• No results found

Bioinspired Designs and Biomimetic Applications of Triboelectric Nanogenerators

N/A
N/A
Protected

Academic year: 2021

Share "Bioinspired Designs and Biomimetic Applications of Triboelectric Nanogenerators"

Copied!
23
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

University of Groningen

Bioinspired Designs and Biomimetic Applications of Triboelectric Nanogenerators

Li, Wenjian; Pei, Yutao T.; Zhang, Chi; Kottapalli, Ajay Giri Prakash

Published in:

Nano energy

DOI:

10.1016/j.nanoen.2021.105865

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from

it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date:

2021

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Li, W., Pei, Y. T., Zhang, C., & Kottapalli, A. G. P. (2021). Bioinspired Designs and Biomimetic Applications

of Triboelectric Nanogenerators. Nano energy, 84, [105865]. https://doi.org/10.1016/j.nanoen.2021.105865

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

Nano Energy 84 (2021) 105865

Available online 11 February 2021

2211-2855/© 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Review

Bioinspired designs and biomimetic applications of

triboelectric nanogenerators

Wenjian Li

a

, Yutao Pei

a

, Chi Zhang

b

, Ajay Giri Prakash Kottapalli

a,c,*

aDepartment of Advanced Production Engineering, Engineering and Technology Institute Groningen, Faculty of Science and Engineering, University of Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands

bBeijing Institute of Nanoenergy and Nanosystems, Chinese Academy of Sciences, No. 8, Yangyandong 1st Road, Yanqi Economic Development Zone, Huairou District, Beijing 101400, China

cMIT Sea Grant College Program, Massachusetts Institute of Technology (MIT), 77 Massachusetts Avenue, NW98-151, Cambridge, MA 02139, United States

A R T I C L E I N F O Keywords: Triboelectric nanogenerator Bioinspired Biomimetic Electronic skin Neuromorphic device A B S T R A C T

The emerging novel power generation technology of triboelectric nanogenerators (TENGs) is attracting increasing attention due to its unlimited prospects in energy harvesting and self-powered sensing applications. The most important factors that determine TENGs’ electrical and mechanical performance include the device structure, surface morphology and the type of triboelectric material employed, all of which have been investi-gated in the past to optimize and enhance the performance of TENG devices. Amongst them, bioinspired designs, which mimic structures, surface morphologies, material properties and sensing/power generation mechanisms from nature, have largely benefited in terms of enhanced performance of TENGs. In addition, a variety of bio-mimetic applications based on TENGs have been explored due to the simple structure, self-powered property and tunable output of TENGs. In this review article, we present a comprehensive review of various researches within the specific focus of bioinspired TENGs and TENG enabled biomimetic applications. The review begins with a summary of the various bioinspired TENGs developed in the past with a comparative analysis of the various device structures, surface morphologies and materials inspired from nature and the resultant improvement in the TENG performance. Various ubiquitous sensing principles and power generation mechanisms in use in nature and their analogous artificial TENG designs are corroborated. TENG-enabled biomimetic applications in artificial electronic skins and neuromorphic devices are discussed. The paper concludes by providing a perspective to-wards promising directions for future research in this burgeoning field of study.

1. Introduction

Triboelectric nanogenerators (TENGs), since their invention by Wang’s group in 2012 [1–4], have become an emerging research hotpot and gained widespread scientific attention, for their prospects in energy harvesting and self-powered sensing [5,6]. TENGs work on the principle of conversion of mechanical energy into electricity based on the coupling effect of contact electrification and electrostatic induction [7]. The key merits of TENGs such as simple structure, wide material choice, excellent output performance, ease of fabrication and possibility to form flexible and stretchable devices enable them to effectively harvest ambient environmental energy from diverse energy forms, such as vi-bration [8,9], wind [10,11], water wave [12–14] and human motion

energy [15,16]. In comparison to other types of energy harvesters which use electromagnetic, piezoelectric and photovoltaic mechanisms, TENGs have unique advantages in terms of their ability to efficiently scavenge random and low frequency ambient environmental energy [17–22]. TENGs can not only be used as energy harvesters but also as self-powered sensors, due to the fact that the amplitude, frequency and/or waveform of their electric output can reveal the changes in external mechanical stimuli and environmental parameters. For instance, self-powered sliding sensors [23], tactile sensors [24] based on the change in amplitude and frequency of the triboelectric signal have been demonstrated to function with high accuracy. It was also by demonstrated that analyzing the waveform of the output of flexible wearable TENGs, respiration, heart/pulse rate and biometrics can be

* Corresponding author at: Department of Advanced Production Engineering, Engineering and Technology Institute Groningen, Faculty of Science and Engineering, University of Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands.

E-mail address: a.g.p.kottapalli@rug.nl (A.G.P. Kottapalli).

Contents lists available at ScienceDirect

Nano Energy

journal homepage: http://www.elsevier.com/locate/nanoen

https://doi.org/10.1016/j.nanoen.2021.105865

(3)

detected with high sensitivity [25–27].

Over the last few years, enormous research efforts have been made towards developing and optimizing TENGs. The device structure, sur-face morphology, triboelectric material and dielectric property are considered as the most important factors in determining TENGs’ per-formance [28–31]. In view of various device structural designs, TENGs are classified into four basic working modes [30]: contact-separation mode [3,32,33], lateral-sliding mode [34,35], single-electrode mode [36,37], and freestanding mode [38,39]. Based on the four basic modes, TENGs with more complex and smart structures are designed to meet different types of energy harvesting and self-powered sensing re-quirements in different scenarios. The surface morphology of the triboelectric materials can greatly affect the effective contact area [3, 40], which plays a crucial role in determining the total surface charges during the contact electrification process, thus affecting the output performance of TENGs. Materials selection for the fabrication of TENGs, especially the triboelectric materials [41,42], also takes an equally important position when it comes to TENGs’ output performance. The greater the difference in triboelectric polarity between the two tribo-electric materials (one material is more likely to lose electrons and another material is more likely to gain electrons), the better the output of the TENGs. The excellent dielectric property of the triboelectric ma-terials can not only improve the surface charge density, but also effec-tively avoid electrostatic breakdown effect [43–45]. In addition, the specific physical and chemical properties of materials employed can significantly benefit TENGs, offering them desirable features such as flexibility [46,47], stretchability [48,49] and biocompatibility [50,51]. As a result, optimization of TENGs’ structure, surface morphology, triboelectric material choice, and dielectric constant is key to improve TENGs’ mechanical and electrical performance of the device.

Nature can provide us remarkable inspirations for scientific research and technology innovation [52–54]. On one hand, bioinspired designs,

which mimic structures [55–57], materials [58,59] and surfaces [60–62] from nature, have become a very important and effective approach to optimize traditional devices and develop novel products such as adhesive interface to overcome gravity for walking on walls inspired by geckos [63,64] and microlenses inspired by brittle stars [65, 66]. On the other hand, traditional devices and materials can be employed for biomimetic applications, especially those realize specific functions like sense organs. For example, multifunctional electronic skins [67–70] functionalized as tactile [71,72], temperature [73,74] and strain sensor [71,75] are widely used in robotics, while artificial neurons [76–78] greatly benefit the development of artificial intelligence and deep learning.

Among the various design and optimization efforts to enhance the electrical and mechanical performance of TENGs, bioinspired designs which mimic structures, surface morphologies, material properties and sensing/power generation principles from nature have proven to offer impressive performance enhancement for TENGs. Bioinspired designs of TENGs also allow them to have better applicability in various applica-tion scenarios which although quintessential are often easily over-looked. In addition, the merits of TENGs like externally-interactive, highly sensitive, easy fabrication and integration, tunable output and self-powered property enable TENG to be developed into various ap-plications, especially self-powered sensing. By mimicking human’s biological functions, biomimetic applications of TENGs such as artificial electronic skin and artificial neuromorphic devices largely promoted the development of the fields and brought more possibilities to realize cost- effective artificial components.

In this review, we focus specifically on bioinspired TENGs and TENG enabled biomimetic applications, as illustrated in Fig. 1. Bioinspired TENGs refer to TENGs with designs inspired from biology, while TENG enabled biomimetic applications involve devices where traditional TENGs are applied to realize the end engineering functionality which

Fig. 1. Bioinspired designs and biomimetic applications of triboelectric nanogenerators. “Fingerprint”. Reproduced with permission [98]. Copyright 2018, Elsevier. “Honey-comb”. Reproduced with permission [90]. Copyright 2019, Wiley-VCH. “DNA”. Reproduced with permission [82]. Copyright 2020, Wiley-VCH. “Lotus leaf”. Reproduced with permission [105]. Copyright 2019, Elsevier. “Moth’s eyes”. Reproduced with permission [106]. Copyright 2019, Elsevier. “C. zebrine leaf”. Reproduced with permission

[108]. Copyright 2019, Wiley-VCH. “Cancellous-bone”. Reproduced with permission [118]. Copyright 2019, Elsevier. “Water hyacinth petiole” [116]. Reproduced with permission. Copyright 2017, Elsevier Ltd. “Silk”. Repro-duced with permission [119]. Copyright 2016, Wiley-VCH. “Jellyfish”. Reproduced with permission [128]. Copyright 2017, Elsevier. “Electric eel”. Reproduced with permission

[127]. Copyright 2016, Wiley-VCH. “Cockroach”. Repro-duced with permission [122]. Copyright 2019, Wiley-VCH. “Multifunctional”. Reproduced with permission [132]. Copyright 2018, Wiley-VCH. “Triboelectric-photonic”. Reproduced with permission [133]. Copyright 2018, Wiley-VCH. “Ultrastretchable”. Reproduced with permis-sion [131]. Copyright 2017, American Association for the Advancement of Science. “Basilar membrane”. Reproduced with permission [136]. Copyright 2016, Wiley-VCH. “Artificial synapse”. Reproduced with permission [138]. Copyright 2019, Elsevier. “Neuromorphic tactile sensor”. Reproduced with permission [137]. Copyright 2020, American Chemical Society.

(4)

mimics a specific biological function. Firstly, TENGs designed and optimized by learning from natural structures, surfaces/interfaces, ma-terial properties, and unique sensing/energy harvesting mechanisms are in-depth introduced, respectively. Secondly, applications of TENGs in

mimicking biological functions for novel biomimetic devices are dis-cussed in detail. TENG enabled biomimetic applications are mainly of two kinds: artificial electronic skin and artificial neuromorphic devices. Finally, we conclude this review and discuss future research

Fig. 2. TENGs with bioinspired structures from plants. a, b) Schematic illustration of the bioinspired TENG tree, leaf-TENG and stem-TENG. c) Schematic diagram of the TENG tree working in the tunnel. Reproduced with permission [79]. Copyright 2018, Wiley-VCH. d) Schematic illustration of the lawn structured TENG arrays with a kelp forest morphology. e) The structure of the lawn structured TENG unit. f) demonstration of lighting up 60 LEDs. Reproduced with permission [80]. Copyright 2015, Wiley-VCH. g) Optical photo of the kelp undersea. h) Schematic illustration of the kelp-inspired biomimetic TENG (K-TENG). i) Output of the K-TENG under different frequencies. Reproduced with permission [81]. Copyright 2019, Elsevier. j) Structure of bioinspired helical TENG (BH-TENG). k) BH-TENG pasted on the soles of shoes to harvest human walking energy. Reproduced with permission [82]. Copyright 2020, Wiley-VCH.

(5)

perspectives for bioinspired TENGs’ design and TENG-based biomimetic applications.

2. Bioinspired designs of TENG

2.1. Bioinspired structural designs for TENG

In a pursuit to broaden the applications of TENG devices in both energy harvesting and self-powered sensing, various novel and opti-mized structural designs were investigated in the past. Optimization of the structural design of TENG devices on one hand enables adaptable application-specific performance suited to diverse application scenarios, ranging from water wave energy harvesting to implantable biomedical devices and on the other hand enhances the performance of the TENG in terms of electrical output and mechanical stability.

Biological structures ubiquitously seen in nature often feature unique morphological designs and undertake specific roles that support the perennial existence of life. Learning from these structures can provide TENGs with excellent characteristics and make TENGs suitable for spe-cific scenarios.

2.1.1. Structural designs bioinspired from plants

Through millions of years of evolution, plants have evolved to be extremely adaptive to the environment. Some plants can adjust their position and posture to constantly adapt to the environmental changes while some others exhibit extreme stability regardless of the environ-mental changes, becoming important models of inspiration for designing and developing new technologies. Since TENGs are widely used to harvest wind and ocean wave energy, the structure of some plants especially trees, lawns and aquatic plants, which can vibrate or swing with environmental winds or oceanic waves, naturally inspired re-searchers to design novel TENG structures. Bian et al. designed a bionic TENG tree which consists of supercells with leaf-TENG and stem-TENG to harvest wind energy, especially in the subway tunnel environment (Fig. 2a–c) [79]. The leaf-TENG was designed taking inspiration from the bionic structure of the elliptical shape-like natural leaf, while the stem-TENG takes inspiration from the bionic structure of the stem. In the presence of external wind flow, both the leaf-TENG and stem-TENG can generate triboelectricity based on wind-induced contact-separation. A TENG tree with four supercells connected parallelly generated an output voltage and current of about 330 V and 59.6 µA, respectively, at a wind speed of 11 m/s. To harvest wind energy from arbitrary flow directions, Zhang et al. designed lawn structured TENG arrays which feature a kelp forest morphology (Fig. 2d, e) [80]. Laminar free-standing strips (TENG units) can sway individually and thus contact and separate with each other. Under a wind flow velocity of 27 m/s, the vibration frequency of the strips can be as high as 154 Hz, and two adjacent TENG units can generate a power density of 2.76 W m−2. With a 60-strip array installed

on the rooftop, the generated powered density can be 2.37 W m−2 which

continuously lights up 60 LEDs (Fig. 2f). Inspired by the motion of kelp that gently sways along with undersea waves, Wang et al. designed a kelp-inspired biomimetic TENG (K-TENG) to harvest oceanic wave en-ergy (Fig. 2g, h) [81]. The K-TENG was comprised of two columns of vertically free-standing polymer strips with Cu electrodes deposited on their back sides. Every two adjacent strips served as a TENG unit and as the K-TENG vibrates in the presence of waves, each strip can sway independently and contact and separate with the neighboring strip. The K-TENG can efficiently harvest ocean wave energy at vibration fre-quency as low as 1 Hz with an output current and voltage of 10 µA and 260 V in a single unit (Fig. 2i).

Not only the external macro structure of plants can benefit the design of TENGs, but the hidden internal microstructure also gives inspiration to TENGs’ structural design. Li et al. designed a bioinspired helical TENG (BH-TENG) from the helical structure of DNA (Fig. 2j, k) [82]. The unique double helical structure of DNA not only guarantees the secur-ity/stability of genetic information, but also allows more genetic

information to be packed in a smaller space. The two triboelectric layers used in the BH-TENG were considered as the two base pairs in DNA. Therefore, the BH-TENG had a larger effective contact area. Compared to the normal contact-separation TENG with planar electrodes with the same size, the helical structure not only induces more effective contact areas but also endows increased stability.

2.1.2. Structural designs bioinspired from animals

In contrast to plants, almost all species of animals have to actively interact with their living environments to survive based on their specific mobility, powerful sensing ability, hunting skill and habitation-building ability. The unique morphological structural organization of some tis-sues, organs and sensors in animals have proved to be quintessential to their perception and interaction ability, and researchers in the past have translated such structural designs to the development of bionic devices, especially bionic sensors. By learning from animals, the structure of TENGs can be greatly optimized to boost the energy harvesting effi-ciency and especially the self-powered sensing ability. Lei et al. designed a butterfly-inspired TENG (B-TENG) to effectively harvest energy from different types of low-frequency waves. The B-TENG consisted of an arc- shaped outer shell and a spring-assisted four-bar linkage-based TENG (Fig. 3a) [83]. Copper triboelectric wings were mounted on the linkage, which can easily deform and recover because of the restoring force of the spring. Due to the special structure of the outer shell and the four-bar linkage, the B-TENG can work under both reciprocating swing mode and flapping mode. The B-TENG is able to effectively harvest water wave energy irrespective of whether it was placed parallel or perpendicular to the direction of the water wave. Learning from the layout of spider net, Shi et al. proposed a bioinspired triboelectric interface based on a single electrode TENG in spider-net layout (Fig. 3b, c) [84]. Two information coding configurations were realized, large/small electrode coding (L/S coding) and with and/or without electrode coding (0/1 coding). L/S coding interpreted coding information in terms of relative amplitude, since large output peak was generated when sliding across large-width electrode. 0/1 coding was designed with and/ or without electrode in a predefined position, interpreting coding information in terms of their positions in time domain, avoiding signal overlapping and large device dimension.

Yoon et al. fabricated a 3D-printed biomimetic-villus structure TENG (BV-TENG), which realized a large enhancement in surface area within a limited structural volume [85]. Compared to the planar structure, the surface area of the bioinspired villus structure had an improvement of 300%. For small-scale wind energy harvesting, Ahmed and co-workers designed a shape-adaptive, lightweight TENG inspired from the spe-cial flutter mechanics of the hummingbird (H-TENG) [86]. Flutter-driven TENGs were designed with lightweight wings to harvest wind energy from different directions and different attack angles.

Xiao et al. proposed a non-spring assisted honeycomb structure inspired TENG (HSI-TENG), which can overcome the two major limi-tations of spring-assisted vibrational TENGs: narrow working bandwidth and single-direction energy harvesting [8,32,87–89]. The HSI-TENG features a 3D printed honeycomb frame with a PTFE ball inside each of the grooves sandwiched with two copper electrode layers (Fig. 3d) [90]. As the HSI-TENG was subjected to an external vibration, the PTFE balls acting as both oscillator and electronegative triboelectric layer started to move up and down. The HSI-TENG was capable of harvesting vibration energy as long as the vibration acceleration exceeds a certain value, despite of the increase in frequency. The honeycomb inspired structural design not only allowed the HSI-TENG to harvest vibrational energy beyond the limitation of resonant frequency (Fig. 3e), but also increased the effective contact area. Compared to the square-grid frame of the same dimensions, the compact nature of the honeycomb structure had 6 more grooves, thus increasing the contact area by 12.2% and power density by 43.2%.

Su et al. developed a wearable alveolus-inspired membrane sensor (AIMS) for nitrogen dioxide (NO2) detection and breath analysis

(6)

Fig. 3. TENGs with bioinspired structures from animals. a) Schematic illustration of the butterfly-inspired TENG (B-TENG). Reproduced with permission [83]. Copyright 2018, Wiley-VCH. b) Schematic diagram of the bioinspired spider-net-coding (BISNC) triboelectric interface. c) The BISNC interface with L/S coding. Reproduced with permission [84]. Copyright 2019, Wiley-VCH. d) Schematic diagram of the honeycomb structure inspired TENG (HSI-TENG). e) Output peak voltage of the HSI-TENG as a relationship with vibration frequency. Reproduced with permission [90]. Copyright 2019, Wiley-VCH.

(7)

(Fig. 4a–c) [91]. The bionic AIMS consisted of a latex membrane, a sensing material layer, a copper electrode and a plastic air conduit. The latex membrane swells like a tent and thus separating from the sensing layer when gas inflation occurs, while shrinking and contacting the sensing layer during gas deflation. The AIMS showed an excellent sensitivity of 452.44% and linearity of 0.976 under a NO2 concentration

of 50 ppm. Breath patterns including normal, rapid, and deep breathing can be recognized through the peak intervals and amplitudes of the generated triboelectric signals. Inspired from the structure of human eardrum, Yang et al. developed a self-powered bionic membrane sensor (BMS) for cardiovascular characterization and voice recognition (Fig. 4d–f) [92]. Inspired by the human tympanic membrane, the BMS takes an oval shape and thereby achieves a wide frequency range response. A nylon thin film coated with indium tin oxide electrode was laminated onto a PET substrate. Then, inspired from the human eardrum, a PTFE tympanic membrane was placed on the tip of a tiny PET umbo, which was centered at the nylon film. Therefore, a conical cavity was formed between the PTFE and nylon film as spacer for

contact-separation. The BMS showed an excellent pressure sensitivity of 51 mV Pa −1 and a detection limit of 2.5 Pa in the low pressure range (<

1.2 kPa) and wide frequency range from 0.1 to 3.2 kHz. Zhang et al. designed a bionic-fin-structure assisted multilayer triboelectric nano-generator (BFM-TENG) for undersea energy harvesting (Fig. 4g–i) [93]. The bionic-fin-structure consisted of a fin, a connecting rod and a wedge shaped cone. Together with another two parts (a motion control slit and two multilayered TENGs), the BFM-TENG realized was able to harvest undersea surges and current energy from 3D directions.

Human skin is not only a protection layer with excellent mechanical properties but also a comprehensive and elegant tactile sensing system, stemming from its intrinsic structure [67,94,95]. Inspired by the struc-ture of biological cells, Wang et al. designed a stretchable and conformable TENG based on the patterned interconnected cellular structure, with physiological saline solution as the electrode and silicone rubber as the encapsulation and triboelectric layer (Fig. 5a, b) [96]. The bioinspired structural design enabled the TENG with excellent me-chanical strength and fracture toughness, thus sustaining the good

Fig. 4. TENGs with bioinspired structures from animals. a) Schematic structure of the alveolus-inspired membrane sensor (AIMS). b) Sensitivity of the sensing layer. c) Breath pattern recognition of the AIMS. Reproduced with permission [91]. Copyright 2020, American Chemical Society. d) Schematic structure of the self-powered bionic membrane sensor (BMS). e) Pressure sensitivity of the BMS. f) Cardiovascular signals monitoring and high frequency voice recognition. Reproduced with permission [92]. Copyright 2015, Wiley-VCH. g and h) Schematic illustration of the bionic-fin-structure assisted multilayer triboelectric nanogenerator (BFM-TENG). i) Photo of the fabricated BFM-TENG. Reproduced with permission [93]. Copyright 2020, Wiley-VCH.

(8)

performance under various strains (Fig. 5c). The gradient stiffness be-tween human-skin epidermis and dermis with interlocked microridge structures enabled effective pressure transmission to the underlying mechanoreceptors. Inspired by this structural mechanism, Ha et al. proposed a highly sensitive triboelectric sensor based on hierarchical nanoporous and interlocked microridge structured polymers with gradient stiffness (Fig. 5d, e) [97]. The skin-inspired hierarchical poly-mers consisted of a stiff P(VDF-TrFE) layer and a soft PDMS layer with different elastic moduli. The formed PDMS layer has open porous structures, while the P(VDF-TrFE) layer possesses inner pores and nanofiber-like surface. The microridges can serve as typical spacers to maintain effective contact and separation, resulting in an ultrathin and spacer-free device. The deformable porous polymers significantly in-crease the effective contact area and the amount of triboelectric charges. In addition to this, the gradient stiffness between the P(VDF-TrFE) and PDMS can provide effective stress transmission and concentration to microridges, leading to the significant deformation of interlocked microridge structures. As a result, the hierarchical nanoporous and interlocked microridge structured sensor demonstrated a pressure sensitivity of ~ 0.55 V/kPa and was able to detect weak artery pulse, bending strain and wide range frequency of dynamic stimuli (Fig. 5f).

Integrated electronic skin (e-skin) system consisting of both sensing elements and power supply elements is highly desirable for closely mimicking the complex human skin. Chen and co-workers developed a

finger-tip skin inspired e-skin system integrated with sliding sensing, pressure sensing and energy storage ability by mimicking the diverse structures and functions of human skin (Fig. 5g) [98]. This e-skin system consisted of three layers: fingerprint inspired triboelectric sliding sensor, epidermal-dermal inspired hybrid porous microstructure pressure sensor and subcutaneous fat inspired fabric based porous super-capacitor. In view of the main topic of this review, we will only discuss the fingerprint inspired triboelectric sliding sensor, which has the same spiral structure as that of the fingerprint. The sliding sensor consists of a PDMS substrate and four CNT-PDMS spiral electrodes, and works in single electrode mode (Fig. 5h). The specialized pattern of electrodes ensures that each electrode can generate signal one by one disregard of the direction of the object sliding across the sensor. In addition to the sliding direction sensing, sliding speed and displacement detection can also be realized by analyzing the number of voltage valleys and struc-tural parameters.

2.2. Bioinspired surface morphologies for TENG

Surface modification of triboelectric materials, such as micro/ nanopatterning [2], ion injection [99] and fluorination [100], is widely adapted to improve TENGs’ performance, since the effective contact area and surface charge density are of ultimate importance to the output. With the rapid development of TENGs, the surface modification

Fig. 5. TENGs with bioinspired structures from human skin. a) Schematic illustration of the biological cells bioinspired TENG. b) Schematic illustration of the working mechanism. c) Output voltage under strains. Reproduced with permission [96]. Copyright 2017, Elsevier. d) Schematic illustration of the epidermis–dermis structure and hierarchical nanoporous e) Structure of the interlocked microridge structure bioinspired triboelectric sensor. f) Pressure sensitivity of the sensor. Reproduced with permission [97]. Copyright 2018, American Chemical Society. g) Schematic illustration of fingertip skin inspired integrated e-skin system. h) Fingerprint inspired triboelectric sliding sensor and its simulated response when sliding along the -X direction. X-direction. Reproduced with permission [98]. Copyright 2018, Elsevier.

(9)

is not limited to only enhancing the instantaneous output power, but also for pursuing the long time stability by endowing the triboelectric material with a specific ability, such as self-cleaning [101,102].

Nature features magical phenomena, such as the self-cleaning lotus leaves, the superhydrophobic and adhesive rose petal, and anti-

reflective moth’s eye, all of which have been linked to the specific sur-face morphologies, also called micro/nanostructures. Even the most common leaves which are present ubiquitously in nature have a unique micro/nanostructure on their surfaces. In order to develop TENGs with such specific surface properties which exist in nature, researchers have

Fig. 6. Bioinspired surfaces through MEMS process. a) Schematic illustration of the interlocked TENG (i-TENG). b) SEM micrograph of the interlocked interface. c) Output comparison of four different types of TENGs. Reproduced with permission [103]. Copyright 2016, Elsevier. d) Microstructures on the toe-pad of the treefrog. e) Replication of convex PDMS. f) Instantaneous power comparison of the TENGs with different surfaces. Reproduced with permission [104]. Copyright 2019, Wiley-VCH. g) SEM micrograph of the lotus leaf surface. h) The fabricated hierarchical superhydrophobic PDMS surface. i) Output recovery after the TENG with the 3D hierarchical PDMS interlayer was dampened. Reproduced with permission [105]. Copyright 2019, Elsevier. j) SEM micrograph of the surface structure of moth’s eye. k) Schematic illustration of the hybrid energy harvester with the MM-TENG and solar cell. l) Comparison of total transmittance between the protective glass plate and MM-TENG. Reproduced with permission [106]. Copyright 2019, Elsevier.

(10)

fabricated biomimetic morphologies similar to the natural surfaces or replicated their morphology through soft lithographic templating. 2.2.1. Surfaces fabricated through MEMS process

Applying micro/nanopatterning methods to create micro/nano-structures on the triboelectric surface is the most straightforward and simple method with controllable total area and unit size. Typically, micro/nanostructures are fabricated through mature but sophisticated fabrication procedures, including lithography and etching, which are widely used in fabricating microelectromechanical systems (MEMS) devices. Inspired by the unique morphology of beetle wings, Choi and co-workers introduced an interlocked TENG (i-TENG) in which the interlocked interface was formed between the nano-pillar Ni electrode and nano-pillar PDMS composite thin film (Fig. 6a–c) [103]. PDMS nano-molding and Ni electrodeposition methods were applied to fabri-cate the PDMS nano-pillar arrays and Ni nano-pillar arrays using a polyvinyl chloride template with nanoholes. The interlocked interface not only dramatically increased the effective contact surface area but also the normal and shear friction between the interface which can led to efficient charge separation. The output performance of the TENG with interlocked interface was demonstrated to be almost 12 fold higher than that of the flat TENG (Fig. 6c). Bui et al. designed a treefrog toe pad-inspired micropatterned TENG in which nonclose-packed microbead arrays were fabricated on the surface of the triboelectric material to enhance the electrical performance (Fig. 6d–f) [104]. The first step of the fabrication approach was the large-scale fabrication of a concave PMMA film via an improved phase separation method and the second step involved replication of convex PDMS by using the concave PMMA as template. The size and separation of micro microlens arrays can be precisely controlled. The bioinspired TENG with micropatterned PDMS and flat Al has a high power density of 8.1 W m−2 under a small

pressure (Fig. 6f).

As the electrical performance of TENGs has drastically improved, the need for improved stability and robustness of TENGs for real-time ap-plications in harsh environments has attracted increasing research attention. Zhou et al. used particle lithography method to fabricate a lotus leaf surface morphology inspired 3D hierarchical super-hydrophobic PDMS surface which is beneficial to the electrical perfor-mance enhancement, humidity resistance and contamination resistance of the TENG (Fig. 6g, h) [105]. The fabrication process includes photolithographically fabricating bottom and top photoresist grid layers with larger square holes, releasing PS particles on the final grid mold, and pouring PDMS on the mold. In comparison to TENGs with other kinds of PDMS interlayer, TENG with the 3D hierarchical PDMS showed a much better electrical performance due to the increase in effective contact area and higher capacitance. In addition to the performance enhancement, the TENG can not only retain up to 86% initial output in high relative humidity condition and recover very fast (Fig. 6i), but also maintains 88% of initial output after contamination and cleaning, which revealed the high stability and robustness of the TENG.

Integration of the traditional solar cell with the water-based TENG was demonstrated to create a possibility to harvest energy both in sunny and rainy days. However, the water-based TENG which is usually located on the outermost surface can cause the degradation of optical characteristics of solar cells, thus optical transparency of the water- based TENG is required to be at least as good as that of the traditional protective glass plates. Inspired from the anti-reflective property of the periodic tapered nanopillar structures of the moth’s eyes (Fig. 6j), Yoo et al. proposed a moth’s eye mimicking TENG (MM-TENG) with outstanding light transmittance [106]. The MM-TENG was composed of all transparent parts, including moth’s eye mimicking quartz glass (MM-glass), interdigitated AgNW electrode, and optically clear adhesive (Fig. 6j, k). The MM-glass was fabricated using plasma etching process with a self-assembled PS nanoparticle mask layer. The MM-TENG demonstrated a total transmittance over 90% ranging from 480 to 800 nm and its solar-weighted transmittance is 89.41%, which was

0.01% higher than that of the traditional protective glass plate (Fig. 6l). As a result, the efficiency of MM-TENG-covered solar cell was increased by 0.17% compared with protective glass plate-covered solar cell. 2.2.2. Surfaces replicated from natural surfaces

In terms of the production cost and process complexity, the sophis-ticated micro/nanofabrication procedures adopted above may be not the best choice, despite the advantages of high precision, mass pro-duction and controllable parameters. An alternate and efficient method to achieve bioinspired surface topographies is to directly replicate the microstructures from existing natural surfaces using soft lithography technology. Inspired by the evolution-optimized microstructure of the leaf surface, Sun et al. replicated the leaf morphology on PDMS to construct an efficient TENG harvester by PDMS templating of a ramified leaf, which was simple, green and environment-friendly (Fig. 7a) [107]. The surface topography of the PDMS, including the size, microstructure, unit density etc., can be tuned by choosing different types of leaves or different regions of the same leaf as the template. Yao et al. fabricated a bioinspired interlocked triboelectric interface through the facile repli-cation of the cone-like array microstructures of the Calathea zebrine leaf (Fig. 7b–d) [108]. Their soft lithography process has two steps: first molding of the original Calathea zebrine leaf endowing a negative PDMS template with reverse patterns and second molding on the negative PDMS template to generate the PDMS layer with the cone-like array microstructures. In their work, micron or submicron scale tiny PTFE burrs were further generated on top of the microstructured PDMS sur-face. The combination of the bioinspired interlocked interface and PTFE tiny burrs dramatically enhanced the triboelectrification, resulting in high sensitivity for self-powered pressure sensing. The e-skin sensor with bioinspired interlocked interface and PTFE burrs not only had much higher sensitivity than other sensors with different surface conditions (Fig. 7e), but also were found to be sensitive than other reported pres-sure sensors demonstrated in previous papers.

Seol et al. compared the ability to enhance the effective contact area of three bioinspired surfaces, which were replicated from the lotus leaf, the rose petal, and the cicada wing, respectively (Fig. 7f–h) [109]. The replicas of the three surfaces showed microscale rod-like structures with nanoscale bumps, semi-circle patterns with nanoscale valleys and nanoparticle-like structures, respectively. The output of TENGs with the three bioinspired surfaces and a control one with flat morphology fol-lowed the order of flat < lotus leaf < rose petal < cicada wing, according to the higher aspect ratio and higher packing density. Interestingly, by using the water-soluble PVA as negative template material for the replication of rose petal, Zhu et al. realized the final PDMS replica with relatively uniform hierarchical structures despite the plump and shriv-eled parts of the petal (Fig. 7i, j) [110]. The output of the TENG with bionic microstructure patterned PDMS is much better than the TENG with smooth PDMS (Fig. 7k). Inspired by the natural liquid-solid contact electrification on the dried lotus leaf and its self-cleaning ability, Choi et al. fabricated a lotus leaf pattern-assisted triboelectric replicable nanogenerator (LL-PATERN) through thermal nanoimprinting replica-tion technique (Fig. 7l, m). The LL-PATERN demonstrated not only both enhancement in short-circuit current and open-circuit voltage compared to the flat-PATERN without topography, but also a good anti-contamination property with a sustained electrical output perfor-mance due to a self-cleaning effect.

2.3. Bioinspired materials for TENG

Despite of the wide selection of triboelectric materials due to the universality of the contact electrification effect, seeking and developing novel functional materials such as the ones with higher electronegativity and dielectric constant are still of great significance. Apart from improving the electrical performance of TENGs, more and more mate-rials’ functions are designed and proposed to meet the requirements of various applications. For example, TENGs are required to be flexible and

(11)

Fig. 7. Bioinspired surfaces through direct replication. a) Schematic illustration of the fabrication process of the patterned PDMS. Reproduced with permission [107]. Copyright 2016, Elsevier. b) Schematic illustration of the e-skin sensor with bioinspired interlocked triboelectric interface. c) Photograph of Calathea zebrine leafs. (Inset: SEM image of microstructures of the Calathea zebrine leaf) d) Replicated biomimetic microstructures. e) Comparison of the sensitivity between sensors with different microstructures and surface conditions. Reproduced with permission [108]. Copyright 2019, Wiley-VCH. f–h) Photographs of the lotus leaf, rose petal, and cicada wing. The insets are SEM micrographs of the PDMS replica of the surfaces. Reproduced with permission [109]. Copyright 2014, Wiley-VCH. i) SEM image of shriveled rose petal and PVA covered petal. j) SEM image of the micropatterned PDMS. k) Comparison of the output of TENGs with micropatterned PDMS and smooth PDMS. Reproduced with permission [110]. Copyright 2020, Elsevier. l) Measurement of the electric potential before and after sequential contact and detachment of the water droplet on the natural lotus leaf. m) The self-cleaning effect of the lotus leaf pattern-assisted triboelectric replicable nanogenerator (LL-PATERN). Reproduced with permission [111]. Copyright 2017, Elsevier.

(12)

must provide use-of-comfort for wearable devices. Biocompatibility and biodegradability are highly desired for wearable/implantable applica-tions and for widely distributed use of TENG-based energy harvesters and self-powered sensors to avoid environmental pollution.

Nature is no doubt a great learning source for developing new ma-terials and improving existing mama-terials. For example, mussel-inspired underwater glue was developed by learning the way mussels stick to surfaces underwater and withstand the strongest waves [112,113]. To reduce the drag force of swimsuits and airplanes, the surfaces of mate-rials were modified with denticles, which are found on the shark skin [114,115]. A diverse range of triboelectric materials have been devel-oped in the literature so far by drawing inspiration from nature. 2.3.1. Bioinspired porous triboelectric materials

Surface treatment of triboelectric materials through micro/nano-patterning and surface functionalization have greatly improved the surface charge density of TENGs. However, there is still enormous

potential for research in this regard to further increase the charge den-sity for the long-term deployment of TENGs. Implementation of porous materials is an ideal method to further enhance charge density because of the ability of porous structures to trap charges, which are called body charges. Inspired by the water hyacinth petiole (WHP) that contains a large number of air-filled chambers, Yu et al. fabricated a biomimetic porous structure nanofiber-based triboelectric nanogenerator (BN- TENG) with ultrahigh transfer charge density (Fig. 8a–c) [116]. The porous-structure nanofiber was fabricated using a two-component co-axial electrospinning method by delivering Polyvinylidene fluoride (PVDF) solution and mineral oil through inner and outer channels, respectively. The fabricated biomimetic WHP structures in the nanofiber were able to generate large amount of induced charges, and can hold charges on the surface of the nanofiber mat, thus enhancing the surface charge density. Since the WHP nanofiber initially was not closed on its end, a pressure treatment with the honeycomb mold was conducted to alleviate the charge dissipation, dividing the nanofiber mat into many

Fig. 8. Bioinspired porous triboelectric materials. a) Schematic illustration of the biomimetic nanofiber-based triboelectric nanogenerator (BN-TENG). b) Schematic illustration of the honeycomb mold pressure-treated PVDF nanofibers. c) FE-SEM micrograph of the porous BWHP structure inside the PVDF nanofibers. The inset is the BWHP. Reproduced with permission [116]. Copyright 2017, Elsevier Ltd. d) Schematic picture of the bird’s nest. e) fabrication of the porous hybrid. f) SEM image of the porous structure of the hybrid. Reproduced with permission [117]. Copyright 2020, Elsevier Ltd. g) Schematic illustration of a TENG with Trimurti PVDF. h) SEM micrograph of the nanoporous cancellous-bone-like structure. i) SEM micrograph and water contact angle of upper and bottom surfaces. Reproduced with permission [118]. Copyright 2019, Elsevier.

(13)

orthohexagonal parts and restricting charges in each part. Finally, an ultrahigh transfer charge density of about 364μC m−2 was achieved,

which was the maximum value measured at the atmosphere compared to other works. Inspired from bird’s nest, Bui et al. developed a hierar-chically porous dielectric-electrode hybrid for TENG (Fig. 8d–f) [117]. The porous hybrid is composed of a PMMA nanosponge which can offer large friction area and humidity-resistance and anticorrosion properties, and a copper woven mesh which serves as flexible support and elec-trode. The TENG assembled with the porous hybrid can generated an output voltage of 560 V and power density of 13.9 W m−2. Moreover,

the TENG can sustain its high output under high humidity and exposed into saline solution.

The key merits such as simple device configuration, flexibility, and stretchability rank TENGs on the top of the list of candidates for wear-able self-powered sensors and energy devices. Apart from the electrical performance, output stability, biocompatibility and comfort-of-use are also important indexes for wearable TENGs. Zhang et al. fabricated a bioinspired hydrophobic/cancellous/hydrophilic Trimurti PVDF material-based TENG with considerable electrical performance, good stability and use comfortability (Fig. 8g–i) [118]. The fabricated bio-inspired Trimurti PVDF had the hydrophobic lotus-leaf-like structure on the front surface, nanoporous cancellous-bone-like structure in the interior and hydrophilic root-xylem-like structure on the back surface. The PVDF mat was termed as Trimurti because it contained three bio-mimetic properties simultaneously, resulting in the TENG with consid-erable electrical performance, good stability and use comfortability simultaneously. Firstly, the inner cancellous-bone-like structure allowed the tribo-material to trap charges not only on its surface but also in the inner pores, thus enhancing the electrical performance. Secondly, the lotus-leaf-like hydrophobic front surface offered the TENG to maintain 22% of the original output under a high relative humidity of 85%, showing good output stability. Thirdly, the root-xylem-like hydrophilic bottom surface made the sweat spread out on the bottom surface and the evaporation process can be faster, improving comfort of the wearable. 2.3.2. Bioinspired biocompatible triboelectric material

With improved performance and robust functionality in real-time applications, TENGs are believed to be applied and distributed glob-ally to address various self-powered sensing and energy harvesting needs. Due to the high demand, such large numbers of TENG devices may encounter the problems of mass fabrication cost and environmental compatibility. Therefore, adapting natural sustainable and eco-friendly materials for TENGs is a very promising method. Silk, one of the strongest natural fibers, possesses the properties of both high mechani-cal strength, good biocompatibility and superior electronegativity. Inspired by the unique characteristics of the silk, Kim at el. reported a silk fibroin based biocompatible triboelectric nanogenerator (Silk Bio- TENG) fabricated using electrospinning (Fig. 9a, b) [119]. This nanofiber-network structure had a positive effect on the triboelectric

power generation because of the enhanced surface roughness and ul-trahigh surface-to-volume ratio. The output peak voltage of the Silk Bio-TENG demonstrated a 1.5 fold performance enhancement as compared to the cast Silk Bio-TENG. In addition, as a result of the excellent mechanical properties of silk fibers, the Silk Bio-TENG dem-onstrates unique stable and durable performance (Fig. 9c).

2.4. TENGs bioinspired from natural mechanisms and properties Nature not only inspires us to develop and optimize structures, sur-face morphologies and materials for TENGs, but also features innovative designs to explore for the creation of novel and functional TENGs, which mimic natural mechanisms and properties. There are many examples of animals which use electrostatic field for sensing, and also some that portray functionalities quite similar to power generation in TENGs. For the sensing example, cockroaches use electrostatic field to detect near- by obstacles [120]. It is also known that bees can also seek flowers though the detection of the weak electrostatic field [121]. The electric eel, a type of electric fish is known to generate high potential in order to prey or defense. The mechanisms employed by these animals to perceive their surroundings by using electrostatic field and to generate electricity can naturally inspire researchers to translate such novel mechanisms to TENGs and design innovative TENG-based devices. In addition to the natural mechanisms, some specific properties in natural creatures also inspire us to design fascinating TENGs from a perspective beyond the structure, surface and material inspiration.

2.4.1. TENGs bioinspired from natural sensing/power generation mechanisms

To perceive and escape from predators, cockroaches use their antennae to detect approaching objects around them. The electrostatic equilibrium between its antennae and leg disturbed by an object re-minds the cockroach to escape. Inspired by the object detection mech-anism of cockroaches’ antennae, Wang et al. proposed a triboelectric bionic-antennae-array (BAA) sensor to detect noncontact motions (Fig. 10a) [122]. The BAA sensor was composed of Ag-coated polyimide fibers, which were sewn on the surface of a PTFE film. Under the PTFE film, a grounded sensory circuit was electrically connected to those fi-bers, thus forming single-electrode TENG. When an object gradually approached the BAA sensor, its electrostatic balance was disturbed and then the open-circuit voltage of the TENG was changed. As a result, by detecting the change in voltage and polarity, approaching distance and speed of the object could be recognized. The effective distance detection range of the bionic BAA sensor could reach 180 mm with a displacement resolution of 1 mm and a maximum sensitivity of 5.6 V mm−1.

Electric eel has the ability to generate high voltage underwater based on ion channels controlled by special electric organs [123]. When stimulated by external incentives, the neuro signals received by the electric organs cause the opening of the ion channels of the electrolyte

Fig. 9. Bioinspired biocompatible triboelectric material. a) Photograph of silk (I) and schematic illustration of the fabrication method: electrospinning (II). b) FE- SEM micrograph of fabricated silk fibroins. c) The stability and durability of the Silk Bio-TENG. Reproduced with permission [119]. Copyright 2016, Wiley-VCH.

(14)

thereby allowing the ions to flow through and then reversing the po-larity momentarily. The transmembrane potential of each electrocyte can reach up to ~ 150 mV and a high voltage of almost 600 V can be generated resulting from the in-series stack of thousands of electrocytes [124]. By mimicking the working mechanism of the ion channels in the electric eel, Zou et al. proposed a bionic stretchable nanogenerator (BSNG) for underwater energy harvesting and sensing (Fig. 10b, c) [125]. The BSNG consisted of two layers. The first layer was PDMS-silicone double layer, with a series of controllable channels and a

fluid chamber filled with electrification liquid, serving as the electrifi-cation layer. The controllable channels were realized by cutting the PDMS layer into multiple sections, so they would open after stretching and close after releasing due to the elastic resilience of the silicone. The second layer was an induction layer, containing two liquid electrodes under the channels and a chamber of the electrification layer. The BSNG can be stretched over 60% due to the excellent stretchability of silicone and the segmented structure of PDMS. The BSNG was demonstrated to generate an open-circuit voltage of 10 V in one working cycle

Fig. 10. TENGs bioinspired from natural mechanisms. a) Schematic illustration of the bionic-antennae-array (BBA) sensor and mechanism of object detecting of cockroaches. Reproduced with permission [122]. Copyright 2019, Wiley-VCH. b) Schematic illustration of the electric eel, electrocytes, and ion channels. c) Schematic illustration of the bionic stretchable nanogenerator (BSNG), artificial ion channels. Reproduced with permission [125]. Copyright 2019, Nature Pub-lishing Group.

(15)

underwater. Inspired by the power stacking mechanism employed in electric eel’s electric organs, Jie et al. designed a bionic multiple-layer TENG for current enhancement [126]. The bionic TENG was made up of 5 TENGs connected in parallel, taking full advantage of the limited space and largely improving the output current, which reached up to 260 µA.

2.4.2. TENGs bioinspired from natural properties

Not only the power generation mechanism of the electric eel but also its flexible, soft and resilient skin can give inspirations to the design of TENGs. Inspired by the soft and resilient properties of the skin of electric eel, Lai et al. presented an intrinsically mechanically durable and resilient skin-like triboelectric nanogenerator (SLTENG) (Fig. 11a, b) [127]. They adapted the super-soft and tough Eco-flex 00-10 silicone rubber, which can penetrate into the voids of silver-nanowire networks and form the single electrode SLTENG. The SLTENG could generate triboelectricity even under various extreme deformations, such as

multiple twists, folds and even stretches of over 300% strain. Inspired by the strong elastic resilience ability of jellyfish underwa-ter, Chen et al. reported a waterproof and shape-adaptive bionic-jellyfish triboelectric nanogenerator (bjTENG) with unique elastic resilience (Fig. 11c, d) [128]. The bjTENG consists of a multilayered TENG encapsulated into a PDMS package, which enabled the bjTENG with resilient property, making it possible to detect the water level and fluctuations with high sensitivity. When the device was tested in an air-water interface, the output of the interface bjTENG increased approximately linearly with the increase of water fluctuation frequency. Therefore, the bjTENG was able to act as a self-powered fluctuation sensor to monitor the pressure changes at the air-water interface (Fig. 11e). With an optimized elastic resilient structure of force assis-tance resembling breaststroke, the bjTENG could effectively harvest underwater wave energy and flow energy (Fig. 11f, g).

Fig. 11. TENGs bioinspired from natural properties. a) Schematic illustration of the skin-like triboelectric nanogenerator (SLTENG). b) SLTENG under various extreme deformation. Reproduced with permission [127]. Copyright 2016, Wiley-VCH. c) Schematic illustration of the bionic-jellyfish triboelectric nanogenerator (bjTENG). d) Elastic resilience ability of the bjTENG. e) Schematic illustration of the water interface bjTENG. f, g) The bjTENG with optimized elastic resilient structure to harvest underwater energy. Reproduced with permission [128]. Copyright 2017, Elsevier.

(16)

3. Biomimetic applications of TENG

Imitating the human perception capabilities, including the environ-mental sensing, signal transduction and neural signal processing ability to realize the intelligent robotics, man-made organs and artificial in-telligence have always been scientists’ pursuit. Biomimetic devices that mimic the functions from specific human body parts have demonstrated huge potential, especially in the biomedical field. However, power supply is a critical problem for current biomimetic devices, wherein rigid, bulky and exhaustible batteries are used. To solve the energy problem, developing self-powered biomimetic systems with self- powered sensing and environmental energy harvesting capability is highly desired. Amongst all the energy transduction technologies, TENGs exhibit great prospects due to their enormous advantages, such as ease of fabrication, low-cost, high-output voltage, stimuli- ultrasensitivity, flexibility, stretchability, easy-integrability and wide material choice.

3.1. TENG enabled electronic skin

As discussed previously, the complicated and advanced structures in human skin on one hand provide us inspiration to develop optimized TENG devices with excellent performance and unique properties. On the other hand, the self-powered sensing, flexible and integrative functions of TENG allow them to act as artificial electronic skins to sense envi-ronmental changes. Until now, many types of TENG-based electronic skin have been proposed, and a detail review on this specific topic was conducted earlier [129,130]. In view of this, here we only focus on some recent representative works.

Pu et al. proposed an ultrastretchable, transparent and soft tribo-electric electronic skin for both biomechanical energy harvesting and tactile sensing (Fig. 12a–c) [131]. Elastomer and ionic hydrogel were used as the electrification layer and electrode, respectively, resulting in the simultaneous achievement of high transparency and ultrahigh stretchability. This electronic skin can not only scavenge human motion energy, but is also pressure-sensitive with a sensitivity of 0.013 mV kPa−1. For demonstration, a 3 pixel × 3 pixel artificial skin

Fig. 12. TENGs enabled artificial electronic skin. a) High transparency of the e-skin to full visible colors. b) Ultrahigh stretchability of the electronic skin. c) A 3 pixel ×3 pixel artificial skin. Reproduced with permission [131]. Copyright 2017, American Association for the Advancement of Science. d) Schematic illustration of the skin-inspired TENG (SI-TENG)-based electronic skin. e) Electronic skin as voice vibration detector. f) Electronic skin enabled intelligent prosthetic hand. Reproduced with permission [132]. Copyright 2018, Wiley-VCH. g) Schematic illustration of the stretchable triboelectric-photonic smart skin (STPS). h) Pressure sensitivity under different strains. Reproduced with permission [133]. Copyright 2018, Wiley-VCH.

(17)

was proposed to perceive the touch position and pressure (Fig. 12c). A stretchable and washable skin-inspired TENG (SI-TENG), composed of silver-coated conductive yarn network electrode and silicone rubber electrification layer, was proposed by Dong et al. as electronic skin (Fig. 12d) [132]. The yarn network electrode was sandwiched using silicone rubber and uniformly distributed, enabling the SI-TENG with good transparency. The SI-TENG-based electronic skin can be applied as a wearable power source for biomechanical energy harvesting, while it can also detect weak physiological signals like voice vibrations and can be integrated into the intelligent prosthetic hand to identify finger tactile forces due to its high pressure sensitivity (Fig. 12e, f). Bu et al. developed a stretchable triboelectric-photonic smart skin (STPS), which was able to sense multidimensional tactile and gesture signals (Fig. 12g) [133]. The STPS consisted of a microcracked metal film, aggregation-induced emission (AIE)-mixed silicone rubber and a layer of conducting AgNW network. When the STPS was stretched, distributed microcracks with their size related to the degree of strain arose in the metal film. The underlying AIE silicone rubber exhibited photo-luminescence phenomenon when exposed to light. The silicone rubber and AgNW electrode made up the TENG part, which was sensitive to the vertical pressure with an extremely high pressure sensitivity of 34 mV Pa−1 in the low region (Fig. 12h).

3.2. TENG enabled artificial neuromorphic devices

A typical neuromorphic action consists of stimulation perception realized by mechanoreceptors, and signal transduction and processing completed by synapses. Developing self-powered devices for artificial mechanoreceptors, self-powered transduction without external power supplies and integrated artificial neuromorphic devices is highly desired for disabled assistance, human–machine interaction and artificial intelligence.

Cochlear implant is the most feasible approach for some irreversible hearing loss caused by the loss of hair cells. By mimicking the functions of cochlear tonotopy: passive frequency selectivity of the basilar mem-brane and acoustic to electric energy conversion of hair cells, artificial basilar membranes (ABMs) are attracting increasing attention for their potential in hearing aid [134,135]. Jang et al. proposed a triboelectric-based ABM (TEABM) to mimic cochlear tonotopy (Fig. 13a) [136]. The TEABM consists of an array of beams of different lengths, where each beam can respond to a different resonance frequency. In the presence of an acoustic stimuli, acoustic-to-electric transduction occurred due to triboelectrification between Kapton film and Al foil. The basal region had low frequency sound response while apical region reacted to high-frequency sound. The frequency selectivity of the TEABM was in the range of 294.8–2311 Hz; and its sensitivity to the incoming sound pressure was measured in the range of 1.74–13.1 mV/P, increasing as the channel number increased (Fig. 13b).

Developing intelligent tactile sensors which can communicate with nerves is the basis for building the artificial nervous system. Inspired by the somatosensory signal generation and neuroplasticity based signal processing, Wu et al. developed a biomimetic neuromorphic tactile sensor with the ability of learning and memory based on a well-designed single electrode TENG (Fig. 13c, d) [137]. The key component of the TENG is a polyimide:reduced graphene oxide (PI:rGO) hybrid negative triboelectric layer. The rGO sheets in the PI:rGO hybrid layer can act as body charge traps which allow charges on the surface spontaneously transferred to the inside rGO sheets, thus increasing the total number of triboelectric charges. The significant feature of the neuromorphic tactile sensor was that its output increased with the increasing number of presses, similar to the biological excitatory postsynaptic potential. From the point of bionics, the gradually increasing output can be regarded as the “learning” behavior of humans, which means that the output con-tained the information not only about current press but also the history of previous stimulations. Besides, the decay characteristics of the output resulted from charge dissipation showed a similarity to “the forgetting

curve” of humans, which has characteristics of a first abrupt drop and following slow decay.

Liu et al. proposed a self-powered artificial synapse (SPST) by combining a TENG and an electric-double-layer organic field effect transistor (Fig. 13e) [138]. The TENG could serve as a tactile perception sensor and provide pre-synaptic spikes to the SPST when touched by external objects, realizing self-powered tactile synapse. The output voltage of the TENG was connected to the gate of the transistor, acting as pre-synaptic stimulation, and the source-drain channel in the transistor served as post-synapse. By touching the TENG, the output characteristics of the SPST exhibited the similarity to essential biological synaptic functions, including excitatory postsynaptic current (EPC) (Fig. 13f), paired-pulse facilitation (PPF), dynamic filtering, and short-term plas-ticity to long-term plasplas-ticity. Zhang et al. applied chitosan electrolyte gate transistor and silk-fibroin triboelectric material in their biological tactile sensory neuron to realize an energy-efficient and less cumber-some fabrication method (Fig. 13g) [139]. Due to the high proton con-ductivity of the chitosan gate electrolyte, there was a clear hysteresis loop when the gate voltage increased and decreased. Apart from suc-cessfully emulating typical synaptic functions like EPC and PPF, the web-weaving spider’s sensing ability to identify preys which are stranded on the spider web and wind flows is also emulated (Fig. 13h).

4. Summary and perspectives

In this review, bioinspired TENGs and TENG enabled biomimetic applications are comprehensively and systematically discussed to the best of our knowledge. The device structure, surface morphology and triboelectric material are three main starting points for optimizing TENGs, whether the pursuit is to improve the electrical performance or mechanical performance. Accordingly, we firstly discussed bioinspired TENGs in which at least one of the three aspects: structure, surface morphology and material is designed by mimicking nature. Structures in nature provide inspirations not only to design integrated, compact and mechanically-stable TENGs for efficient ambient energy harvesting but also to improve the sensitivity and accuracy of TENG-based self-pow-ered sensors. Replicating natural surface morphologies in triboelectric interfaces and in other parts of a TENG can on one hand increase the total contact area, and on the other hand equip the TENG with special functions such as self-cleaning ability and antireflective nature. Nature provides various prototypes to develop porous triboelectric materials which can greatly increase the total charge density due to the extra body charge traps. Reprocessing natural biocompatible and durable materials into forms suitable for TENGs while maintaining the original properties is an effective and low cost method to develop bionic triboelectric ma-terials. Thereafter, novel TENGs bioinspired from the unique mecha-nisms and properties in nature are also discussed. The sensing and power generation mechanism of some animals are to some extent similar to TENGs in the physical principal or function, offering inspirations to design novel TENG-based devices. In addition to the natural mecha-nisms, some specific properties in natural creatures also inspire us to design fascinating TENGs from a perspective outside the structure, surface and material. Secondly, TENG enabled biomimetic applications including artificial electronic skin and artificial neuromorphic devices are reviewed. To mimic human skin, TENGs are designed to be stretchable, flexible and conformable in order to be intimately attached to objects of any shape, and multifunctional to sensing environment changes. For artificial neuromorphic devices, TENGs serve as both self- powered sensors and stimulators with relatively simple structure but complicated integration and materials.

Tremendous progress have been achieved in improving the perfor-mance of TENGs, and various TENG-based applications have been suc-cessfully demonstrated. However, there are still challenges to further optimize the TENG in order to bring it close to real-life applications. Drawing inspirations from nature seems to be the right and optimal path ahead to address this challenge to provide opportunities to design and

(18)

Fig. 13. TENG enabled artificial neuromorphic devices. a) Schematic illustration of the triboelectric-based ABM (TEABM). b) Frequency response of the TEABM. Reproduced with permission [136]. Copyright 2016, Wiley-VCH. c) The TENG-based neuromorphic tactile sensor with functionally emulated mechanoreceptor and neuromorphic systems. d) Schematic illustration of the tactile sensor and PI:rGO hybrid layer. Reproduced with permission [137]. Copyright 2020, American Chemical Society. e) Schematic illustration of a biological synapse and the self-powered artificial synapses (SPST). f) Excitatory postsynaptic current (EPSC) with different touch frequencies of TENG. Reproduced with permission [138]. Copyright 2019, Elsevier. g) Schematic illustration of the tactile sensory neuron with chitosan electrolyte gate transistor and silk-fibroin triboelectric material. h) Simulation of the web-weaving spider’s sensing ability. Reproduced with permission

(19)

optimize TENGs for practical applications since the continuous self- optimization with evolution in nature provides a plethora of innova-tive and simple solutions. Besides the reviewed works, it is anticipated that more and more paradigms that combine bionics and TENGs will be proposed in the near future to optimize TENG and develop TENG- enabled biomimetic devices. The following perspectives, research di-rections and futuristic research ambitions can be considered to help solve the existing challenges of TENG and develop new applications, as illustrated in Fig. 14.

1) More structures can be inspired from nature to expand the applica-tion scenarios and improve the electrical/mechanical performance of TENGs. Even though the basic working modes of the TENG are fixed, the morphological structures of TENG devices should be rationally designed to better adapt to the application scenarios. Besides, structural optimization of TENG should be conducted to realize effective contact-separation and micro-stimulus (ultra-low frequency and amplitude) response, which can largely improve the perfor-mance and application field of TENG. Learning the way natural creatures interact with the specific environments can help to develop and optimize the structures of TENGs.

2) Damage-tolerant, cost-effective and functional bioinspired surface morphologies are expected to be learnt and translated from nature. The micro/nanostructures on the contact interfaces are often easily damaged after continuous contact and friction, thus damage-tolerant surface morphologies are desired to create durable TENGs with sustainable long-term performance. However, current bioinspired surface morphologies are mainly fabricated using complicated and expensive MEMS processes or the direct replication method which is almost only applicable for soft materials. Besides, bioinspired func-tional triboelectric surface morphologies are gaining increasing

attention for the benefit of both TENGs’ instantaneous output and long-term stability. Therefore, more efforts should be made to develop cost-effective methods to fabricate damage-tolerant bio-inspired surface morphologies for TENGs.

3) Bioinspired materials which are biodegradable, biocompatible and porous are highly desired. It can be anticipated that more and more TENG-based self-powered systems will be applied in the oncoming wave of implant and wearable devices and distributed internet of things, especially in the uninhabited wild. As a result, materials applied in TENGs should not have the adverse impact on the sur-rounding environment when it no longer works, since replacing or collecting the damaged devices would consume a lot of manpower, material and financial resources. Developing porous materials with large surface area to volume ratio is still highly desired and in the future more novel 2D and 3D porous materials are expected to draw inspiration from natural materials.

4) From the point of view sensing principles and power generation mechanisms, there exists a vast scope of knowledge in nature to be derived and mimicked. There are many animals that use electrostatic field to navigate and detect potential predators, which can inspire the development of special sensors based on TENG. The mechanism through which electric eels realize voltage stack may inspire us to achieve effective multitude increase in the output integration of different TENG units since the output phase of every TENG is quite different. Besides, specific properties in natural creatures may also give us inspiration to develop novel TENG devices, such as durable but flexible TENGs.

5) More research efforts should be made to develop integrated and multifunctional, self-healable TENG-based electronic skins. In the future e-skins are definitely expected to have integrated multi-

Referenties

GERELATEERDE DOCUMENTEN

De enige beoordelings- maatstaf zou daarbij moeten zijn of er voldoende mogelijkheden voor de Nederlandse politie en Justitie zijn respectievelijk gevonden kunnen worden om

Slepende beleidscontroverses hebben maar al te vaak betrekking op conflicterende frames of denkkaders en kunnen dan ook meestal het best worden opgelost door reframing, dat wil

In de voorbije 450 miljoen jaren manifesteerden zich meerdere ijs- tijden, met lage zeewaterspiegels, gevolgd door zeer warme perio- des waarbij de lage landen volle- dig onder

3.2 The native entrepreneurs and their role in the Indies economy Many European scholars such as Boeke state that indigenous merchants played an insignificant role in developing

We provide direct measurements of the boundary layer properties in highly turbulent Taylor-Couette flow up to Re ¼ 2  10 6 (Ta ¼ 6:2  10 12 ) using high-resolution particle

In our model the most ef ficient gait (i.e. with the least energy losses) was achieved with highest ankle plantar- flexion velocity just before contralateral foot strike, a

row (bright colours) is for light off condition and bottom row (faded colours) for light on. Normalized responses differed between light off and light on, in a layer-dependent