting. Wang, Hou, and Priya have reported the use of perovskites for TENG. Xu and Hao have elaborated on the use of smart polymer materials for TENG. The remaining articles are about the novel applications of TENGs for a variety of fields, such as acoustic sensors (Wand and Lee), biomedical research (Wand and Long), powering of wireless sensor systems (Chew, Kuang, Ruan, and Zhu), e-textiles (Beeby), human–machine interfacing (Pu, Guo, and Hu), and powering body-implantable medical devices (Karan and Kim). Although the collections of articles in this special issue is a small portion of the current progress in nanogenerators and piezotronics, a broad range of research is being carried out worldwide. Nanogenerators will find major applications in micro–nano power sources, self-powered sensors, blue energy, and high voltage sources. Piezotronics and piezophototronics will find applications related to devices made for the third generation semiconductors. We anticipate that the field will be advanced very fast, which will soon impact industrial technology in specific areas.

Bhaskar Dudem, R. D. Ishara G. Dharmasena, Sumanta Kumar Karan, S. Ravi P. Silva (*s.silva@surrey.ac.uk)

1. Abstract

Triboelectric nanogenerators (TENGs) are one of the most promising energy harvesting methods available for next-generation wearables, autonomous devices and sensors, and the Internet-of-things (IoT), which can efficiently convert ambient mechanical energy into useful electricity. It can be implemented in clothing, shoes, walkways, and moving parts in automobiles, harvest suitable energy to drive many types of portable/wearable/implantable electronics that at present, and are predominantly powered by batteries. In order to move the current state-of-the-art in practical devices to realistic technologies, much development is still needed. These requirements we envisage will be accelerated with the help of theoretical models and simulations, which can be verified and refined using an empirical route to best fit experimental data. This will give rise to self-validated models that allow for predictive design of TENG devices for specific applications using computer-aided design (CAD) and simulators.

2. Introduction

TENGs were first proposed by Yang et al. in 2012,25 and it is well-known that they generate electricity based on the triboelectric effect, the frictional contact between two triboelectric materials resulting in static charge generation, and electrostatic induction resulting from the relative movement of such charged surfaces.2,26–28 While contact-electrification (CE) is believed to be the fundamental phenomenon for charge generation during the friction of a TENG, the exact mechanism of this charge generation is not yet fully understood. In addition, the mechanism to enhance the output performance of TENGs by surface modification of the triboelectric materials is also not fully elucidated. Thus far, an extensive effort has been made to understand and optimize the output performance of TENGs by distinct types of theoretical models and simulations,29–42 many of which are represented on a timeline as shown in Fig. 2.

FIG. 2.

Timeline displaying the progress in various types of theoretical approaches, thus far, to optimize the performance of TENGs. 2013 and 2014: Capacitive model for (i) sliding-mode,29 vertical contact–separation (ii) dual and (iii) single electrode mode,30,31 and (iv) grating structured TENGs.32, 2015: (v) Theoretical prediction for contact-mode free-standing TENG.33, 2017: (vi) Distance-dependent electric field (DDEF) model of a vertical contact–separation mode TENG;35 (vii) and (viii) stress simulation analysis across the nano-architecture triboelectric layers.36,37 2018: (ix) the universal DDEF models to simulate and optimize different modes of TENGs38 and (x) TENG impedance pots and TENG power transfer theory to understand its output power transfer;39 (xi) electron-cloud-potential-well model for elucidating electron transfer mechanism in contact electrification;40 (xii) the electrostatic potential distribution generated across the distinct atoms like F–C–Al and C–F–Al during the superposition of Al and the PTFE.41, 2019: (xiii) photon excitation effect to explain the charge transfer mechanism in triboelectrification.42 

FIG. 2.

Timeline displaying the progress in various types of theoretical approaches, thus far, to optimize the performance of TENGs. 2013 and 2014: Capacitive model for (i) sliding-mode,29 vertical contact–separation (ii) dual and (iii) single electrode mode,30,31 and (iv) grating structured TENGs.32, 2015: (v) Theoretical prediction for contact-mode free-standing TENG.33, 2017: (vi) Distance-dependent electric field (DDEF) model of a vertical contact–separation mode TENG;35 (vii) and (viii) stress simulation analysis across the nano-architecture triboelectric layers.36,37 2018: (ix) the universal DDEF models to simulate and optimize different modes of TENGs38 and (x) TENG impedance pots and TENG power transfer theory to understand its output power transfer;39 (xi) electron-cloud-potential-well model for elucidating electron transfer mechanism in contact electrification;40 (xii) the electrostatic potential distribution generated across the distinct atoms like F–C–Al and C–F–Al during the superposition of Al and the PTFE.41, 2019: (xiii) photon excitation effect to explain the charge transfer mechanism in triboelectrification.42 

Close modal

3. Development of various theoretical models

a. Exploring the electrical response of TENGs based on mathematical modeling.
TENGs typically contain triboelectric charges on their contact surfaces, and the relative movement of these charged surfaces causes the electric fields acting on their electrodes to vary, which results in an induced output current flow between the electrodes. The power output of the TENG is extracted by driving this current through an external load. This behavior is commonly expressed using Maxwell’s displacement current4 
JD=Dt=εEt+PSt.
(1)
Herein, JD = displacement current density, D = displacement field, t = time, E = electric field, and PS = polarization of the medium. Two fundamental approaches have been presented so far to simulate the output behavior of TENGs: the parallel plate capacitor approach and the distance-dependent electric field approach.
The first generation of theoretical models representing TENGs was derived using the parallel-plate capacitor approach [Figs. 2(i) and 2(ii)]. In this approach, each TENG architecture was represented using a unique combination of parallel plate capacitors [Figs. 2(i)2(v)] to approximate their output trends. Using these capacitor assemblies, a relationship between the voltage (V), charge (Q), and the separation distance of TENG layers (x) was developed (known as the VQx relationship), which is given by
V=1CQ+VOC.
(2)
Herein, VOC is the open-circuit voltage, and C is the overall capacitance of the TENG. Initially derived for the vertical contact separation mode TENG (VCSTENG), this concept has been expanded to represent other contact modes as well as sliding mode TENG architectures [Figs. 2(i)2(v)].29–31,33 An universal edge approximation-based equivalent capacitance method has also been proposed as an extension of this concept, in which the capacitances of the edge effect are taken into consideration to estimate the VQx relationship that applies to all modes of TENGs.43 However, the parallel plate capacitor models contain drawbacks, such as the relatively low accuracy in predicting TENG outputs, the necessity of deriving bespoke capacitor models for each TENG architecture, difficulty in explaining the polarization of dielectrics, induction of output charges on electrodes, etc.35,38,39,44,45
More recently, the distance-dependent electric field (DDEF) concept has been presented based on Maxwell’s equations as a universal platform to describe the output trends of TENGs [Fig. 2(vi)]. As opposed to the parallel plate capacitor model, the DDEF model studies the electric fields originating from triboelectrically charged surfaces by considering their finite dimensions. For instance, according to the DDEF model, the overall electric field originating from a triboelectrically charged surface (along its perpendicular axis) Ez, with charge density σ, dimensions L, W, permittivity ɛ, and at a perpendicular distance of x, is presented by
Ez=σπεarctanLW2xW4xW2+LW2+1=σπεf(x).
(3)
The DDEF equation [Eq. (3)] can be applied to triboelectrically charged surfaces as well as the electrode surfaces to derive the outputs of TENG. Assuming a TENG with m number of triboelectric surfaces where the charge density (of an i-th surface) is given by σT,i and output charge density is σu [Fig. 2(ix)], the potential (Φ) of the electrodes [a (Φa) and b (Φb)] are calculated as39 
Φa=σuπεa0yfxdx+1πi=1mσT,iεaxa,ifxdx,
(4)
Φb=σuπεb0yfxdx+1πi=1mσT,iεbxb,ifxdx.
(5)

Therefore, the current, charge, voltage, and power outputs can be derived using Eqs. (4) and (5). The DDEF approach results in several advantages over the parallel plate model, providing a detailed understanding of the electric field propagation, TENG polarization, and output induction, along with higher accuracy in describing the TENG outputs.

Furthermore, some of the recent studies have expanded on these developments based on Maxwell’s displacement current theory to develop three-dimensional theoretical models to describe the output power generation of TENGs.46 

b. Optimizing TENG electrical outputs using mathematical modeling.

The DDEF model has been used to understand the power generation behavior of the TENG, resulting in the derivation of the TENG power transfer theory and TENG impedance plots [Fig. 2(x)].38 TENG power transfer theory consists of a combination of the DDEF model and Norton’s theorem, presenting the power output of a TENG by means of current output and an impedance element. This allows for its implementation more readily into circuit simulators by means of circuit blocks. TENG impedance plots are a similar strategy that helps in visualizing the maximum power generation conditions. Both these techniques, along with the DDEF model simulations, have been used to examine the outputs of a TENG and to optimize its power generation conditions. Accordingly, the primary factors affecting the power generation of a TENG have been defined, which are divided into the material, structural, and kinetic parameter sub-categories.

Considering material parameters, triboelectric charge density and dielectric constant of the TENG layers hold critical importance.38 Higher charge densities help in increasing the power outputs, which can be achieved via appropriate material selection, surface structuring, and physical and chemical surface modifications.47–49 Theoretically, having lower dielectric constants have been shown to enhance to electrostatic induction,38 however, some experimental studies have shown that the use of high dielectric constant nanomaterials results in an overall increase in the power output. Structural parameters, such as the area of the TENG layers and their thickness, also affect the TENG outputs. Larger surface areas of triboelectric layers result in higher power outputs as well as the reduction of the internal impedance of the device, both of which are desirable for TENGs. On the other hand, smaller thicknesses of dielectric triboelectric surfaces are desirable to increase the electrostatic induction, hence the power output. However, it has been shown that a minimum level of thickness of the (dielectric) triboelectric layers (typically few micrometers) is required to ensure that the triboelectric charges can accumulate stably. With regard to the motion parameters, higher rates of movement of TENG layers (high velocity or high frequency) increase the power output while reducing the TENG impedance. The behavior of a TENG at increasing amplitudes (maximum separation) of movement shows a relatively similar trend up to a threshold value. Furthermore, conformal contact between TENG surfaces has been shown to provide higher power outputs and reduced impedances, in comparison to the non-contact mode TENG operations. Therefore, the design of a TENG for a given application needs to be conducted considering an appropriate balance of the above parameters, in which the theoretical models, TENG power transfer theory as well as TENG impedance plots will act as design tools.

In addition to these theoretical models, a standard method has been proposed to quantitatively evaluate the performance of TENGs, which consists of a structural figure-of-merit (FOM) related to the design of the TENG and a material figure-of-merit (FOMm) as the square of the surface charge density.50 Within the built-up V-Q, the TENG operation cycle with maximized energy output is initially proposed. Based on this maximum energy output per cycle and considering both the maximized energy conversion efficiency and the maximized average output power, the performance FOM was derived to evaluate each TENG design, composed by a FOM and a FOMm. However, the breakdown condition is not considered in this process, which can seriously affect the maximum energy output. Therefore, a standardized method that considers the breakdown effect is further proposed for output capability assessment of nanogenerators, which is crucial for the standardized evaluation and application of nanogenerator technologies.51 In addition, these structural FOM of TENGs are extended to quantitatively evaluating and comparing output performance under different load resistances as well as in charging systems (powering capacitors).52,53 Therefore, these FOM standards will set the foundation for all further applications and industrialization of the TENGs, including operation in a hybrid energy storage system.

c. Optimizing the electrical response of TENGs based on the mechanical stress simulation models.

The surface modification of the frictional materials by creating nano- or micro-architectures is one of the major routes to improve the performance of TENGs.3,27,36,54–59 This is because such architectures can be expected to increase the surface roughness and result in an enhanced contact area as well as a high output performance of TENG.36 However, the exact mechanisms by which they influence the performance are not yet fully clear. For example, even though some of the surface morphologies exhibit a high surface roughness, the resultant contact area can be rather reduced while they contact together. As a result, finite element method (FEM) simulations have been utilized to estimate the mechanical stress at the contacting interface between the nano-architecture polymer [such as polydimethylsiloxane (PDMS)] and metallic layers [Fig. 2(vii)].36 In this study, the mechanical stress distributed across the surface of nano-pillars is considered as a crucial parameter, since the number of trapping electrons and electron transportation at the higher stress sites can be expected to maximize owing to the higher deformation of internal structures. Thus, by estimating the contact area and mechanical stress together (i.e., the product of contact area and stress is defined as contact force), an optimal structural dimension such as diameter and period of nano-pillars is determined to attain a high electrical output response of TENGs. These studies are elucidating that an effective contact force distributed across the surface of nanostructures can largely contribute to the effective charge density as well as the electrical output performance of TENGs. Along with the dimensions, the geometric shape of the nanostructures may also expect to influence the durability and performance of TENGs. Thus, the effect of geometric shapes on the output characteristics of TENG is also investigated by considering again the contact force as a reference [Fig. 2(viii)].37 The similar type of mechanical stress simulations used and realized within the nano-architectures that undergo greater deformation (like dome-shape) can result in a larger contact surface as well as the higher electric output owing to the larger contact force or mechanical stress distributions across their surface. Such nano-architectures are force sensitive, and less durable or not reliable against long-term contact/separation cyclic operations. In contrast, nanostructures with moderate deformability (like pillar-shape) can significantly improve the performance and are durable and longer-lasting under repetitive cyclic operations. Therefore, such mechanical simulation models can be used as a roadmap to decide on the optimistic shapes and dimensions of nano-architectures on polymer for designing the force sensitive, durable, and high-performance TENGs. Along with these surface features, the mechanical forces applied on TENG devices also play a key role to improve its surface charge density and output performance. Thus, a comprehensive theoretical model has been proposed to understand the energy conversion in various TENG modes, in which both the mechanical energy inputs indicated by the F-x (force–displacement) plot and electrical energy output as reflected by the V-Q (voltage–charge) plot are simulated and analyzed.60 Such a model can also play an important role for optimizing the energy conversion efficiency in TENGs by considering both the mechanical and electrical energies, simultaneously.

d. Theoretical models to explain the charge transfer mechanism in contact–electrification.

All the above mathematical and mechanical simulation models postulate that the electrical output of TENGs is mainly dictated by the triboelectric surface charge density and it is enhanced by altering the surface features of triboelectric materials. This is despite the fundamental mechanism of contact–electrification is still not fully understood.61 Therefore, along with optimizing the parameters of TENGs, understanding the charge transfer mechanism in contact electrification is crucial for the TENG research community. Previous literature suggests that either the electron or ion transfer is involved in the contact electrification processes.61–65 However, Xu et al. recently examined this age-old problem by the study of surface charge density evolution across the surface of triboelectric material with time at various high temperatures, with data consistent with the electron thermionic emission model.40 From the results, they have reported that the contact electrification is dominated by the electron transfer in solid–solid materials, rather than the ion transfer.40 They have also proposed an electron cloud/potential-well model [Fig. 2(xi)] based on fundamental electron cloud interactions to explain all types of contact–electrification phenomena for general materials in the contact–separation mode TENGs. As is well known, atoms in the materials consist of electrons within the inner atomic or molecular orbitals that are tightly bound, whereas electrons in the outermost orbits are mostly loosely bound. Therefore, here an atom was considered as a potential well in which the outermost loosely bound electrons form an electron cloud around the atom. As shown in Fig. 2(xi), the electron clouds of two atoms belonging to two different materials are separated by an interatomic distance before an external compression force is applied, while the electrons cannot transfer due to the local trapping effect in their potential wells. Once the force is applied, the electron clouds overlap owing to the physical contact between the corresponding materials, and their initial single potential wells become an asymmetric double-well potential. Consequently, the electrons can transfer from the atom of one material to the atom of another, resulting in contact–electrification. The key role of the external compression force is to bring the two materials into contact as well as to shorten the distance between the atoms that can cause a strong overlap of their electron clouds in the repulsive region. Likewise, the charges/electrons can be transferred if one material rubs against the other. After the separation of these materials, most of the transferred electrons can remain as static charges on the surface of materials owing to the energy barrier present in the corresponding material if the temperature is not too high. As a result, one of the materials gets charged positively and the other charged negatively. With the elevated temperature, the electrons transferred into negatively charged material atoms are more likely to hop out of the potential well, and they either return into the positively charged atoms or emit thermionically into the air. This study is one of the breakthroughs to understand the charge transfer mechanism in TENGs, which can be generally applied to explain all types of contact electrification processes in conventional materials, such as the combination of metal–semiconductor, metal–polymer, and polymer–polymer. Furthermore, a typical aluminum (Al)–polytetrafluoroethylene (PTFE) material pair of TENG has been tested by Wu et al. [Fig. 2(xii)] to analyze the mechanism of metal–polymer contact electrification, and the outcomes provide direct evidence for the above-stated cloud/potential-well model.41 In 2020, a modified electron cloud-potential well model has been expanded for explaining the contact–electrification and charge transfer and release between two materials in the sliding-mode TENG.66 Lin et al. further examined the effect of photon excitation on contact electrification [Fig. 2(xiii)].42 By illuminating the surface of an insulator (such as SiO2 or PVC) with UV light at a specific wavelength and intensities, the surface electrostatic charges can be released under photon excitation. To verify the photoelectron emission of electrons in contact electrification, the surface of an insulator (such as SiO2 or PVC) illuminated with the light and effects of its wavelength and intensity on the irradiation-induced triboelectric charge decay were studied. These studies show that there exists a threshold photon energy above which surface electrostatic charges will be released. Therefore, both the electron thermionic emission and photon excitation studies indicate that the electron transfer plays a dominant role in contact electrification, particularly in solid–solid cases.

4. Concluding remarks

A number of theoretical approaches have been presented in the literature to explain the working principles and the output trends of different TENG architectures, which can be categorized under the parallel-plate capacitor approach and the DDEF approach. These models, along with the accompanying tools, such as the TENG impedance plots and TENG power transfer theory, have revealed a number of new methods on the optimization of the TENG power outputs, targeting enhanced power generation, efficiency, and effectiveness as energy harvesters and self-powered sensors. Moreover, recently presented theoretical models have shed new light toward the origin of the triboelectric effect, as well as the surface modification related performance enhancement, which will potentially contribute toward designing better TENGs. Such theories and advancements can be further customized to suit a range of different applications, for example, textiles, large-scale energy harvesting, such as ocean waves, and IoT applications.67–70 Therefore, the work related to TENG theoretical modeling continues to develop a comprehensive understanding on the origin of the triboelectric effect, static charge transfer between surfaces, novel material developments, and output induction of TENGs and their fine tuning on macro-, micro-. and nanoscales, leading toward the design and fabrication of practical and sustainable TENG applications.

With regard to TENGs, their typical large internal impedance has been a major issue (which normally is in Mega-Ohm to Giga-Ohm range at low operating frequencies) in using them for practical energy harvesting applications, with typical electronic devices or energy storage units containing relatively lower impedance (several Ohms).38 This makes the transfer of power from the TENG to such practical loads extremely low efficiency due to the load mismatch. However, theoretical works on impedance characterization, impedance engineering, and device engineering have the potential to provide a viable solution to this issue as evident in recent studies,38,71–73 which will significantly impact the overall energy conversion efficiency of TENGs. Furthermore, this can help to manage the high voltage and low current outputs, which are characteristic to TENG outputs, that need to co-exist with the high impedance issue. Moreover, designing efficient energy storage systems backed by the theoretical advances, which match to the TENG impedance characteristics, would also help to manage drawbacks caused by the high voltage generation within the TENGs.

5. Acknowledgments

The authors would like to acknowledge the support from the EPSRC Research Project (Grant No. EP/S02106X/1) for funding this work. This work was also supported by the Royal Academy of Engineering under the Research Fellowship Scheme.

Jialu Li, Zhaoling Li (*zli@dhu.edu.cn), Zhong Lin Wang (*zhong.wang@mse.gatech.edu)

1. Status

Nowadays, the rapid development of human civilization requires the support of a large amount of energy supply. However, the world’s energy structure is still dominated by low-entropy energy that is high-concentration and high-quality, such as coal, oil, and natural gas.74 In comparison with low-entropy energy, energy, widely distributed in the environment, that is low-quality and irregular is called high-entropy energy, such as wind, water wave, mechanical vibration, and human activities.10 After combustion, transportation, and application, low-entropy energy eventually becomes high entropy energy dissipated in nature. According to the first law of thermodynamics, the massive use of fossil energy will inevitably increase the disordered and low-quality high entropy energy distributed in the environment, aggravate the greenhouse effect, and cause global climate warming, which is contrary to the strategy of low carbon and sustainable development. Meanwhile, with the advent of the Internet of Things, the demand for social development is distributed, and our energy supply needs to be distributed. Currently, China has changed from “concentrated energy” to “distributed energy” in the global energy transformation. Therefore, how to collect high entropy energy with instability, low quality, low frequency, and wide distribution into effective output electric energy is a hot issue that urgently needs to be solved.75 Triboelectric nanogenerator (TENG) is proposed for effective collection and utilization of high entropy energy. In 2012, we fabricated the first TENG that is an energy acquisition method based on the coupling effects of contact electrification and electrostatic induction.2 This emerging technology is capable of converting extensively existed environmental mechanical energy into continuous electricity.76 

At present, most TENGs are fabricated based on thin polymer films or elastic rubbers. The advantages of these triboelectric materials include lightweight, prominent mechanical properties, large power output, and high stability. Simultaneously, breathability is very crucial factor for the practical application of TENGs in the field of intelligent wearables. Meanwhile, hydrogels and aerogels are also common constructing materials. Unfavorably, most hydrogels are prepared with organic solvents, which are not environmentally friendly. Besides, hydrogels are prone to collapse when subjected to repeated mechanical stretching. Aerogels are porous materials with space network structure exhibiting low density and prominent gas permeability.

Alternatively, sustainable textiles could be an optimal option to construct breathable and high performance TENGs. To start with, textiles have favorable softness and flexibility to conformably comply with human’s skin with ease. Moreover, they possess high specific surface area and diversified micropore structure, endowing textile-based TENGs with excellent portability, brilliant breathability, and high power output. Additionally, the low-cost and simple fabrication process is indispensable for sustainable textiles, which is more conducive to the realization of large-scale production and industrialization of textile-based TENGs. For instance, Huang et al.77 developed a washable textile TEGN by adopting commercially available knitting method, and it delivered an open-circuit voltage of up to 800 V and a maximum power density of 203 mW/m2. This washable and comfortable energy textile can drive warning indicator and smart watch or act as a motion sensor to monitor human movement signals. As portable and reliable power supply sources, textile-based TENGs have opened up promising possibilities for broad application prospects,78 ranging from mechanical energy harvesting, wearable electronics driving, to self-powered environmental sensing and health monitoring (Figs. 3 and 4).79 

FIG. 3.

Schematic illustrations and application demonstrations of wearable textile-based TENGs. (a) Reproduced with permission from Xiong et al., Nat. Commun. 9, 4280 (2018). Copyright 2018 The Authors, published by Springer Nature. (b) Reproduced with permission from Sun et al., Nat. Commun. 11, 572 (2020).80 Copyright 2020 The Authors, published by Springer Nature. (c) Reproduced with permission from Guo et al., Nano Energy 48, 152 (2018).81 Copyright 2018 Elsevier. (d) Reproduced with permission form Zhou et al., Sci. Rep. 7, 12949 (2017).82 Copyright 2017 The Authors, published by Springer Nature. (e) Reproduced with permission from Kwak et al., ACS Nano 11, 10733 (2017).83 Copyright 2017 American Chemical Society. (f) Reproduced with permission from Qiu et al., Nano Energy 58, 750 (2019). Copyright 2019 Elsevier. (g) Reproduced with permission from Zhao et al., Adv. Mater. 28, 10267 (2016).84 Copyright 2016 Wiley-VCH. (h) Reproduced with permission from Guo et al., ACS Appl. Mater. Interfaces 8, 4676 (2016).85 Copyright 2016 American Chemical Society. (i) Reproduced with permission from Pu et al., Adv. Mater. 27, 2472 (2015).86 Copyright 2015 Wiley-VCH. (j) Reproduced with permission from Gong et al., Nat. Commun. 10, 868 (2019).87 Copyright 2019 The Authors, published by Springer Nature. (k) Reproduced with permission from Seung et al., ACS Nano 9, 3501 (2015).88 Copyright 2015 American Chemical Society.

FIG. 3.

Schematic illustrations and application demonstrations of wearable textile-based TENGs. (a) Reproduced with permission from Xiong et al., Nat. Commun. 9, 4280 (2018). Copyright 2018 The Authors, published by Springer Nature. (b) Reproduced with permission from Sun et al., Nat. Commun. 11, 572 (2020).80 Copyright 2020 The Authors, published by Springer Nature. (c) Reproduced with permission from Guo et al., Nano Energy 48, 152 (2018).81 Copyright 2018 Elsevier. (d) Reproduced with permission form Zhou et al., Sci. Rep. 7, 12949 (2017).82 Copyright 2017 The Authors, published by Springer Nature. (e) Reproduced with permission from Kwak et al., ACS Nano 11, 10733 (2017).83 Copyright 2017 American Chemical Society. (f) Reproduced with permission from Qiu et al., Nano Energy 58, 750 (2019). Copyright 2019 Elsevier. (g) Reproduced with permission from Zhao et al., Adv. Mater. 28, 10267 (2016).84 Copyright 2016 Wiley-VCH. (h) Reproduced with permission from Guo et al., ACS Appl. Mater. Interfaces 8, 4676 (2016).85 Copyright 2016 American Chemical Society. (i) Reproduced with permission from Pu et al., Adv. Mater. 27, 2472 (2015).86 Copyright 2015 Wiley-VCH. (j) Reproduced with permission from Gong et al., Nat. Commun. 10, 868 (2019).87 Copyright 2019 The Authors, published by Springer Nature. (k) Reproduced with permission from Seung et al., ACS Nano 9, 3501 (2015).88 Copyright 2015 American Chemical Society.

Close modal
FIG. 4.

Categories of textile-based TENGs according to the internal geometries and structural dimensions: 1D fiber or yarn-based TENGs, 2D fabrics-based TENGs, and 3D fabrics-based TENGs. (a) Reproduced with permission from Yang et al., ACS Appl. Mater. Interfaces 10, 42356 (2018).89 Copyright 2018 American Chemical Society. (b) Reproduced with permission form He et al., Adv. Funct. Mater. 27, 1604378 (2017).90 Copyright 2017 Wiley-VCH. (c) Reproduced with permission from Liu et al., Nanoscale Adv. 2, 4482 (2020).91 Copyright 2020 Royal Society of Chemistry. (d) Reproduced with permission from Zhong et al., ACS Nano 8, 6273 (2014).92 Copyright 2014 American Chemical Society. (e) Reproduced with permission from Li et al., Nano Energy 36, 341 (2017).93 Copyright 2017 Elsevier. (f) Reproduced with permission from Zhou et al., ACS Appl. Mater. Interfaces 6, 14695 (2014).94 Copyright 2014 American Chemical Society. (g) Reproduced with permission from Cao et al., ACS Nano 12, 5190 (2018). Copyright 2018 American Chemical Society. (h) Reproduced with permission from Zhao et al., ACS Nano 10, 1780 (2016).95 Copyright 2016 American Chemical Society. (i) Reproduced with permission from Li et al., Adv. Energy Mater. 7, 1602832 (2017).96 Copyright 2017 Wiley-VCH. (j) Reproduced with permission from Dong et al., Adv. Mater. 29, 1702648 (2017).97 Copyright 2018 Wiley-VCH. (k) Reproduced with permission from Dong et al., Nat. Commun. 11, 2868 (2020). Copyright 2020 The Authors, published by Springer Nature.

FIG. 4.

Categories of textile-based TENGs according to the internal geometries and structural dimensions: 1D fiber or yarn-based TENGs, 2D fabrics-based TENGs, and 3D fabrics-based TENGs. (a) Reproduced with permission from Yang et al., ACS Appl. Mater. Interfaces 10, 42356 (2018).89 Copyright 2018 American Chemical Society. (b) Reproduced with permission form He et al., Adv. Funct. Mater. 27, 1604378 (2017).90 Copyright 2017 Wiley-VCH. (c) Reproduced with permission from Liu et al., Nanoscale Adv. 2, 4482 (2020).91 Copyright 2020 Royal Society of Chemistry. (d) Reproduced with permission from Zhong et al., ACS Nano 8, 6273 (2014).92 Copyright 2014 American Chemical Society. (e) Reproduced with permission from Li et al., Nano Energy 36, 341 (2017).93 Copyright 2017 Elsevier. (f) Reproduced with permission from Zhou et al., ACS Appl. Mater. Interfaces 6, 14695 (2014).94 Copyright 2014 American Chemical Society. (g) Reproduced with permission from Cao et al., ACS Nano 12, 5190 (2018). Copyright 2018 American Chemical Society. (h) Reproduced with permission from Zhao et al., ACS Nano 10, 1780 (2016).95 Copyright 2016 American Chemical Society. (i) Reproduced with permission from Li et al., Adv. Energy Mater. 7, 1602832 (2017).96 Copyright 2017 Wiley-VCH. (j) Reproduced with permission from Dong et al., Adv. Mater. 29, 1702648 (2017).97 Copyright 2018 Wiley-VCH. (k) Reproduced with permission from Dong et al., Nat. Commun. 11, 2868 (2020). Copyright 2020 The Authors, published by Springer Nature.

Close modal

2. Current and future challenges

Textile-based TENGs with compelling features have emerged as an effective energy conversion device and renovated the mechanical energy harvesting technology. Qiu et al.98 designed a single-electrode TENG textile by one-step surface coating method via simultaneously electrospinning and electrospray processes. This TENG textile possessed satisfactorily tailorable and washable properties, but the power density and durability were still far away from enough. Comparatively, Cheng et al.99 designed a flame-retardant textile-based TENG with high electrical output and mechanical stability. However, it cannot be washable owing to the adoption of layer-by-layer self-assembly technique for self-extinguishing ability. Even though tremendous progress has been achieved in textile-based TENGs, there are still many problems to be well addressed.

First, the electrical output performance is supposed to be further enhanced for practical applications. Generally, the electrical output of TENG largely depends on the effective contact area of those two friction materials. Textiles have intrinsically porous structure to obtain excellent gas permeability. The more holes textiles possess, the less triboelectric charge density they can produce, causing an inefficient energy conversion. Besides, substantial pore spaces will induce capillary effect, especially for hydrophilic fabrics, which absorb massive moisture from the air and results in a low power generation.100 

Second, multiple working models are needed to be integrated to efficiently harvest versatile and various human motions. TENG specifically includes contact–separation mode, sliding mode, single-electrode model, and freestanding layer mode. Since body motions concurrently cause mechanical friction and deformation in many forms, including contact, extrusion, and sliding, textile-based TENGs using single working model cannot take full advantage of these biomechanical motions to achieve an optimized energy harvesting efficiency.101–105 The structural design with complex geometric configuration is required to maximize the power output.

Third, washability is one of the key prerequisites for sustainable textiles to construct wearable TENGs. After washing in water for many times, actually, most textile-based TENGs would be prone to creep deformation, either volume expansion or volume reduction, thereby having poor dimensional stability. Worse still, moisture and liquid contaminants are common in daily scenarios, which will speed up the electron dissipation and largely impair the power output of textile-based TENGs.106 It is very necessary to improve the dimensional stability of textile materials while maintaining excellent power output in the presence of the water scrubbing.

Finally, in practice, a single fiber may have perfect electrical conductivity and mechanical flexibility, but after spinning and weaving, the resistance of fiber assembly or fabric increases sharply and the flexibility decreases dramatically. Conductance-stable triboelectric materials and advanced textile processing techniques are of significance to ensure the uniformity and stability of mass produced textile-based TENGs, while not sacrificing the electrical performance and flexible texture in the course of real applications.107 

3. Advances in science and technology to meet challenges

There are many approaches in science and technology available to address current and future challenges in textile-based TENGs. First, for the insufficient power output, one simple and straightforward way to further improve the power output is rational selection of triboelectric materials. Although all textiles exhibit triboelectricity, finding the right paired positive and negative materials can generate maximum output. Other effective way is surface functionalization on textiles. By using nanoparticles and possibly nanocomposites, nanomaterials demonstrating stronger capabilities to gain or lose electrons can be successfully introduced. Besides, surface roughening and three-dimensional (3D) textile TENGs configuration can considerably increase the effective contact area and separation distance during operations, respectively,108,109 and, in consequence, generate a distinctly high power output.

Second, for the sake of solving the restriction of single working model, at present, TENG has developed from the single working model to the multiple working models, which can be divided into the spherical 3D TENG, rotating-disk-based direct-current TENG, cylindrical spiral TENG, etc. Using multiple working models is more conducive to the effective collection of various mechanical energies in the ambient environment, which will perform a high charge density output even triggered by tiny deformations. In addition, to scavenge waste energy from all kinds of human motions, piezoelectric nanogenerators (PENGs) can be combined with TENGs and exhibit their own advantages and characteristics.103 The hybrid energy harvesters are quite compatible and can complement each other.

Third, waterproofing modification treatment enables the textile-based TENGs with outstanding washable property. For instance, hydrophobic or oleophobic coating materials, such as hydrophobic nanoparticles, can be adopted to endow liquid repellence.110 Also, it is applicable to directly use commercial waterproof fabric (i.e., Gore-Tex) as sacrificial substrate to accommodate the triboelectric materials.106 Since the TENG is based on the surface charging effect, its performance is greatly affected by the environmental humidity. Packaging technology is required to protect the device from vapor contamination or liquid permeation, but without reducing too much of the flexibility. Preserving its hydrophobicity and flexibility is important for improving the washability and energy conversion efficiency.

Finally, efficient modern textile technology lays the foundation of realizing commercially mass production of textile-based TENGs with favorable flexibility. Improving and perfecting the spinning and weaving techniques are beneficial to fabricate the prominent textile-based TENGs, especially in producing very soft yarns and fabrics. For example, in terms of textile TENGs with coaxial yarn intersection, a roller-guided assembly line has demonstrated impressive efficiency in scalable manufacture. Besides, liquid metals are proved to be conductance-stable materials in comparison with conventional metals, and liquid metal can be widely used in fabricating high performance textile-based TENGs.

4. Concluding remarks

Mechanical energy exhibits the characteristics of abundant source, wide distribution, diverse forms, and easy conversion, which is a preferred choice for effective environmental energy collection. With continuous advancement and development in recent years, textile-based TENGs have attracted worldwide attention and achieved remarkable research results. Textile-based TENGs open up a new possibility for mechanical energy harvesting and is expected to relieve the shortage of world energy problem and promote the sustainable development. The TENG not only represents an emerging type of energy conversion technology but also demonstrates versatile functions in self-powered sensing. Even though vast progress has ever been made to realize the widespread application of textile-based TENGs, it is still essential to further optimize the structure design and material selection, enhance the power output, develop multiple working models, improve the washable property, boost the flexibility and stability, and expand the industrial production. In the future, textile-based TENGs have great potential in the extensive fields of high-efficiency energy harvesters, high sensitivity sensors, human–machine interaction, Internet of Things, environmental monitoring, and intelligent wearable.

5. Acknowledgments

This work was financially supported by the National Natural Science Foundation of China (Grant Nos. 52073051, 51873030, and 51703022), the National Key R&D Program of China (Grant No. 2018YFC2000900), the Natural Science Foundation of Shanghai (Grant No. 18ZR1402100), and Shanghai Committee of Science and Technology (Grant No. 19QA1400100).

Yan Zhang, Chris Bowen (*c.r.bowen@bath.ac.uk)

1. Abstract

Piezoelectric energy harvesting has attracted substantial interest from both academia and industry due to its potential to achieve long-lifespan sensing and self-powered autonomous operation of low-power electronics. Porous piezo-composites that combine a low dielectric constant with a high piezoelectric coefficient are promising materials with respect to an enhancement of energy harvesting capability. Here, we review recent achievements in the field of design and optimization of the material and provide insights and an outlook for the current and future challenges of the materials and their applications.

2. State-of-the-art

Energy harvesting, or energy scavenging, is a process that harvests small amounts of energy present in the ambient environment that would otherwise be wasted, such as heat, light, sound, vibration, or movement. This ambient energy can be converted into electricity for autonomous and self-powered low-power electronic devices and is a promising technology of interest to both academia and industrial applications to reduce or remove battery replacement and provide power in inaccessible or remote conditions. Among the range of harvesting approaches, piezoelectric energy harvesting converts oscillatory mechanical energy into electrical energy and is attracting due to its high energy conversion efficiency, ease of implementation, and potential for miniaturization.111 

If we consider a piezoelectric material in an energy harvester with the surface area of A, thickness of h, that is subjected to a mechanical force of F, the energy E generated can be generally estimated by E = 12CV2, where C and V are the material capacitance and the voltage between the opposite electrode surfaces of the piezoelectric element, respectively. Based on this, the energy E from a mechanical load F can be calculated as E = 12dij2ε0ε33TF2hA, where ε33T is the relative permittivity (or dielectric constant), ε0 is the permittivity of free space, and dij is the piezoelectric charge coefficient. Figures-of-merit that include the physical properties of the material have been developed from this simple analysis to assess the performance of energy harvesting materials for practical applications. For the selection and design of materials for piezoelectric energy harvesting, the harvesting figure-of-merit FoMij, derived from the energy described above and has been widely used,112–114 where FoMij=dij2ε0ε33T. The FoMij can be viewed as a direct indicator of the amount of piezoelectric energy harvested for a specific force, area, and thickness of material.

In order to achieve the desired optimized functional property with a high value of FoMij, significant effort has been made to improve the FoMij, such as chemical doping,115 growth of single crystal materials,116 and the combination of both;117 all of the above-mentioned methods mainly focus on the enhancement of the piezoelectric charge coefficient dij, which is a measure of the charge generated from an applied load. In many cases, the increase in the piezoelectric charge coefficient also leads to an increase in the dielectric constant, therefore limiting the degree of improvement in the FoMij=dij2ε0ε33T. Moreover, chemical doping is highly sensitive to the subtle changes in the stoichiometry of the material and needs complex processing steps due to the requirement for the close control of the composition for each element in the composite. In addition, single crystal materials normally exhibit poor mechanical properties and are often high cost and have a low Curie temperature, which can limit their operating temperature. Another promising approach to increase the FoMij is the careful introduction of porosity into the piezoelectric, which can lead to a large decrease in the dielectric constant while being able to simultaneously maintain a relatively high piezoelectric charge coefficient.113,118 Therefore, the formation of piezoelectric composites that consist of a piezo-active ceramic and a passive polymer, or air phase, can lead to high performance materials for energy harvesting applications.

There are five main preparation techniques to create energy harvesting materials based on porous piezoelectrics:119 the burnt-out polymer spheres method (BURPS), the replica template method, gel casting, freeze casting, and additive manufacturing. The common approach to realize an effective device is to make it as a piezo-composite with a 0-3-0, 3-0, 3-1, 3-2, 3-3, and 2-2 connectivity by introducing a polymer into the pore space. The polymer can either act as a porous structure with a piezoelectric active phase that dispersed within the polymer using ferroelectric particles, with pores as an air phase to create a 0-3-0 structure,120 or as a polymer that acts as a second phase as it infills the pore space of the porous piezoelectric ceramic to form a 3-0,121 3-1,122 3-2,123 3-3,124 or 2-2118 structure.

Based on the distribution of the pores in the composite, there are two types of pore structures, which can be regarded as isotropic and anisotropic architectures. In an isotropic porous piezoelectric composite, the pores are randomly distributed with a spherical-like or net-like morphology achieved via traditional processing techniques, such as gel casting,125 BURPS126,127 (“burned-out plastic spheres”), the replica method by coating the ceramic suspension on the wax/coral,128 or using a polymeric sponge129 that is burnt out during sintering, and additive manufacturing that allows for the fabrication of complex structures without the need for a mold.130 Gel casting has attracted attention due to its simplicity and low cost, but many of the organic solvents used in this process are toxic and carcinogenic. The BURPS process has the advantage of being an easy and low-cost fabrication method, while its disadvantages include the relatively poor dispersion of the additive in the ceramic powders and the potential of gases to form cracks/defects during the volatilization process. Additive manufacturing is gaining increasing attention since it can produce a wide range of shapes with geometrical complexity; however, the pores generated can be relatively large in size and the integrity between printed layers can limit mechanical or dielectric properties.

By tailoring the amount of porosity and pore morphology, the piezoelectric performance and dielectric constant can be controlled to achieve a combination of high piezoelectric activity and low dielectric constant. A near three-fold increase of the FoMij compared with the dense material was observed in an isotropic porous BaTiO3 composite with the porosity level of 60 vol. %.127 In order to increase the piezoelectric coefficient in a porous composite with reduced dielectric constant, a sandwich structure with “dense layer–isotropic pores–dense layer” arrangement has also been proposed.131,132 An optimum volume fraction of the porous layer was found to be 20 vol. %, with a relative thickness of the porous layer of 0.52 and the porosity within of ∼60 vol. %, which were shown to exhibit the highest piezoelectric energy harvesting capability.132 The typical morphologies of the isotropic pores and sandwich structure are shown in Figs. 5(a) and 5(b).

FIG. 5.

(a) PZT-coated polymeric net by applying several layers of ceramic suspension.129 (b) SEM of fracture surface of cross section of BaTiO3 with porous sandwich layer (porosity ∼60 vol. %).132 (c) Schematic of freeze casting technique and the corresponding pore morphology with aligned pore structure: i. homogeneous ceramic suspension, ii. the frozen ice crystals grow unidirectionally along the temperature gradient, iii. aligned porous ceramic after ice sublimation, and iv. aligned pore morphology after sintering.

FIG. 5.

(a) PZT-coated polymeric net by applying several layers of ceramic suspension.129 (b) SEM of fracture surface of cross section of BaTiO3 with porous sandwich layer (porosity ∼60 vol. %).132 (c) Schematic of freeze casting technique and the corresponding pore morphology with aligned pore structure: i. homogeneous ceramic suspension, ii. the frozen ice crystals grow unidirectionally along the temperature gradient, iii. aligned porous ceramic after ice sublimation, and iv. aligned pore morphology after sintering.

Close modal

Due to the inherently low mechanical properties, such as compressive strength, of the isotropic porous composite, new materials with aligned pore channels and exhibiting an anisotropic structure have been considered for improving both the mechanical properties and piezoelectric coefficient. An efficient route to achieve a highly anisotropic morphology is freeze casting, where ice crystals in a ceramic suspension [see Fig. 5(c-i)] are grown unidirectionally along a temperature gradient [see Fig. 5(c-ii)]. These ice crystals act as a replica of the aligned pores [see Fig. 5(c-iii)] in the piezoelectric composite, thereby forming an aligned pore morphology after ice-sublimation and sintering, as shown in Fig. 5(c-iv). The benefit of this structure is the improved connectivity of the piezo-active ceramic and the reduced electric field concentration in the pore space under the application of an external electric field during the poling process.113,121,132 As an example, an energy harvester fabricated from a highly aligned porous BaTiO3 ceramic exhibited a 2.4-fold increase in capacitor charging during piezo-energy harvesting compared to the dense material under the same conditions.113 A comparison of the harvesting Figure of Merit (FoMij) of porous ferroelectric materials is summarized in Table I.

TABLE I.

Comparison of figure of merit (FoM) for energy harvesting using porous ferroelectric materials.

MethodMaterial/compositePorosity vol. (%)d33 (pC/N)Relative permittivity, εrFoM (pC/N)2References
BURPS BCZT and air 10–25 285–424 1026–2158 8.9–9.4 133  
PZT-PCN and air 24–45.6 140–300 110–290 20.1–31.7 134  
PZT and air 35–54.5 161–312 241–1608 12.1–16.4 135  
PZT and air 5–45 208–350 300–1600 12.4–16.3 136  
BS-0.64PT 17.1 ∼485 400–1500 22.3 137  
PMN-PZT and air 33 510 ∼1580 18.6 138  
NKN and air 40 ∼100 ⋯ ⋯ 139  
BCZT and air 20 381 ∼3350 4.9 131  
BT and air 30 ∼124 ∼1100 1.6 132  
Gel-casting PZT and air 27.8–72.4 260–560 400–3500 10.1–19.1 125  
PZT and air 31.3–58.6 424–635 446–3418 13.3–45.5 140  
Freeze-casting PZT and air 28.1–68.7 608–690 1400–3500 15.4–29.8 141  
PZT-PZN and air ∼90 450 100–120 207.9 142  
PZT-PZN and air 50–82 380 284–853 29.9–57.4 143  
PZT and air 20–60 ∼350 ∼600 23.1 118  
NKNS ∼60.5 ∼130 ∼1319 1.45 144  
Direct ink writing PLZT ⋯ 481 ⋯ ⋯ 145  
PZT and epoxy resin ⋯ ∼360 ⋯ ⋯ 123  
KNN ⋯ 280 ⋯ ⋯ 146  
MethodMaterial/compositePorosity vol. (%)d33 (pC/N)Relative permittivity, εrFoM (pC/N)2References
BURPS BCZT and air 10–25 285–424 1026–2158 8.9–9.4 133  
PZT-PCN and air 24–45.6 140–300 110–290 20.1–31.7 134  
PZT and air 35–54.5 161–312 241–1608 12.1–16.4 135  
PZT and air 5–45 208–350 300–1600 12.4–16.3 136  
BS-0.64PT 17.1 ∼485 400–1500 22.3 137  
PMN-PZT and air 33 510 ∼1580 18.6 138  
NKN and air 40 ∼100 ⋯ ⋯ 139  
BCZT and air 20 381 ∼3350 4.9 131  
BT and air 30 ∼124 ∼1100 1.6 132  
Gel-casting PZT and air 27.8–72.4 260–560 400–3500 10.1–19.1 125  
PZT and air 31.3–58.6 424–635 446–3418 13.3–45.5 140  
Freeze-casting PZT and air 28.1–68.7 608–690 1400–3500 15.4–29.8 141  
PZT-PZN and air ∼90 450 100–120 207.9 142  
PZT-PZN and air 50–82 380 284–853 29.9–57.4 143  
PZT and air 20–60 ∼350 ∼600 23.1 118  
NKNS ∼60.5 ∼130 ∼1319 1.45 144  
Direct ink writing PLZT ⋯ 481 ⋯ ⋯ 145  
PZT and epoxy resin ⋯ ∼360 ⋯ ⋯ 123  
KNN ⋯ 280 ⋯ ⋯ 146  

In addition to the formation of porous ceramics, there has also been recent interest in forming porous polymers. A porous polymer with elongated unidirectionally pore channels can form a dipole-like structure by applying an external electric field that charges the pores to form a ferroelectret material. Such piezoelectrically active polymer composites are also promising for piezoelectric energy harvesting applications due to their combination of low permittivity and high piezoelectric coefficient.147 

3. Current challenges and future prospects

Over recent decades, flexible device technologies have developed at an unprecedented rate, resulting in the improvement of wearable, stretchable, bendable, foldable, and lightweight electronics. Polymer based piezoelectric composites have been widely employed due to their low permittivity, low density, and high flexibility. However, the low piezoelectric coefficient of piezoelectric polymers (for example, PVDF and its copolymers have a d33 ∼ 20 pC/N, compared with that of PZT ∼ 500–700 pC/N), together with the low working/Curie temperature, Tc (PVDF and its copolymers have Tc ∼ 100 °C, compared to PZT with a Tc > 300 °C) are the main factors that constrain the development of such materials in energy harvesting applications. For porous piezoelectric ceramics, while the introduction of porosity can provide some mechanical flexibility, effort is needed on the micro-structural design and polymer infiltration to improve the flexibility while retaining a high piezoelectric activity.148 Impregnating the pore space of a porous ceramic with a polymer of high dielectric constant and lower stiffness would also be beneficial to piezoelectric properties since the electric field concentration on the passive polymer can be reduced,121 and the applied mechanical load from the environment exerted on the active phase would not be shared by the passive polymer phase.

For practical applications on harvesting large-scale of energy sources, such as the vibrations of tall buildings, bridges, vehicle systems, railroads, ocean waves, and human motions, industrial scale up of processing methods is needed. Cracks/defects [such as those indicated in Fig. 5(b)], compositional homogeneity, and property stability are the main concern for the realization of the readily controlled materials and devices. Another challenge is that large-scale vibrations often vary with time and frequency, which makes efficient and reliable energy conversion difficult, in such a case, more efforts for the development of broader-band or off-resonance energy harvesting materials and systems are of interest.

There is also scope for the development of new materials, new architectures, and more efficient processing techniques for the creation of new ferroelectret candidates with moderate porosity, optimum ratio between the length and the width of pore channels, and enhanced piezoelectric capability.149 Commercially available ferroelectrets are currently made of polypropylene (PP, Emfit, Finland) with a d33 of 25–200 pC/N, where the pore space is formed by foaming process and stretching particulate filled polymers. Sandwich layer structures with a porous ferroelectret inner layer clamped between two dense polymer layers have also been realized by hot pressing, leading to an improved charge density. Hot pressing, therefore, provides an attractive route to create a variety of composite structures.150,151

The enhancement of the piezoelectric performance has been widely reported by forming single crystal with appropriate dopants; therefore, the formation of porous single crystal-like materials would truly combine a high piezoelectric activity with low dielectric constant for exceptional harvesting figures-of-merit. Li et al. presented Sm-doped Pb(Mg1/3Nb2/3)O3–PbTiO3 (Sm-PMN-PT) single crystals that exhibit a d33 of 3400–4100 pC/N, compared with the traditional PMN-PT, PZN-PT single crystals that have a d33 of 1200–2500 pC/N.117,152 It would be of interest to fabricate porous pseudo-single crystals to reduce the dielectric constant together with remaining this high piezo-response. This could be achieved by using freeze casting to align small single crystal templates [see Fig. 6(a)] and growing the templates during sintering to form complex porous architectures whose microstructure is single-crystal like,153 as shown in Fig. 6. Such freeze casting could also be combined with 3D printing techniques to obtain complex and novel physical or geometrical configurations with customized design of shape, high dij, and low ε33T for variety of needs in piezoelectric energy harvesting and sensing sectors.

FIG. 6.

Schematic of the fabrication of porous pseudo-single crystals via freeze casting. (a) Freeze casting small single crystal templates, (b) freeze drying the frozen green body, and (c) sintering for the realization of the porous pseudo-single crystals.

FIG. 6.

Schematic of the fabrication of porous pseudo-single crystals via freeze casting. (a) Freeze casting small single crystal templates, (b) freeze drying the frozen green body, and (c) sintering for the realization of the porous pseudo-single crystals.

Close modal

4. Concluding remarks

Piezoelectric composite materials that consist of the piezo-active and passive phases have been widely explored in energy harvesting applications due to their low dielectric constant and relatively high piezoelectric coefficient; such a combination of properties leads to a high harvesting figure-of-merit. The material selection and structural design have been utilized for the optimization of the energy harvesting capability. We expect significant progress in enhancing the poling efficiency of the porous composite by design of the pore space and impregnating a high dielectric constant but low stiffness polymer into the pore space. For harvesting large-scale vibrations, industrial scaling up of processing techniques and broader-band energy harvesting systems are of interest. The use of new materials, such as lead-free ferroelectrics and ferroelectrets, more efficient processing techniques, additive manufacture, and an ability to create porous pseudo-single crystals will be of increasing interest in the future development of this topic.

5. Acknowledgments

This work was supported by the Academy of Medical Sciences GCRF Fund (Grant No. GCRFNGR2-10059), ERC project (ERC-2017-PoC-ERC-Proof of Concept, Grant No. 789863), and The Leverhulme Trust (Grant No. RGP-2018-290). The authors acknowledge the Key Research and Development Project of Hunan Province (Grant No. 2020WK2004), Overseas Talent Introduction Project of China, and Hundred Youth Talents Program of Hunan.

Philippe Basset (*philippe.basset@esiee.fr), Ahmad Delbani, Hemin Zhang, Dimitri Galayko

1. Abstract

Triboelectric energy harvesters, as for any electrostatic transducer, need to maximize their bias voltage and capacitance variation for maximizing the electrical energy conversion from the mechanical domain. Internal (almost) steady bias of several hundreds of volts can be easily obtained directly by triboelectrification with a good triboelectric material, and even higher bias, increasing with time, can be obtained with a poor material if using unstable charge pumps for rectification. However, whatever the chosen strategy, the high-voltage/low-current output of the TENGs need to be adapted to a lower voltage/higher current output to power typical electronics.

This section presents an overview of the architectures for power management systems proposed to date for triboelectric nanogenerators to improve the global energy conversion conditions efficiency in the future.

2. Introduction

Triboelectric energy harvesters (TENG) typically generate high AC open-circuit voltage of several hundred volts peak, with relatively low short-circuit current of few tens of μA peak maximum. However, most applications require DC voltages around a few volts, with DC currents of at least a few hundred of μA in order to generate a minimum average power of 100 μW. Therefore, a power management circuit (PMC) is needed to generate a low DC voltage for the load, while implementing a high voltage interface at the TENG’s side to maximize the converted power. Hence, PMC for TENGs is an important topic that has recently attracted significant attention.

A complete PMC needs to fulfill two important tasks that are strongly interlinked: (1) the generation of a dynamic electrical bias on the TENG to optimize the mechanical-to-electrical energy transduction, while (2) performing the AC–DC conversion to a low DC voltage. For electrostatic transducers, such as TENGs, the converted power is proportional to the square of the bias voltage: hence, the first intention is to maximize the bias voltage level. A high bias is typically generated by the triboelectric effect, but may also be produced externally by some special conditioning circuits (CC), allowing high energy yield even with low cost TENG devices. In addition, the bias voltage generated by the CC on the TENG is necessarily variable and depends on the TENG motion: this represents the main challenge in the PMC implementation.

3. Passive charge pumps

Passive charge pumps are capacitive charge pumps where the switches are implemented with diodes (even if, for energy saving purpose, the diodes may be implemented with active switches154–156). These circuits can be classified into two categories: stable and unstable charge pumps. Stable charge pumps reach a constant saturation voltage after some time, for instance, full-wave and half wave rectifier, Cockroft–Walton multiplier, etc. On the contrary, unstable charge pumps, such as Bennet’s electricity doubler, generate a DC voltage that is continuously increasing, until component limitations or electrostatic discharge occur: such charge pumps are able to generate high voltage biasing even with basic TENGs having low internal polarization.

4. Stable charge pumps: The diode rectifier based conditioning circuits

The full-wave (FW) diode rectifier is the most used CC for TENGs that can be found in the literature. It is a stable charge pump generating a saturation voltage proportional to the maximum transferred charge in the short-circuit mode.157 The saturation voltage of the half-wave (HW) rectifier is higher by a factor (η − 1), where η is the Cmax/Cmin capacitance ratio of the TENG.158 Even higher saturation voltage can be obtained with a voltage multiplier, also known as the Cockroft–Walton doubler,159,160 which is a generalization of the HW network.

During the early cycles, as the output voltage is close to zero, the output voltage of the FW rectifiers increases with twice the slope of the HW, although in that case both powers are far from the optimum of each circuit. However, if the output voltage of the rectifier can be set to half of its saturation voltage, HW outperforms the FW by a factor (η + 1)/2.161 This can be seen from Fig. 7(c), where the extracted power is calculated as the time derivative of the energy of the load capacitor.

FIG. 7.

Circuits (a), spice simulations (b), and output power (c) of a half- and full-wave diode bridges, a Bennet doubler, and a Cockroft-multiplier (×3) for a TENG having a charge density of 50 μC/cm2 and a Cmax/Cmin ∼ 1 nF/300 pF.

FIG. 7.

Circuits (a), spice simulations (b), and output power (c) of a half- and full-wave diode bridges, a Bennet doubler, and a Cockroft-multiplier (×3) for a TENG having a charge density of 50 μC/cm2 and a Cmax/Cmin ∼ 1 nF/300 pF.

Close modal

5. Unstable charge pumps: The Bennet’s doubler family

A new class of CC, inspired from the electrical machines of the 18th century and also made of diodes and capacitors only, has recently emerged for TENGs conditioning: the Bennet doubler.162 These circuits have the ability to exponentially increase the charge on the TENG’s electrodes during operation and to increase its bias and conversion efficiency.163 A minimum value of η is necessary, which is typically 2 for the simplest architecture, but it can vary depending on the circuit configuration.164–166 While the output energy can be greatly improved by the unstable charge pumps, their main drawback is that it may take some time for reaching the conditions for a high energy conversion. This time is mostly depending on the value of η163 and the ratio between Cmax and the load capacitance. Figure 7 shows a simulation comparing the output rectified voltages and the output power for various CC.

6. Circuits based on active charge extraction

PMC based on synchronous techniques, such as SECE (Synchronous Electrical Charge Extraction) or SSHI (Synchronous Switching Harvesting on Inductor), consist in externally controlling the transducer’s current by using an inductive DC–DC convertor activated synchronously with the motion of the mobile electrode. They can be seen as an improvement of the passive FW conditioning circuit: Their purpose is to actively enhance the correlation between the transducer’s current and voltage to maximize the extracted power. The “charge extraction” term means that a negative (positive) current is generated at the phases of high positive (negative) voltage to maximize the average voltage–current product (the power). These circuits use a coil connected in series with a switch synchronously activated at the extremum voltages of the kinetic energy harvester, i.e., several times per mechanical cycle. We note that the term “synchronous” in their names is misleading: the passive conditioning circuits discussed earlier have also switches (diodes) activated synchronously with the mechanical motion, yet this synchronization is automatic, whereas in SECE and SSHI networks the synchronization needs to be external and requires sophisticated circuitry. Various implementations and variations of the basic SECE and SSHI techniques exist.167–169 In the recent applications with TENG, the DC–DC convertor used for charge extraction has been controlled electronically,170,171 by movement-induced electrostatic force172 or movement-induced mechanical contacts.155 

7. TENG generators for low voltage loads

A common limitation of the above-mentioned conditioning circuit is a high output voltage required for optimal energy conversion [cf. Figs. 7(b) and 7(c)], which is not compatible with the low voltage of the load supply. The FW and HW rectifiers work optimally when the output voltage is the half of the saturation voltage. The Bennet’s doubler and the SSHI converter optimal output voltage is the maximum voltage supported by the technology and the used components. The SECE circuit is an exception: the output voltage may be as low as required for the load. Because of the generally requested high voltage at the output of the conditioning circuit, two-stage architectures are required: the second stage is a DC–DC convertor achieving a voltage adaptation between the high output voltage of the primary conditioning circuit and the low voltage generated for the load.

The DC–DC converters used for the second stage are generally of two kinds. The first one, working in the continuous mode, performs an active impedance synthesis for the output of the conditioning circuit. This technique is mainly used for low coupling piezoelectric transducers used with full wave rectifiers to set the duty cycle or switching frequency of a Buck or Buck–Boost DC–DC converter at a frequency much higher than the mechanical oscillation.173 The term “continuous mode” is due to the fact that the current in the inductor never goes to zero. However, a high-frequency control of the switch may be complex to implement and is power consuming.

Therefore, many works employ a DC–DC conversion in the discontinuous mode: first, the circuit accumulate the harvested charges in the output capacitor of the TENG CC, then this capacitor is fully or partially discharged to the load capacitance through the DC–DC converter by activating the Buck switch at a frequency much lower than the mechanical frequency,174 and only when a significant energy is accumulated in the output capacitor of the TENG CC. This allows an event-based control of the DC–DC converter: the latter operates only when enough energy is harvested, and a significant reduction of the consumption needed for the power management is obtained. The event-driven control of the DC–DC convertor is achieved with a comparator provided with a hysteresis: either the hysteresis is narrow and then the CC output capacitor is discharged only partially or the hysteresis is wide and the output capacitor is fully discharged. The advantage of the narrow hysteresis is to keep the output voltage of the CC close to the optimal level. The wide hysteresis is easier to implement, it minimizes the activation frequency, but the CC operates suboptimally a non-negligible amount of time.

Moreover, a two-stage architecture allows the use a CC output capacitance of intermediate value (∼10–100 times the TENG’s capacitance) to minimize the set-up time required to reach the maximum harvested power. The output capacitor of the second stage (the load capacitor) is usually much larger (∼103–106 times the TENG’s capacitor), which provides a stabilization of the output voltage available for the load.

While actively driven MOS switches can be used to control the operation of the DC–DC stage of the power management circuit, recently self-powered/self-actuated switches have been proposed: a switch with wide hysteresis made of a SCR thyristor and a Zener diode,175 and a high-voltage micro-plasma switch with a narrow hysteresis, to minimize the period while the TENG is biased at low voltage (Fig. 8).176 

FIG. 8.

A two-stage CC with an unstable charge-pump rectifier and high-voltage autonomous micro-plasma switch.176 

FIG. 8.

A two-stage CC with an unstable charge-pump rectifier and high-voltage autonomous micro-plasma switch.176 

Close modal

8. Comparison and choice criteria of power management architectures

The circuits described earlier have different features regarding the performances and the facility of implementation. Depending on the available triboelectric device and the design objective, one of them should be preferred. In the following, we list some typical situations that a TENG generator designer must face, and we provide guidance for selecting the most appropriate conditioning circuit.

Passive charge pumps: The great advantage of the passive charge pumps is their simplicity of implementation. However, they all need an additional stage in order to control the voltage at its output (with two-stage architectures or an LDO–low drop voltage–regulator).

The full bridge rectifier should be preferred when the output voltage [Vout in Fig. 7(a)] is much lower than VTE, the build-in voltage of the TENG that is assumed to be high. Under this condition, this circuit provides the highest conversion power over all possible passive charge pumps. The half bridge rectifier should be preferred when VTE is high and high Vout is possible. However, this usually involves additional DC–DC conversion to obtain a low voltage supply for the load.

When the built-in voltage of the TENG is low, unstable charge pumps should be preferred, as half-bridge and full-bridge circuits would have poor performance. In this case, the unstable charge pump increases the output voltage Vout up to the optimal (high) value, independently on VTE. As with the half-wave rectifier, these circuits must be supplemented with a DC–DC converter to generate a low voltage for the load.

Circuits based on the active charge extraction (typically, SECE) provides a very good compromise between efficiency and the complexity of the control. Its advantage is to generate a low voltage on the load without requiring an additional DC–DC stage. It should be preferred when the TENG voltage is high and when the technology used for the implementation of the circuit allows the use of a complex control involving an actively controlled switch, an internal inductive DC–DC convertor, etc. Usually, this involves using CMOS technology (on-chip integration) to minimize the parasitics and optimize overall circuit operation.

Two stage power management architectures are needed when the operational voltage of the TENG (the voltage on Cbuf in Fig. 8) is much higher than the voltage on the load supply on Cstore. Such architectures are necessary if we wish to maximize the energy extracted from the TENG by biasing it with a high voltage (on Cbuf). The classical configuration shown in Fig. 8 corresponds to the combination of an unstable charge pump and of a DC–DC convertor regulating the voltages on Cbuf. Note that a DC–DC convertor can only regulate one voltage, on Cbuf or Cstore. If both are to be regulated, the DC–DC convertor is designed to regulate the voltage on Cbuf, and an additional stage (an LDO or another load management circuit) must be used for the regulation of the load voltage.

9. Conclusion

The conditioning electronics for TENGs remains a challenge that will continue to be addressed in the future. The main challenge is to match the high output voltage required for the biasing TENGs with the low voltage needed to power the load in most applications. Reconciling of these requirements requires the use of a two-stage power interface: the first stage is a conditioning (primary) circuit that generates power and the second stage is a DC–DC that matches the output voltage to the load. Variations in operating conditions may require the architectures to be adaptive. Adaptive behavior is typically implemented in the second stage (the DC–DC converter), which monitors the operation of the primary conditioning circuit and keeps it optimal. The high operating voltage of TENGs makes this task difficult to implement. In addition, another fundamental difficulty is the minimizing of the leakage in the system: at 100 V, 1 µA leak results in a power loss of 100 µW, which can be prohibitive.

In all cases, the two-stage power interface includes a switch to control the charge transfer between the two stages. Regardless of the switch technique chosen, the switch will induce significant losses with each actuation. Mechanical contact switches are easy to implement because they require no electrical control and can be operated precisely at the extreme variations in the TENG voltage. However, they are actuated at least once per mechanical cycle, which does not allow for charge accumulation over several cycles and therefore generates a lot of losses because of their high actuation frequency. Electronic switches need additional energy for their “cold” start-up, control, and actuation. They also drastically limit the allowable output voltage across the TENG, especially with IC technologies, which greatly affects the energy conversion. On the other hand, MPPT can easily be implemented to adapt the system to irregular mechanical inputs. Plasma switches can store energy up to very high voltages, they are self-actuated, and even a narrow hysteresis can be implemented by design. However, they also generate losses and their hysteresis cannot be matched to the external mechanical force, unless the gap between the switch electrodes can be adjusted by some MPPT electromechanical systems.

Feng Ru Fan (*frfan@xmu.edu.cn), Wenzhuo Wu (*wenzhuowu@purdue.edu)

1. Abstract

Nanogenerators and piezotronics have emerged as promising candidates to meet the needs in harvesting and interfacing the mechanical signals in numerous emerging technologies. The recent advances in atomically thin materials have promoted the development of 2D materials based nanogenerators and piezotronics with ultrathin form factors. In this roadmap, we intend to provide a brief discussion focusing on our perspectives on prospects and challenges associated with the fields of 2D materials for nanogenerators and piezotronics. The development of convergent, trans-disciplinary approaches is expected to remove the barriers in the design, synthesis, integration, characterization, and application of 2D materials for nanogenerators and piezotronics. Such collective efforts from the research community would also stimulate extensive investigations for piezoelectricity, electronic transport, triboelectricity, as well as many other scientific and technological aspects of atomically thin materials.

2. Background and state-of-the-art

The capabilities of devices to scavenge, detect, and interact with a rich spectrum of ubiquitous mechanical signals (e.g., force, strain, vibration, etc.) would allow the next-generation machines to sense, monitor, and communicate with the environment with greater intelligence in a multitude of emerging technologies, e.g., wearable devices, soft robotics, medical prosthetics, and human–machine interface.177–181 Such capabilities are largely lacking in the state-of-the-art technologies for mechanical harvesters/sensors.182,183 To this end, emerging technologies, such as nanogenerators and piezotronics, have attracted intensive interest. Wang and Song invented the piezoelectric nanogenerator (PENG) for harvesting the mechanical vibrations into electricity.1 Wang and co-workers further demonstrated the concept of piezotronics, where the mechanically induced piezoelectric polarization function as the controlling signal.21,23 Early studies in PENG and piezotronics were mainly exploring piezoelectric nanowires (e.g., ZnO and GaN).

The recent advances in scientific understandings and technological applications of 2D materials have promoted the development of nanogenerators and piezotronics, leveraging the unique characteristics of these materials.184 2D materials’ atomically thin geometries and superior mechanical properties enable the introduction of enormous strain (e.g., >1%) without fracture, and broad tunability inaccessible to bulk or thin-film materials.185 2D semiconductors and dielectrics (e.g., MoS2, BN, etc.)186–200 with noncentrosymmetric lattices have received significant attention owing to their unique electrical, optical, and other physical and chemical properties. Due to the strain-induced lattice distortion and the associated charge polarization, these 2D materials exhibit piezoelectricity. They are capable of direct mechanical-to-electrical transduction for static and dynamic signals and appealing for PENGs and piezotronics with ultrathin form factors. It should be noted that the piezoelectricity is closely related to the structure and thickness of 2D materials. For PENG, single-layer 2D materials with broken inversion symmetry has a strong intrinsic piezoelectric response, whereas centrosymmetric bilayers and bulk crystals are non-piezoelectric.189 For TENG, the selection range of 2D materials is relatively wide, which can be either single-layer or bulk crystals. Figure 9 shows the chronological timelines for the milestones in the experimental efforts on 2D material based nanogenerators (PENGs and TENGs) and piezotronics.189,197,201–204 We would like to point out that, due to the space limit, these only represent a small fraction of representative work in the related fields. Instead of comprehensively reviewing the progress in related fields, which can be found in recent publications, here we intend to provide a brief discussion focusing on our perspectives on prospects and challenges in related fields.

FIG. 9.

The chronicle timeline showing the experimental milestones in 2D materials nanogenerators (above the timeline) and 2D materials piezotronics (below the timeline). (a) The first PENG based on the piezoelectric ZnO nanowires. Adapted with permission from Z. L. Wang and J. Song, Science 312, 242 (2006). Copyright 2006 AAAS, USA. (b) The first TENG. Adapted with permission from Fan et al., Nano Energy 1, 328 (2012). Copyright 2012 Elsevier Ltd. (c) The first 2D materials-based PENG based on monolayer MoS2. Adapted with permission from Wu et al., Nature 514, 470 (2014). Copyright 2014 Springer Nature. (d) Transparent TENG used monolayer graphene as electrode. Adapted with permission from Kim et al., Adv. Mater. 26, 3918 (2014). Copyright 2014 Wiley-VCH. (e) MoS2-based PENG. Adapted with permission from Kim et al., Nano Energy 22, 483 (2016). Copyright 2016 Elsevier Ltd. (f) PENG based on chemical-vapor-deposition grown monolayer WSe2. Adapted with permission from Lee et al., Adv. Mater. 29, 1606667 (2017). Copyright 2017 Wiley-VCH. (g) TENG based on monolayer MoS2 nanocomposites. Adapted with permission from Wu et al., ACS Nano 11, 8356 (2017). Copyright 2017 American Chemical Society. (h) Triboelectric series of 2D materials. Adapted with permission from Seol et al., Adv. Mater. 30, 1801210 (2018). Copyright 2018 Wiley-VCH. (i) Textile-TENG based on BP composites fabrics. Adapted with permission from Xiong et al., Nat. Commun. 9, 4280 (2018). Copyright 2018 Springer Nature. (j) Flexible PENG based on multilayer BP. Adapted with permission from Ma et al., Adv. Mater. 32, 1905795 (2020). Copyright 2020 Wiley-VCH. (k) Flexible PENG based on SnS. Adapted with permission from Khan et al., Nat. Commun. 11, 3449 (2020). Copyright 2020 Springer Nature. (l) The first piezotronic device based on ZnO nanowires. Adapted with permission from Z. L. Wang, Adv. Mater. 19, 889 (2007). Copyright 2007 Wiley-VCH. (m) The first piezo-phototronic device based on ZnO nanowires. Adapted with permission from Hu et al., ACS Nano 4, 1234 (2010). Copyright 2010 American Chemical Society. (n) The first 2D materials-based piezotronics prepared with the monolayer MoS2. Adapted with permission from Wu et al., Nature 514, 470 (2014). Copyright 2014 Springer Nature. (o) The direct measurement of piezoelectric coefficient in monolayer MoS2. Adapted with permission from Zhu et al., Nat. Nanotechnol. 10, 151 (2015). Copyright 2015 Springer Nature. (p) The piezoelectricity in 2D graphene nitride. Adapted with permission from Zelisko et al., Nat. Commun. 5, 4284 (2014). Copyright 2014 Springer Nature. (q) The first piezo-phototronic device based on monolayer MoS2. Adapted with permission from Wu et al., Adv. Mater. 28, 8463 (2016). Copyright 2016 Wiley-VCH. (r) Piezoelectricity in CVD-grown monolayer MoS2. Adapted with permission from Qi et al., Nat. Commun. 6, 7430 (2015). Copyright 2015 Springer Nature. (s) Piezoelectricity in Janus monolayers of transition metal dichalcogenides. Adapted with permission from Lu et al., Nat. Nanotechnol. 12, 744 (2017). Copyright 2017 Springer Nature. (t) Piezoelectricity in bilayer WSe2. Adapted with permission from Lee et al., Adv. Mater. 29, 1606667 (2017). Copyright 2017 Wiley-VCH. (u) Piezoelectricity in a-In2Se3 nanoflakes. Adapted with permission from Zhou et al., Nano Lett. 17, 5508 (2017). Copyright 2017 American Chemical Society. (v) 2D piezotronics in ZnO nanosheets. Adapted with permission from Wang et al., Nano Energy 60, 724 (2019). Copyright 2019 Elsevier Ltd. (w) Piezoelectricity in multilayer black phosphorus. Adapted with permission from Ma et al., Adv. Mater. 32, 1905795 (2020). Copyright 2020 Wiley-VCH.

FIG. 9.

The chronicle timeline showing the experimental milestones in 2D materials nanogenerators (above the timeline) and 2D materials piezotronics (below the timeline). (a) The first PENG based on the piezoelectric ZnO nanowires. Adapted with permission from Z. L. Wang and J. Song, Science 312, 242 (2006). Copyright 2006 AAAS, USA. (b) The first TENG. Adapted with permission from Fan et al., Nano Energy 1, 328 (2012). Copyright 2012 Elsevier Ltd. (c) The first 2D materials-based PENG based on monolayer MoS2. Adapted with permission from Wu et al., Nature 514, 470 (2014). Copyright 2014 Springer Nature. (d) Transparent TENG used monolayer graphene as electrode. Adapted with permission from Kim et al., Adv. Mater. 26, 3918 (2014). Copyright 2014 Wiley-VCH. (e) MoS2-based PENG. Adapted with permission from Kim et al., Nano Energy 22, 483 (2016). Copyright 2016 Elsevier Ltd. (f) PENG based on chemical-vapor-deposition grown monolayer WSe2. Adapted with permission from Lee et al., Adv. Mater. 29, 1606667 (2017). Copyright 2017 Wiley-VCH. (g) TENG based on monolayer MoS2 nanocomposites. Adapted with permission from Wu et al., ACS Nano 11, 8356 (2017). Copyright 2017 American Chemical Society. (h) Triboelectric series of 2D materials. Adapted with permission from Seol et al., Adv. Mater. 30, 1801210 (2018). Copyright 2018 Wiley-VCH. (i) Textile-TENG based on BP composites fabrics. Adapted with permission from Xiong et al., Nat. Commun. 9, 4280 (2018). Copyright 2018 Springer Nature. (j) Flexible PENG based on multilayer BP. Adapted with permission from Ma et al., Adv. Mater. 32, 1905795 (2020). Copyright 2020 Wiley-VCH. (k) Flexible PENG based on SnS. Adapted with permission from Khan et al., Nat. Commun. 11, 3449 (2020). Copyright 2020 Springer Nature. (l) The first piezotronic device based on ZnO nanowires. Adapted with permission from Z. L. Wang, Adv. Mater. 19, 889 (2007). Copyright 2007 Wiley-VCH. (m) The first piezo-phototronic device based on ZnO nanowires. Adapted with permission from Hu et al., ACS Nano 4, 1234 (2010). Copyright 2010 American Chemical Society. (n) The first 2D materials-based piezotronics prepared with the monolayer MoS2. Adapted with permission from Wu et al., Nature 514, 470 (2014). Copyright 2014 Springer Nature. (o) The direct measurement of piezoelectric coefficient in monolayer MoS2. Adapted with permission from Zhu et al., Nat. Nanotechnol. 10, 151 (2015). Copyright 2015 Springer Nature. (p) The piezoelectricity in 2D graphene nitride. Adapted with permission from Zelisko et al., Nat. Commun. 5, 4284 (2014). Copyright 2014 Springer Nature. (q) The first piezo-phototronic device based on monolayer MoS2. Adapted with permission from Wu et al., Adv. Mater. 28, 8463 (2016). Copyright 2016 Wiley-VCH. (r) Piezoelectricity in CVD-grown monolayer MoS2. Adapted with permission from Qi et al., Nat. Commun. 6, 7430 (2015). Copyright 2015 Springer Nature. (s) Piezoelectricity in Janus monolayers of transition metal dichalcogenides. Adapted with permission from Lu et al., Nat. Nanotechnol. 12, 744 (2017). Copyright 2017 Springer Nature. (t) Piezoelectricity in bilayer WSe2. Adapted with permission from Lee et al., Adv. Mater. 29, 1606667 (2017). Copyright 2017 Wiley-VCH. (u) Piezoelectricity in a-In2Se3 nanoflakes. Adapted with permission from Zhou et al., Nano Lett. 17, 5508 (2017). Copyright 2017 American Chemical Society. (v) 2D piezotronics in ZnO nanosheets. Adapted with permission from Wang et al., Nano Energy 60, 724 (2019). Copyright 2019 Elsevier Ltd. (w) Piezoelectricity in multilayer black phosphorus. Adapted with permission from Ma et al., Adv. Mater. 32, 1905795 (2020). Copyright 2020 Wiley-VCH.

Close modal

3. Challenges and opportunities

Despite the rapid progress achieved in related fields, roadblocks exist for the synthesis, integration, characterization, and application of 2D materials based nanogenerators and piezotronics. One of the biggest challenges in exploring the known 2D piezoelectric materials for practical applications is to induce consistent piezoelectric responses (e.g., aligned polarity) in all the 2D flakes using the external macroscopic mechanical strains (e.g., those induced by the substrates). Such a challenge is primarily due to the following facts: (1) most of these materials possess symmetries that lead to in-plane piezoelectricity; (2) the ionic nature of these compound materials results in orientation-dependent piezoelectric responses; and (3) these materials are often placed on the host substrate with poor or little control over the in-plane orientations due to the limitations in synthesis and/or assembly control. Many of the experimentally explored 2D piezoelectrics have small piezo-coefficients, making applications elusive. Although a wide range of 2D materials has been theoretically predicted to exhibit intrinsic piezoelectricity,186,187,198,199 most of these materials have yet to be investigated experimentally due to the challenges associated with the current approaches for preparing those materials, e.g., exfoliation and vapor depositions. These ongoing efforts have vague potential in scaling-up with desirable yield and material properties (e.g., shape, orientation, dimensions, carrier density, etc.), and suffer from the restrictions in growth substrates (e.g., for epitaxy) and process conditions (e.g., atmospheric control due to material instability). Also, the symmetry in most 2D piezoelectrics results in a thickness-dependence that imposes formidable challenges for the process control, where the piezoelectricity can disappear when the material thickness varies by only one atomic layer. The atomic thickness of the 2D materials also renders their sensitivity (and instability) to the environmental conditions. Moreover, the carrier concentration of 2D materials, which can significantly impact the piezoelectric property through charge screening, is prone to significant variations due to unintentional environmental doping. Such a challenge necessitates the development of stable 2D materials through the optimization of material properties, device structure, and the incorporation of the packaging process. An extensive understanding of the mechanism of electromechanical transduction and charge transfer at the atomic scale, which is still lacking, will facilitate the optimization of the output/sensing performance of related devices. The ultrathin nature and robust mechanical properties of 2D materials allow for potential integration into 3D architecture with diverse functionalities and boosted performance compared to planar counterparts. Still issues such as the electrical interconnection across layers and the robustness of the manufacturing process need to be addressed appropriately. The search, design, and production of 2D materials with robust out-of-plane piezoelectricity is favored for such 3D integration.192,193,204

These challenges also lead to abundant research opportunities. The reduction of dimensionality in 2D materials leads to strong, accessible piezoelectricity compared to its bulk counterpart. 2D piezoelectric materials are unique when compared to traditional piezoelectrics, which are brittle and insulating, and can host significant strain-induced electric field couplings through extremely strong Coulomb interactions in the 2D limit. Such characteristics offer unexplored possibilities for probing intriguing science in the atomically thin limit due to the strong coupling of piezoelectricity to various solid-state excitations involving charges, photons, and spins in 2D systems for engineering novel functionalities.205–207 From a practical application point of view, 2D piezoelectrics have been explored as the active components for electromechanical sensors,208 the optically active layers for strain-engineered optoelectronics,194 and the interfaces for mechanically enhanced catalysis.209 2D materials, in general, can also be used to construct efficient triboelectric nanogenerators (TENGs)2 due to their flexible structures and tunable surface/dielectric properties for engineered triboelectrification and electrostatic induction.110,210–212 Such flexibility allows for the design and implementation of ultrathin triboelectric devices using 2D materials. The thin thickness of 2D materials also favors the occurrence/observation of flexoelectricity, a physical process that can induce electrical polarizations when a material is subjected to an inhomogeneous deformation.213 The knowledge in the impact of flexoelectricity on the nanogenerator and piezotronics operations, as well as in 2D materials with centrosymmetry,214 is expected to guide the design of future energy and sensor devices with enhanced performance.

The advances in material design and nanomanufacturing can enable new technology for scalably producing substrate-agnostic, high-performance 2D materials with designer properties. A comprehensive examination, through combined theoretical and experimental efforts, of the interfacial characteristics between 2D materials and electrodes will provide the fundamental understandings, e.g., the effects of metals and electrode configurations on the charge transfer/transport in metal-2D material contacts.215 Such knowledge is critical for the rational design and optimization of future nanogenerators and piezotronics. The exploration of the fundamental doping mechanism and defect chemistry in 2D materials is essential for not only providing versatility in modulating the material properties by design216 but may also enable novel device concepts and applications in nanogenerators and piezotronics.217,218 The development of quantitative in situ or even in operando characterizations of 2D materials (e.g., for piezoelectricity, ferroelectricity, triboelectrification, and charge transfer) under controlled complex straining conditions, e.g., biaxial strain and patterned strain field,219 will add to the critically required fundamental insights and instrumental toolbox toward physics-based design and development of 2D nanogenerators and piezotronics. Last but not least, the design and fabrication of 2D heterostructured artificial crystals can introduce more device physics and diversified functionalities.218 The introduction of additional encapsulation layers (e.g., transferred boron nitride220 or grown dielectrics221) can improve the stability of nanogenerator piezotronics devices based on 2D materials.

4. Conclusions

The development of convergent, trans-disciplinary approaches that bridge disciplines such as advanced manufacturing, data science, material science, device physics, and chemistry is expected to spur data-driven, physics-based theoretical and experimental advances in 2D materials based nanogenerators and piezotronics. Such a confluence of collective efforts from the research community would stimulate extensive investigations for piezoelectricity, electronic transport, ferroelectricity, triboelectricity, and many other scientific and technological aspects of atomically thin materials.

5. Acknowledgments

W.W. acknowledges the College of Engineering and School of Industrial Engineering at Purdue University for the startup support and the Ravi and Eleanor Talwar Rising Star Assistant Professorship. F.R.F. acknowledges Nanqiang Young Top-notch Talent Fellowship from Xiamen University.

Miso Kim (*smilekim@skku.edu)

1. Abstract

Metamaterials are artificially engineered structures capable of yielding effective material properties not found in nature, which leads to desirable wave control functionalities not only in optics but also in acoustics and vibration regimes. Metamaterial-based energy harvesting has recently emerged as an enabling technology for drastic enhancement of harvesting performance by manipulating and amplifying input mechanical sources, such as vibration, sound, and elastic waves. Here, we review major achievements in the field of metamaterials-based energy harvesting using various kinds of phononic crystals and metamaterials, including phononic crystals with defects, gradient-index phononic crystals, locally resonant metamaterials, metasurfaces, and mechanical metamaterials. As in the early stage of development, metamaterial-based energy harvesting research entails challenges, which again form the basis for further development. Emergent directions, such as innovative metamaterial designs for broadband operation within a compact size, artificial intelligence-based design algorithms, and interface technology for efficient energy transfer between the metamaterial and energy harvesting devices, are addressed together with relevant challenges.

2. State-of-the-art

Energy harvesting (EH) technology has experienced a robust and remarkable development, thanks to high performance mechanical-to-electrical conversion materials, devices, and efficient power management circuits. Along with disruptive materials and systems for energy conversion and management,222,223 researchers recently began to turn their eyes toward the idea of actively manipulating and amplifying input mechanical wave energies via metamaterials in order to drastically improve energy harvesting performance. As directly implied from the prefix meta that means “beyond” in original Greek, metamaterials are artificially designed structures that can exhibit material properties beyond the conventional scheme, such as negative mass density, negative bulk modulus, negative refractive index, and bandgap. Metamaterials derive their wave manipulation capability from these unconventional properties, providing a wide range of potential applications including super- and hyper-focusing lenses, cloaking, wave filtering, and absorbers in both optics and acoustics.224–227 Here, we focus on metamaterials that are related to mechanical waves, such as vibration, sound, and elastic waves. Note that the term “metamaterial” can include phononic crystals, acoustic/elastic, and mechanical metamaterials in a broader sense while “metamaterial” or “acoustic metamaterial” sometimes denotes only the artificially engineered composite materials with functionalities due to local resonances. Metamaterials have provided a new paradigm for energy harvesting technology in that it is about active control of input wave energies rather than the conventional energy harvesting research focus: conversion and power management. When mechanical waves are amplified either through wave localization or focusing using metamaterials, subsequent mechanical-to-electrical energy conversion and management through high performance piezoelectric materials, devices, and harvesting power management circuits can make a synergetic effect to yield substantially enhanced harvesting performance.228,229 Here, we review the present state-of-the art in metamaterial-based energy harvesting using various phononic crystal and metamaterial concepts ranging from phononic crystals, acoustic metamaterials, and metasurfaces to mechanical metamaterials.

Phononic crystals (PnCs) or sonic crystals are artificial structures consisting of a periodic repetition of unit cells. Similar to any periodic structure, interaction of acoustic waves with PnC structures results in a dispersive band structure with a bandgap due to Bragg’ scattering, which is governed by the Bloch and Floquet theorem.230 The characteristic length, i.e., unit cell size, of the PnC is of the same order of magnitude as the wavelength: a few millimeters to centimeters for acoustic and elastic waves. Within the bandgap frequency range, the energy trapped can be localized at a defect that is introduced inside a PnC structure, which can subsequently be converted into electrical energy by attaching a piezoelectric energy harvesting device at the defect. Locating a piezoelectric PVDF film in the cavity of PnCs consisting of periodic cylinders has proved useful for substantial enhancement in both vibration and acoustic energy harvesting when compared with the case without PnCs.231–233 Similarly, elastic wave energy can be localized at the imperfection of a two-dimensional lattice constituted by stubs,234 circular,235 or octagonal holes236 in a metal thin plate (e.g., aluminum), leading to substantial amplification of harvesting performance. As highly dense energy localization occurs at the defect resonance frequency inside a phononic bandgap range, a systematic unit cell design for bandgap maximization is crucial for enhanced PnC-based energy harvesting. In this regard, octagonal hole-type unit cells with its optimized dimensions in an aluminum plate were numerically and experimentally demonstrated to yield more than 20 times of power enhancement when coupled with a PZT ceramic disk in comparison with a bare plate [Figs. 10(a) and 10(b)].236 Parametric studies on the physical relation of PnC design parameters, such as material properties, supercell size (i.e., number of unit cells), and defect location, with PnC-based EH system characteristics, including bandgap size and frequency level, as well as mechanical and electrical output performance of elastic wave energy harvesting.234,237,238 PnC-based energy localization and harvesting study has not been limited to a single defect. Creating double or more defects inside a PnC structure has unveiled different physical mechanisms not found in a single defect case, such as defect band splits and corresponding defect resonance modes, which can provide more versatile ways of exploring PnCs for broadband and tunable energy harvesting.239 Instead of using only sonic crystals, researchers proposed a coupled resonance structure of sonic crystals and Helmholz resonators (HRs), the latter of which is one of the most conventional ways to amplify the sound pressure.240–242 Hierarchical sonic crystals consisting of two different levels of periodic orders of PnCs are also an intriguing design proposed for broadband gaps and acoustic energy confinement, exhibiting potentials for broadband sound energy harvesting.243–245 Phononic crystal cantilever beams coupled with a PEH device246,247 or periodic arrangement consisting of piezoelectric cantilevers248 have also been used to take advantage of both cantilever resonance and bandgap due to periodicity simultaneously.

FIG. 10.

Various concepts of metamaterials for mechanical wave energy focusing and energy harvesting: (a) harmonic simulation results showing energy localization at the defect of octagonal-hole type phononic crystals and (b) the resulting amplified output power, (c) an image of fabricated omnidirectional gradient-index (GRIN) phononic crystals (PnCs), and (d) its harmonic analysis results to show sound energy amplification at the center of the GRIN-PnCs, (e) numerical sound pressure field and (f) the output power of the two-sided multilateral metasurface. Figures taken with permission from Park et al., Nano Energy 57, 327 (2019). Copyright 2019 Elsevier Ltd. All rights reserved. Figures from Hyun et al., Appl. Phys. Lett. 116, 234101 (2020). Copyright 2020 with permission of AIP Publishing and S. Qi and B. Assouar, Appl. Phys. Lett. 111, 243506 (2017). Copyright 2017 with permission of AIP Publishing.

FIG. 10.

Various concepts of metamaterials for mechanical wave energy focusing and energy harvesting: (a) harmonic simulation results showing energy localization at the defect of octagonal-hole type phononic crystals and (b) the resulting amplified output power, (c) an image of fabricated omnidirectional gradient-index (GRIN) phononic crystals (PnCs), and (d) its harmonic analysis results to show sound energy amplification at the center of the GRIN-PnCs, (e) numerical sound pressure field and (f) the output power of the two-sided multilateral metasurface. Figures taken with permission from Park et al., Nano Energy 57, 327 (2019). Copyright 2019 Elsevier Ltd. All rights reserved. Figures from Hyun et al., Appl. Phys. Lett. 116, 234101 (2020). Copyright 2020 with permission of AIP Publishing and S. Qi and B. Assouar, Appl. Phys. Lett. 111, 243506 (2017). Copyright 2017 with permission of AIP Publishing.

Close modal

Controlling directional nature of acoustic and elastic wave propagation toward the targeted location has also been a subject of significance for focusing and harvesting. Gradient-index (GRIN) phononic crystals, a periodic arrangement of unit cells with a spatial variation of the refractive index, have proved to be effective to manipulate the wave propagation direction for focusing both in acoustic249,250 and elastic regimes.251–254 Gradually changing indices in a GRIN PnC can be achieved by tailoring the geometric dimensions of the unit cell (e.g., the size of circular or cross-shaped holes and cylindrical stubs) that satisfy the desired target refractive index profile. When a piezoelectric device is integrated into the GRIN PnC structure particularly at the point of focusing [Figs. 10(c) and 10(d)], the focused energy can then be converted into substantially amplified electrical energy, exhibiting the wave manipulation capability in terms of direction as well as amplitude. Embedded acoustic black holes are another design framework that allows producing high energy density by tailoring the wave propagation characteristics. Tuning the local stiffness and thus wave velocity is possible by embedding structural tapering into the host structure. As vibration energy can be trapped in the tapered acoustic black hole area, much enhanced harvesting is attainable.255–258 

Locally resonant metamaterials are another innovative engineered structures that can possess extraordinary physical properties such as negative elastic modulus, negative mass density and negative refractive index.226,259–263 Such unconventional features are associated with the phenomenon such as bandgap or negative refraction, which enables input acoustic wave localization and focusing for enhanced harvesting. Since PnCs require the unit cell size that is comparable to the wavelength, low frequency acoustic energy localization and harvesting can only be achieved with considerable lattice constants, inevitably causing volume issues for practical applications. In contrast, sub-wavelength structural designs of acoustic metamaterials (AMMs) based on local resonance allow us to achieve the desired functionalities with the benefit of much reduced operating volume. Considering the importance of spatial efficiency and easy fabrication, a planar defected AMM design was reported with the theoretical analysis on its acoustic bandgap, wave localization, and harvesting performance.264 Ma et al. reported the dual functionality of a deep sub-wavelength membrane-type AMMs where hybridized resonances served the total sound absorption with no reflection with an additional harvesting functionality.265 Similar membrane-type AMM designs with the dual functionality of noise insulation and harvesting follow as in Refs. 266 and 267. A coiling-up space structure offers a feasible way to tune and elongate the propagating path within a limited volume. A subwavelength-scale acoustic metastructure, consisting of an array of doubly coiled-up AMM cavities, was proposed for the strong confinement and harvesting of the sound energy.268 Recent advances include a helix structure-based AMM,269,270 a compact dual-layer of AMM271 for low frequency acoustic energy harvesting, and mechanical locally resonant metastructures coupled with piezoelectric energy harvesters for enhanced vibration harvesting.272 

The long wavelength of acoustic waves that can range up to the meter scales poses a fundamental limit on the miniaturization of relevant acoustic and elastic applications, including energy harvesting. Recently, the concept of a metasurface, a patterned planar structure with a deep subwavelength dimension, has emerged from the desire to manipulate and control acoustic waves on a subwavelength scale.262,273,274 Phase control of the reflected or transmitted waves by tailoring material or structural designs is the key to the wave manipulation of metasurfaces. Metasurfaces opened doors for tremendous opportunities to realize novel wave functionalities including anomalous refraction,275 total reflection,276 full transmission,277 and total absorption,278,279 which enable perfect absorbers, broadband focusing lenses, acoustic cloaking, diodes, and vortexes. Various metasurface designs, not many yet though, have recently been demonstrated to implement acoustic or elastic wave focusing, energy confinement, and harvesting as well, including multilateral metasurfaces consisting of labyrinthine units280 [Figs. 10(e) and 10(f)] and graded resonant metasurfaces based on rainbow trapping.281 

One more important development we cover here is mechanical metamaterials. Mechanical metamaterials are architected materials that can exhibit unconventional properties or functionalities previously inaccessible, including but not limited to negative Poisson’s ratio,282,283 tunable stiffness,284 and/or mass density, and even pattern and shape reconfigurability.285,286 Auxetic metamaterials,287 pentamode metamaterials,288,289 and origami-based and kirigami-based metamaterials290,291 are the representative mechanical metamaterials potentially useful for extensive structural and functional applications, which is well reviewed in Ref. 292. Pertaining to utilizing mechanical metamaterials for energy harvesting, only a few case studies can be found thus far yet. Unique properties and multifunctionalities of mechanical metamaterials are promisingly expected to contribute to flourishing the energy harvesting research field beyond enhancing output performance. Achieving controlled trapping of elastic strain energy via programmed structures that has been reported as energy absorbing materials293 can be possibly utilized for energy harvesting. Triboelectric nanogenerators embedded in mechanical metamaterials demonstrated not only the tunable mechanical properties inherited by the mechanical metamaterials but also energy harvesting and self-powered deformation sensing capabilities.294 

3. Challenges and future prospects

Manmade structures, such as phononic crystals and metamaterials, have experienced a remarkable development of uncovering unconventional effective material properties and wave functionalities that can break the traditional limits in acoustic and vibration engineering over the last two decades. A multitude of innovative concepts and designs have been proposed along with theoretical basis behind the novel characteristics of metamaterials, followed by experimental validations sometimes in the later years. Now the field is moving to enter the next level where real-world applications of the recently developed phenomena come into the picture. Viewed in this context, energy harvesting is one of the great candidate applications that can benefit from the wave manipulation capability of metamaterials. Research advances of drastic power-enhanced energy harvesting using various types of metamaterials reviewed in this article are the evidence. As still in the early stage of research, there remain substantial challenges to overcome, from which we can identify tremendous research opportunities.

First, we need more innovative metamaterial design schemes that can further increase the degree of input wave energy amplification either by focusing or localization toward the target location for harvesting. In addition, such energy focusing or localization should take place in a wide range of frequencies for broadband operation. Phononic crystals derive their bandgap features from the multiple scattering, thus not based on resonances. Maximizing the bandgap of phononic crystals via geometry or material property tailoring can be a solution to broaden the operating bandwidth. In contrast, local resonance-based acoustic metamaterials or metasurfaces have an inevitably intrinsic issue of narrow working bandwidth while offering advantages of subwavelength or even ultrathin subwavelength scales. Ideas for possible solutions can be drawn upon the coherent perfect absorption approach295 and the integration of multiple metamaterials or metasurfaces over different frequencies,296 both of which were utilized to realize broadband perfect absorbers. In order to develop metamaterials-based energy harvesting adaptable to the environmental changes including frequencies, active metamaterials that have capability to tune the bandgaps as well as the resonance properties by changing effective properties like stiffness in real-time can also be considered as well.

Size reduction also matters. The wavelength of acoustic and elastic waves range from millimeters to centimeters. In this regard, subwavelength structures such as locally resonant metamaterials and metasurfaces have received considerable attention and deserve further study to realize focusing and harvesting functionalities. Mechanical metamaterials also offer compactness within a finite volume, making it possible to realize portable sensing and energy harvesting applications. Considering that the first experimental demonstration of pentamode mechanical metamaterials appeared only in 2012,288 it is no wonder that only very few article can be found on mechanical metamaterial for energy harvesting applications. Hence, there are vast opportunities in exploring intriguing mechanical metamaterial concepts that already exist and will be reported in the near future for energy harvesting purposes.

Developing machine learning-based design algorithms is necessary to identify metamaterials suitable for specific target applications and such artificial intelligence-based design methodology is not just applicable to energy harvesting but to all metamaterial-related applications.297–299 The present fabrication technology is limited to individual custom-made metamaterial specimens using laser-cutting or 3D printing. Together with the advances in 3D printing technology for mass production, automated fabrication processes for both metamaterials and energy harvesting devices are required in order to make this technology include viable commercialization.

Once input wave energy is amplified through metamaterials, efficient energy transfer from metamaterials to energy harvesting devices becomes important.300 In that regard, acoustic or mechanical impedance mismatch between metamaterials and piezoelectric energy harvesting devices can pose a challenge. Therefore, it is required to identify energy harvesting device designs and operating conditions that allow impedance matching, which can vary depending on the input source: sound in air or underwater, elastic waves in a plate, mechanical vibrations. Frequency alignment between energy harvesting devices and metamaterials is also an important factor to consider. When the resonance and the bandwidth of the energy harvesting devices coincide with those of metamaterials, the output harvesting performance can be maximized. Both impedance matching and resonance tuning can be realized by tailoring constituent material and geometric properties of metamaterials and energy harvesting simultaneously.

4. Concluding remarks

Metamaterials, starting from optics and having been extended to acoustics and mechanics, have fascinated scientific community in various fields due to their novel properties and functionalities that break the conventional barriers. The wave control capability of metamaterials has recently infiltrated into energy harvesting research and proved their potential for drastic enhancement in harvesting performance using vibration, sound, and elastic waves. Metamaterials-based energy harvesting undoubtedly offers a promising way to realizing sufficient power generation suitable for practical applications but still poses challenges, such as broadband operation, size reduction, impedance matching, and frequency tuning. Innovative metamaterial designs and deep physical understanding of the mechanisms at the interface between the metamaterial and the energy harvesting device will underpin the next generation of metamaterial-based energy harvesting.

5. Acknowledgments

This research was supported by the National Research Council of Science and Technology (NST) grant by the Korean Government (MSIP) (Grant No. CAP-17-04-KRISS) and the Creative Materials Discovery Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (Grant No. 2018M3D1A1058794).

Kai Wang, Yuchen Hou, Shashank Priya (*sup103@psu.edu)

1. Abstract

Halide perovskites (HPs) have emerged as a promising solution-processible material in the field of photovoltaics and optoelectronics. Recently, broader applications of these modified perovskite materials in mechanical energy harvesting have been proposed due to their high dielectric and piezoelectric properties that are analogous to those of conventional inorganic piezoelectric materials. At this juncture, fundamental origin of ferroelectric and piezoelectric properties of organohalide perovskites remains under investigation. Results reported in literature, such as high piezoelectric coefficient of ∼25 pm/V in MAPbI3 perovskite301 under illumination, ferroelectric domains,302 and optically enhanced piezoelectric properties of halide perovskite films,302,303 imply that these materials will have relevance in multimodal energy harvesting. In this perspective, we describe recent theoretical and experimental advances in understanding the dielectric and piezoelectric properties of halide perovskite materials followed by a brief outlook on the challenges and prospects for HP-based Multimodal Piezoelectric Generators (HP-MPG).

2. Motivation

Halide perovskites (HPs) have been well recognized in photovoltaics and hold the potential to be engineered into a piezoelectric material. This could induce the possibility to simultaneously harvest multiple type of energy sources using a single material. Prior reviews mostly focus on the single applied area of HPs. Besides the revisit to the frontier research on these materials, we also discuss the opportunities to simultaneously harvest different types of energies using an advanced device with HPs.

3. State-of-the-art in piezoelectric organohalides

Halide perovskites (HPs) are ABX3 “perovskite” structured crystals with A-site occupied by monovalent cations caged in the [BX6]4− corning-sharing octahedra network. Figure 11(a) shows the lattice structure of perovskites represented by formulation ABX3. The wide range of options available for each crystallographic site enables thousands of combinations for developing new materials within the generic constraint of the Goldschmidt tolerance rule. Under the applied mechanical or electric field, dynamic tilting of the octahedra and orientations of the A-site polar cations drives the dielectric properties of perovskites. Figure 11(b) shows the molecular ordering based on the alignment of the A-site dipolar molecule. Recently, Govinda et al. reported the tetragonal–orthorhombic phase transition in both MAPbI3 and MAPbBr3 perovskite during a cooling process and suggested that the dipole from MA+ units prominently contributes to dielectric value of the whole material.304 In addition, illumination induced charge excitation also modulates the dielectric constant from ∼103 (dark) to ∼106 under AM 1.5 illumination.301 Extrinsically doped/substituted HP materials, such as (Me2NH2)PbI3, (benzylammonium)2PbCl4, (3-pyrrolinium)CdCl3, Me3NCH2ClMnCl3, TMCM-CdCl3, show high dielectric properties and exhibit a characteristic Curie-temperature (Tc).305–307 Coll et al.301 first observed a giant increase of piezoelectric coefficient (∼25 pm/V) upon illumination of HP, compared to 5–6 pm/V in dark, corresponding to the formation of large photoinduced dipole moments through the molecular orientation of MA+ unit in the lattice. Dong et al.308 directly measured the effective piezoelectric coefficient (d33) of 2.7 pm/V in MAPbI3 single crystal. Song et al.309 found that casting of the HP film on epitaxial PZT could further enlarge its d33 to ∼4 pm/V. By engineering the composition derived from the prototype MAPbI3, higher figure-of-merits such as effective piezoelectric coefficient (d33) of ∼25 pm/V in FAPbBr3 nanoparticles, ∼17.0 ± 6.0 pm/V in MAPb1−xFexI3 films,310 49.51 pm/V in MA2CuCl4,311 and 185 pm/V in TMCM-MnCl3 (with a Tc of 406 K, higher than that of BTO306) have been achieved [Fig. 11(c)]. Figures 11(d-i) and 11(d-i) show the typical piezoelectric force microscopy (PFM) phase and amplitude images of the TMCM-MnCl3306 as an example, where the characteristic piezoelectric domains with lamellar-shape structure are clearly observed. The piezoelectric performance (d33) of the above-mentioned HP materials is also listed in the table in Fig. 11(e) for better comparison. It should be noted that compositional engineering and thus design of the new HP material could provide effective d33 over 103 pm/V. Such a compositional design strategy provides foundation for developing the roadmap for HP-based piezoelectrics. In addition to piezoelectric properties, HP materials are also attractive due to their simplified processing such as low-temperature, solution-spinning and cost-effective single-step synthesis route. The large potential design capacity of this material in combination with optical and mechanical response could provide tremendous potential for developing new generation of multi-modal energy harvesting technologies. These materials will be responsive to mechanical and optical stimulations simultaneously due to their piezoelectric and photovoltaic behavior. Furthermore, these materials can be tailored to respond to thermal excitation by tuning their thermal conductivity and Seebeck coefficient.

FIG. 11.

(a) Schematic crystal structure of ABX3 “perovskite.” Adapted with permission from Wang et al., Sol. Energy Mater. Sol. Cells 147, 255 (2016).312 Copyright 2015 Elsevier B.V. All rights reserved. (b) Molecular origin of polarization in a 2D perovskite: spinning of the A-site polar cation. Adapted with permission from Li et al., Nat. Commun. 8, 16086 (2017).313 Copyright 2017 Springer Nature. (c) List of d33 values in multiple organohalide perovskites. Data are obtained from various resources. (d) Domain structures of TMCM-MnCl3 seen in PFM images constructed by (i) phase and (ii) amplitude signal of the out-of-plane piezoresponse. Adapted with permission from You et al., Science 357, 306 (2017). Copyright 2017 AAAS. (e) Table summarizing the piezoelectric coefficient (d33) of the reported organo halide perovskite materials. Data are obtained from various resources.

FIG. 11.

(a) Schematic crystal structure of ABX3 “perovskite.” Adapted with permission from Wang et al., Sol. Energy Mater. Sol. Cells 147, 255 (2016).312 Copyright 2015 Elsevier B.V. All rights reserved. (b) Molecular origin of polarization in a 2D perovskite: spinning of the A-site polar cation. Adapted with permission from Li et al., Nat. Commun. 8, 16086 (2017).313 Copyright 2017 Springer Nature. (c) List of d33 values in multiple organohalide perovskites. Data are obtained from various resources. (d) Domain structures of TMCM-MnCl3 seen in PFM images constructed by (i) phase and (ii) amplitude signal of the out-of-plane piezoresponse. Adapted with permission from You et al., Science 357, 306 (2017). Copyright 2017 AAAS. (e) Table summarizing the piezoelectric coefficient (d33) of the reported organo halide perovskite materials. Data are obtained from various resources.

Close modal

HP based piezoelectric generators can scavenge mechanical energy from ambient environment and convert it into useful electrical energy. In comparison to traditional inorganic oxide perovskites, which require complicated synthesis processes such as high-temperature calcination and sintering, HP materials could be easily integrated onto target substrates using solution-processing based techniques. Furthermore, they can provide ability to respond to low acceleration levels (<0.1g). In 2015, Kim et al.314 measured the output performance of piezoelectric generator based on the MAPbI3 perovskite thin film casted by a spin-coating process. Figure 12(a) shows the device structure which consists of a MAPbI3 thin film inserted between two electrode layers of PET/indium–tin oxide (ITO) and Au/Ti/PET. After poling, the device exhibits a peak voltage of 2.7 V with a current density of 140 nA/cm2 under 0.5 MPa stress normal to the surface on an active area of 1 × 1 cm2. Liu et al.303 theoretically calculated the effect of the atomic substitution on the piezoelectric behavior. They found that the displacement of the lattice B-site cation contributes to most of the piezoelectric response of the material and the competition between A-X hydrogen bond and B-X metal-halide bond in HP material dominates its piezoelectric properties. Ippili et al.310 developed the Fe-doped HP with formulation of MAPb1−xFexI3 (x = 0.07) exhibiting polarization of ∼1.6 µC/cm2 and a corresponding piezoelectric device with output performance of ∼7.29 V as well as a current density of ∼0.88 µA/cm2 after poling at 30 kV/cm. This magnitude is much higher than that from the MAPbI3 device and could power a commercial red LED [Fig. 12(b)].

FIG. 12.

(a) A piezoelectric device using MAPbI3 and corresponding working mechanism. Adapted with permission from Kim et al., J. Mater. Chem. A 4, 756 (2016). Copyright 2016 The Royal Society of Chemistry. (b) Photograph of the instant LED light during applied mechanical pressure and its schematic rectifying circuit diagram for powering the LED. Adapted with permission from Ippili et al., Nano Energy 49, 247 (2018). Copyright 2018 Elsevier Ltd. All rights reserved. (c) Device configuration and working mechanism for harvesting thermal, mechanical, and solar energies. Adapted with permission from Jella et al., Nano Energy 52, 11 (2018). Copyright 2018 Elsevier Ltd. All rights reserved.

FIG. 12.

(a) A piezoelectric device using MAPbI3 and corresponding working mechanism. Adapted with permission from Kim et al., J. Mater. Chem. A 4, 756 (2016). Copyright 2016 The Royal Society of Chemistry. (b) Photograph of the instant LED light during applied mechanical pressure and its schematic rectifying circuit diagram for powering the LED. Adapted with permission from Ippili et al., Nano Energy 49, 247 (2018). Copyright 2018 Elsevier Ltd. All rights reserved. (c) Device configuration and working mechanism for harvesting thermal, mechanical, and solar energies. Adapted with permission from Jella et al., Nano Energy 52, 11 (2018). Copyright 2018 Elsevier Ltd. All rights reserved.

Close modal

In addition to the mechanical energy harvesting, the HP-based devices can be used for sensing light due to their excellent optical properties. Eom et al.315 reported self-powered pressure and light-sensitive bifunctional sensor using a CVD-MAPbI3 HP films. They found that CVD-MAPbI3 exhibits better stability than solution-prepared samples as solvents such as DMF and DMSO could induce the intermediate solvent–perovskite complexes that can interact with humidity. The device generates output voltage due to intrinsic piezoelectric coefficient of MAPbI3 films and exhibits sensitivities of 8.34 mV kPa−1 and 0.02 nA kPa−1. In terms of the light-stress bi-mode operation, the response time for applied pressure and light is 0.066 and 0.320 s, respectively, under a pressure of 30 kPa. Such a design could be the pathway for self-powered portable electronics and wireless sensors.

The multi-energy sensitivity of HP materials as well as their wide range of properties, such as excellent electronic band structure, ambipolar feature, ferroelectric nature, and ultra-high Seebeck coefficients, is opening new opportunities for multimodal energy conversion by harvesting multiple ambient energies, such as thermal, mechanical, and solar. Multimodal devices incorporating HP materials to simultaneously harvest thermoelectric, piezoelectric, and solar (TPS) energies are being proposed recently. Jella et al.316 developed the TPS-multimodal energy harvesting device consisting of interdigitated electrodes (IDEs) and multifunctional MAPbI3 to harvest thermal, mechanical, and solar energies. Figure 12(c) shows the architecture and operating mechanism of the TPS-device. The device exhibits a thermoelectric performance of 0.012 nW at ΔT of 15 °C, a piezoelectric performance of 1.16 V and 0.61 µA under 0.2 MPa, and a solar cell performance with a VOC of 0.8 V and JSC of 0.014 mA/cm2. This multimodal device design introduces an innovative framework for self-powered, portable, and mobile electronic systems. However, questions about the interaction between each type of energy harvesting mechanism remains unknown. For example, is there any coupling between harvesting energy from light, heat, and mechanical stress? How do they mutually affect each other and how to quantify these coupling effects? From molecular level, piezoelectric relies on the mechanical strain to change the lattice and induce the electric dipole alignment that eventually generates electricity. For semi-conductive materials, such as HP, not only the intrinsic leakage (semiconducting nature) but also the strain induced electronic band structure change will interfere the dipole dynamics during the piezoelectric process. Moreover, secondary or higher-level dynamics within the lattice (photon and thermal effect) will also be involved in these HP materials (due to their “soft lattice” nature) and eventually affect the piezoelectric process. These more complex questions and underlying physics will need expertise from cross-disciplinary teams and more investigations from different perspectives in a collaborative framework.

Another concept being investigated is the composite consisting of HP and polymeric matrix. Ding et al.317 reported a composite film, comprising FAPbBr3 HP nanocrystals uniformly distributed in a PDMS polymer matrix, and inserted between ITO/PET substrate and an aluminum foil to assemble a piezoelectric generator device. The poled device (50 kV/cm) exhibits an output voltage of 8.5 V and a current density of 3.8 μA/cm2, when tested under 0.5 MPa of vertical compression at a frequency of 6 Hz. The generated output was stored in a capacitor to drive a commercial red LED. Dhar et al.318 reported another polymer composite nanogenerator using MAPbI3 and PDMS and observed a significant lattice-defect-induced piezoelectric responses. The increase in the lattice B-site defects will enlarge the ionic polarization, which in turn enhances the permittivity and thereby piezoelectricity. The device fabricated with 5 wt. % PDMS composite generated voltage of over 100 V with a maximum power density of 0.3 mW/cm3, which could illuminate 30 commercial blue LEDs. PVDF has also been used as the matrix, which could provide better dispersion of the perovskite nanocrystals. Ding et al.319 demonstrated a FAPbBr3-PVDF composite for piezoelectric nanogenerators displaying highest outputs with a voltage of 30 V and current density of 6.2 µA/cm2. A uniform-distribution is key to transfer the stress and induce a more efficient piezoelectric behavior at device level. Sultana et al.320 synthesized MAPbBr3 nanoparticles and embedded them into the PVDF matrix in the form of nanofiber processed by electrospinning. These composite fibers exhibit acoustic vibration with a high degree of acoustic sensitivity of ∼13.8 V Pa−1 and efficiency of 58.5%. Due to the wide scope of variety of HP composition, other HP materials, such as MAPbI3321 and MA2CuCl4,311 have also been incorporated as filler in the HP-PVDF composite for piezoelectric generators.

4. Current and future challenges

Self-sustainable, self-power technologies are becoming important not only because of the growing adoption of portable and mobile electronic implementations and edge/node sensors but also due to the fast integration of cloud computing, remote-communication, and Internet of thing (IoT) technologies. Harvesting ambient energy, such as acoustic and mechanical vibration, through piezoelectric effect can be one of the promising solutions. In the earlier discussions, we have highlighted intriguing facets of HP materials in terms of piezoelectric generators. Compared to their inorganic counterparts, these HP materials are advantageous in terms of facile low-temperature processing and hybrid material nature that combines both organic and inorganic solids at a molecular level. From manufacturing level, HPs are good candidates for the roll-to-roll fabrication and excellent compatibility to target systems with specific flexibility requirement. So far, there has been limited research on the HP-based piezoelectric generators, but the interest is growing rapidly. To achieve sustainable progress, several essential aspects that need to be investigated are summarized below:

  • Stability. The intrinsic structural instability, such as phase instability and their extrinsic thermal/optical/ambient instability, need to be considered and addressed. Although multiple encapsulation techniques could boost their extrinsic stability against various environmental stimulus, improving their intrinsic instability still requires significant research efforts. Strategies, such as lattice strain engineering, and atomic-scale entropic modification might be potential directions.

  • Novel material exploration. The wide tunability of the perovskite ranging from crystallographic substitutions at atomic scale to crystal phase modulation and structural dimensionality, and to nanocomposites (0D, 1D, 2D, 3D, hierarchical nanoparticles, etc.) provides opportunity to design material architectures with suitable electromechanical parameters. Designing new compositions and microstructures and studying their piezoelectric and other relevant fundamental properties will provide pathway to improve the energy harvesting capabilities.

  • Clarification between multiple energetic stimulus. The HP materials are responsive to multiple energetic stimulus such as optical, thermal, mechanical, electrical, and even magnetic signals. Each of them may incorporate a coupling effect that could be mutually affected by each other from the material to device level. Decoupling them to gain more fundamental understanding, particularly at the nanoscale level, and/or manipulating their synergistic effects for energy harvesting might be excellent research direction.

  • Device-level research. Research and development at the device-level could provide practical direction to compare performance with current generation piezoelectric sensors and generators. This could improve the performance of HP-based piezoelectric device for mechanical energy harvesting, design of the photoferroelectric and photopiezoelectric devices, developing hybrid energy harvesters for multi-source energy harvesting, and developing eco-friendly, flexible, and high-performance devices.

5. Concluding remarks

In summary, HP materials are emerging as a new class of piezoelectric materials with advantages of simpler-processing and large compositional/structural tunability. The research on the HP-based piezoelectric materials is in early stages with much work required to improve the material properties and device implementation. With the growing research on inventing novel HP materials exhibiting variety of multifunctional properties, including piezoelectric, thermoelectric, and photovoltaic, there is significant opportunity to realize new generation of multimodal harvesters. We envision that future investigations in the field of HP materials will be more cross-disciplinary as it requires knowledge from multiple domains to resolve the intrinsically coupled effects.

6. Acknowledgments

K.W. acknowledges the financial support through SBIR Program through Nanosonic. K.W. also acknowledges IEE Stewardship Seed Grant Program (Penn State). S.P. would like to acknowledge the financial support from Office of Naval Research through Award No. N000141912461. Y.H. acknowledges the support through the National Science Foundation (Award No. 1936432).

Wei Xu, Jianhua Hao (*jh.hao@polyu.edu.hk)

1. Abstract

As important mechanical energy transfer devices, triboelectric nanogenerators (TENGs) and piezoelectric nanogenerators (PENGs) have achieved fast development and shown various applications. Polymers are playing critical roles in TENGs and PENGs, and especially, smart polymer materials (SPMs) possessing ability of responding to the external stimuli are attracting increasing attention for their combinations with TENGs and PENGs. Considering the outstanding advantages of these nanogenerators combining SPMs and the lack of relevant overview, here we briefly summarize the progress about the integration of SPMs with TENGs and PENGs. Meanwhile, some challenges and opportunities based on the status of this research field are presented and discussed.

2. State-of-the-art

Triboelectric nanogenerators (TENGs) and piezoelectric nanogenerators (PENGs) can generate electric output by converting ambient mechanical motions with broad frequency and scale range. This ability enables them to be the potential complement of traditional power sources and become an ideal candidate for forming self-sufficient system and powering emerging electronics.20,28,322 Among them, TENGs, working based on triboelectric and electrostatic effects, possess various unique advantages and their fabrication requires using amount of polymers as contact materials, substrates, or spacers. The properties of the employed polymers will obviously affect the performance of TENGs.323–327 

Compared to traditional polymers, smart polymer materials (SPMs), a special type of polymers that possess the inherent ability of making response to the external stimuli,328 have been attracting considerable attention in the field of nanogenerators. The external stimuli to trigger SPMs could be temperature, light, and solvent. Under these stimuli, SPMs will generate the corresponding responses and show certain changes in one or more aspects, such as shape, color, and conductivity.328,329 The common stimuli and corresponding responses for SPMs are shown in Fig. 13.

FIG. 13.

The stimuli and response for smart polymer materials.

FIG. 13.

The stimuli and response for smart polymer materials.

Close modal

Given the unique stimuli-response ability, the integration of SPMs into TENGs is considered as a feasible approach to conceive devices with attractive functions. Based on the current achievements, the purpose of combining SPMs with TENGs can be broadly classified into two aspects:

a. To improve the certain performance of devices.

In this case, the most typical example is to enhance output signals of TENGs by using piezoelectric and ferroelectric polymers, such as polyvinylidene fluoride (PVDF) and poly(vinylidenefluoride-co-trifluoroethylene) [P(VDF-TrFE)], as the triboelectric materials of TENGs. By means of the electrically induced piezoelectric/ferroelectric polarization, the surface charge density of triboelectric materials is significantly increased, leading to an enhancement in electric output. This improvement can be attributed to the shift in Fermi level of poled polymers instead of the piezoelectric effect.330,331

Another representative example is the use of healable polymers in TENGs for improving their lifespan and durability.332–334 Healable polymer is one type of SPMs that enables fractures healing under certain stimuli. Specifically, the healing can be realized by means of the existence of some special bonds, including dynamic covalent bonds, reversible covalent bonds, and non-covalent bonds, in the molecular structures of healable polymers. These bonds enable the molecular chain diffuses on fracture interface and the re-building of broken bonds to achieve the recovery in structure and characteristics.335,336 The healing process usually requires some external stimuli (e.g., temperature), while part healable polymers can fulfill the healing process under ambient conditions, which is categorized as autonomously healable or self-healing polymers.336,337 By employing healable polymers as constituent materials, the achieved healable TENGs are endowed with healability to recover the structure and performance when fracture has occurred, which could improve the durability, lifespan, and reliability of devices. On the other hand, the mechanical energy conversion of TENGs requires frequent mechanical impacts, which easily leads to the failure in both material and device. Therefore, the combination of healable polymers with TENGs is practically significant. In addition, attributed to healability, healable TENGs also present shape-tailorability and shape-adaptability for broadened application range.338–340 

Numerous healable polymers have been prepared for forming healable TENGs, as summarized in Table II. Among them, self-healing TENGs using autonomously healable polymers can conduct the recovery without stimuli but usually show weak inherent mechanical strength. The reason is that autonomously healable polymers usually require the low glass transition temperature (Tg) for effective molecular chains diffusion under ambient conditions. In contrast, the stimuli triggered healable polymers perform higher mechanical properties. For instance, a thermal triggered healable polyurethane with tensile strength of about 10.5 MPa has been recently reported by our group for TENGs.341 It should be mentioned that the recovery in electric output of TENGs mainly depends on the healing of electrodes conductivity.342 The traditional electrodes used in healable TENGs usually consist of conductive fillers, and their conductivity recovery depends on the simple recontact of broken electrodes that need to be fixed and remained by the healed polymer substrates. Some TENGs, therefore, include intrinsically healable electrodes, such as hydrogel or magnetic electrodes, to arrive the fast healing of performance, where healable polymer substrates will mainly be charge of the mechanical property recovery of devices.338,343

TABLE II.

Comparison of healable polymers employed in TENGs.

PolymerDynamic bonds or structureStimuliHealing efficiency (%)Tensile strength
H-PDMS344  Imine bonds Room temp (12 h) 94 120 kPa 
HTS-PDMS345  Bipyridine groups-zinc ions metal–ligand Room temp (30 min) ∼100 ∼70 kPa 
CNT-putty composite346  Boron–oxygen bonds and hydrogen bonds Room temp (40 min) 92 ∼12 kPa 
IU-PDMS347  Imine bonds and quadruple hydrogen bond Room temp (24 h) ∼100 1.09 MPa 
Healable PPUH polymer348  Fe(III)–PDCA interactions and Room temp (24 h) 98.2 ∼1.4 MPa 
 quadrupolar hydrogen bonding    
Healable poly(dimethylsiloxane) H bonds Room temp (72 h) 98 1.27 MPa 
elastomer349      
MWCNTs-UPy/IU-PAM347  Imine bonds and quadruple hydrogen bonds Room temp (36 h) 97 2.13 MPa 
Near-infrared (NIR) 88 
PVA/agarose hydrogel350  Boron–oxygen bonds and crystallites Water/NIR 93 83 kPa 
PVA/PDAP/graphene hydrogel351  Hydrogen bonds Room temp (60 s) 98 85 kPa 
NIR (30 s) 94 
Laponite/poly(AMPS-co-AA-co-DMAPMA)/ Hydrogen bonds Heat 99 106 kPa 
GO–LiCl hydrogel352      
Epoxy based polysulfide353  Disulfide bonds NIR 90 2.6 MPa 
Healable PU-PDMS338  Disulfide bonds and hydrogen bonds Room temp (1 h) 56 0.87 MPa 
Heat 97 
Vitrimer elastomer354  Disulfide bonds Heat ∼100 1.3 MPa 
Healable polyurethane acrylate339  Multivalent H-bonds Heat 45.1 ∼7 MPa 
Healable PU341  Disulfide bonds Heat 96.7 10.5 MPa 
Poly(hindered urea)355  Bulky urea linkages Heat >95 1.7 MPa 
PolymerDynamic bonds or structureStimuliHealing efficiency (%)Tensile strength
H-PDMS344  Imine bonds Room temp (12 h) 94 120 kPa 
HTS-PDMS345  Bipyridine groups-zinc ions metal–ligand Room temp (30 min) ∼100 ∼70 kPa 
CNT-putty composite346  Boron–oxygen bonds and hydrogen bonds Room temp (40 min) 92 ∼12 kPa 
IU-PDMS347  Imine bonds and quadruple hydrogen bond Room temp (24 h) ∼100 1.09 MPa 
Healable PPUH polymer348  Fe(III)–PDCA interactions and Room temp (24 h) 98.2 ∼1.4 MPa 
 quadrupolar hydrogen bonding    
Healable poly(dimethylsiloxane) H bonds Room temp (72 h) 98 1.27 MPa 
elastomer349      
MWCNTs-UPy/IU-PAM347  Imine bonds and quadruple hydrogen bonds Room temp (36 h) 97 2.13 MPa 
Near-infrared (NIR) 88 
PVA/agarose hydrogel350  Boron–oxygen bonds and crystallites Water/NIR 93 83 kPa 
PVA/PDAP/graphene hydrogel351  Hydrogen bonds Room temp (60 s) 98 85 kPa 
NIR (30 s) 94 
Laponite/poly(AMPS-co-AA-co-DMAPMA)/ Hydrogen bonds Heat 99 106 kPa 
GO–LiCl hydrogel352      
Epoxy based polysulfide353  Disulfide bonds NIR 90 2.6 MPa 
Healable PU-PDMS338  Disulfide bonds and hydrogen bonds Room temp (1 h) 56 0.87 MPa 
Heat 97 
Vitrimer elastomer354  Disulfide bonds Heat ∼100 1.3 MPa 
Healable polyurethane acrylate339  Multivalent H-bonds Heat 45.1 ∼7 MPa 
Healable PU341  Disulfide bonds Heat 96.7 10.5 MPa 
Poly(hindered urea)355  Bulky urea linkages Heat >95 1.7 MPa 

To enhance the output performance of TENGs, microstructures usually are fabricated on the surface of the triboelectric layers. These microstructures can increase the effective contact area, but they may suffer from collapse and damage under strenuous mechanical impacts, causing the failure in performance and stability of TENGs. This obstacle can be overcome by employing shape-memory polymers as constituent materials of microstructures. Shape-memory polymer is a class of SPMs capable of retaining a temporary shape and recovering to its original permanent shape via transforming from any temporary shape in the presence of an external stimulus.356 That means the deformed shape-memory polymer-based microstructures can recover to the original morphology under stimuli, leading to the recovery of device normal state and the improvement in stability and durability of TENGs.356 Similarly, the recovery and renewal of distorted microstructures can also be realized by using the light-transformable polymer as constituent materials.340 Another role of shape-memory polymers in TENGs is to serve as the main material instead of microstructures materials to improve shape-adaptability of device.357 

Another strategy proposed by our group to improve the device’s durability and robustness is the fabrication of magnetic-assisted non-contact TENGs by coating a magnetic responsive polymer composite layer on the surface of device. The interaction between triboelectric materials, therefore, can be remotely controlled by the motion of external magnet via magnetic field mediation. This will avoid the direct mechanical impact to TENGs and, therefore, delay the device failure, leading to improvements in the device’s durability.358,359

b. To develop new application of TENGs.

It is known that the amplitude of electric outputs of TENGs depends on the coupling effects of a few factors, including the contact area between triboelectric materials, the friction surfaces properties, and the electrodes conductivity. For the TENGs combining SPMs as constituent parts, the SPMs employed in the energy device may generate certain responses to the external stimuli. Once these responses of SPMs can affect the above factors, the electric output of TENGs will be correspondingly changed. This enables TENGs to serve as sensors that are capable of detecting these external stimuli by observing output signal change. Therefore, a series of stimuli sensors based on TENGs have been developed,360–363 such as the ammonia sensor based on ammonia-responsive resistance change of polyaniline electrode,363 and the temperature sensor based on heat triggered morphology recover of shape-memory polymer-based friction surface.360 

On the other hand, the electricity generated by TENGs can perform as a stimulus to trigger the electrically responsive polymers. The common electrically responsive polymers mainly include piezoelectric polymers, dielectric elastomers, electroluminescence, and electrochromic polymers, and they can produce the corresponding responses, including the deformation, luminescence, and color change, to the electric stimuli for further application. By combining TENGs with electrically responsive polymers, a series of TENG driven systems have been developed as actuator, artificial muscle, and memorization.364–367 Since the widely distribution of mechanical energy is available, these TENG driven systems will show self-powered ability of harvesting energy from operation environment as the power supply themselves. This brings them potential advantages for application in portable and implanting electronics, and some remote occasions.

Similarly, for PENGs working based on the piezoelectric effects, they can also drive electrically responsive polymers to arrive the self-powered smart systems.368 Meanwhile, SMPs also contributes to the performance improvement of PENGs. The most important and representative SPMs used for PENGs are supposed to be the piezoelectric polymers, such as PVDF and its copolymers. Just as its name implies, piezoelectric polymers can perform the piezoelectric behavior of generating electricity when the mechanical force is applied on it (direct piezoelectric effect) and vice versa (converse piezoelectric effect), which directly leads to the birth of polymer-based PENGs.369,370 Compared with the piezoelectric ceramics possessing higher piezoelectric coefficient in regular PENGs, piezoelectric polymers usually show better flexibility, facile processing, and lower density. This endows the polymer-based PENGs with good flexibility and robustness for suitable application in some occasions demanding large deformation and strong mechanical impact.369,370 For acquiring piezoelectricity required by PENGs, piezoelectric materials usually need accept the “poling” treatment for oriented alignment of dipoles by being applied strong electric field at phase transition temperature. However, for some piezoelectric polymer nanowires prepared by electrospinning and template-wetting technique, the orientation of dipoles can be formed during nanowire’s growth without further treatment.20,371 The corresponding nanowire-based generators present an improved energy conversion efficiency. Besides, the ferroelectric polymer with the spontaneous polarization is also a good candidate for achieving polymer-based PENGs without poling procedure.372 

Similar to TENGs, PENGs’ durability is also challenged by mechanical impacts. The combination of healability into PENGs, therefore, is significant from the practical application level. Recently, a lactate-based piezoelectric polymer elastomer was designed by introducing flexible chain segments into the molecular structure.373 Attributed to the decreased crystallinity, lower glass transition temperature and long linear molecular chain nature, this piezoelectric polymer shows electrical and mechanical healing ability at body temperature, leading to the formation of a healable polymer-based PENG. Besides piezoelectric polymers, PENGs can also be prepared based on piezocomposite, which is a mixture of the polymer matric and the inorganic piezoelectric material. For piezocomposite, the use of SPMs as matric will endow it with additional functions for performance improvement of PENGs. For instance, Yang et al. fabricated a self-healing piezocomposite through mixing piezoelectric PZT particles with self-healing PDMS for achieving a fully self-healing PENG.374 The current achievements about the integration of SPMs with TENGs and PENGs are summarized in Fig. 14.

FIG. 14.

Schematic diagram showing the motivation and achievement of integrating SPMs with TENGs and PENGs.

FIG. 14.

Schematic diagram showing the motivation and achievement of integrating SPMs with TENGs and PENGs.

Close modal

3. Challenges and future prospects

Despite many SPMs have been successfully coupled with TENGs and PENGs, some challenges and opportunities still exist for extending these concepts into practical operation. One of motivations of using SPMs in nanogenerators aims at improving device’s performance. However, for current prototypes, the improvement in certain performance of a device usually accompanies with the sacrifice of some other performances. For instance, healable polymers may improve the device’s lifetime and durability by means of the recovery of structure integrity. However, since the usually weaker mechanical strength of healable polymers (especially self-healing polymers), healable devices themselves become easier to be damaged and require more frequent recovery than regular TENGs. For those healable polymers with high strength, their healing process requires external stimuli and severe healing condition. Considering that TENGs and PENGs usually operate under mechanical impacts, to develop and employ