Sweat-based wearable energy harvesting-storage hybrid textile devices

Jian Lvab, Itthipon Jeerapana, Farshad Tehrania, Lu Yina, Cristian Abraham Silva-Lopeza, Ji-Hyun Janga, Davina Joshuiaa, Rushabh Shaha, Yuyan Lianga, Lingye Xiea, Fernando Sotoa, Chuanrui Chena, Emil Karshaleva, Chuncai Kongb, Zhimao Yangb and Joseph Wang*a
aDepartment of NanoEngineering, University of California, San Diego, La Jolla, California 92093, USA. E-mail: josephwang@ucsd.edu; Fax: +1 8585349553; Tel: +1 8582460128
bSchool of Science, Xi’an Jiaotong University, Xi’an 710049, Shaanxi, China

Received 22nd September 2018 , Accepted 11th October 2018

First published on 11th October 2018

This study demonstrates the first example of a stretchable and wearable textile-based hybrid supercapacitor–biofuel cell (SC–BFC) system. The hybrid device, screen-printed on both sides of the fabric, is designed to scavenge biochemical energy from the wearer's sweat using the BFC module and to store it in the SC module for subsequent use. The BFC relies on lactate, which is oxidized enzymatically to generate electricity. The generated bioenergy is stored directly and rapidly in the printed in-plane SCs. The SC energy-storage module employs MnO2/carbon nanotube composites that offer high areal capacitance and cycling electrochemical stability. Both printed SC and BFC devices rely on optimal elastomer-containing ink formulations and serpentine structure patterns that impart a stable electrochemical performance after a variety of mechanical deformations. Such a fabrication route ensures that the energy-harvesting and storage properties of the two integrated devices are not compromised. The SC–BFC hybrid system can thus deliver stable output over long charging periods, boost the voltage output of the BFC, and exhibit favorable cycling ability. Such attractive performance, demonstrated in successful on-body testing, along with the unique architecture and low-cost scalable fabrication, make the new garment-ased hybrid energy device useful for meeting the power and mechanical resiliency requirements of wearable electronics and smart textiles.

Broader context

Realizing the high performance of wearable electronic devices requires the judicious integration of powerful energy sources onto conformal platforms. The flexible textile-based integrated hybrid energy device described in this study represents the first demonstration of a wearable self-power energy system, in which a supercapacitor (SC) is used for storing bioenergy harvested by the sweat-based biofuel cell (BFC). On-body BFCs scavenging energy from biofluids represents a promising ‘green’ and sustainable approach to power wearable electronic devices. However, BFCs can generate and deliver power only in the presence of biofluid, and the sweat biofuel concentration often fluctuates. Therefore, using BFCs alone limits constant power output due to the irregular perspiration of the user. This new textile-based hybrid energy device bridges the gap of the requirements. Such a flexible system extracts, boosts, and stores energy harvested from human sweat, using lactate as a fuel, and delivers an enhanced and stable output over long periods. Leveraging the screen-printing technique of strain-enduring inks allows the high-throughput cost-effective fabrication of the textile-based hybrid device. Such successful integration of textile-based SCs and BFCs promotes the development of self-sustained on-body energy and represents a vital step towards self-powered wearable electronics for healthcare, fitness, security, and environmental monitoring applications.


Wearable electronics are in the forefront of the emerging fields of sensors, electronic displays, fitness monitoring, thermal controlling, and motion monitoring.1–6 Such devices commonly require energy over extended periods of time. Despite the accelerated growth of wearable devices, most of these devices rely on batteries to provide power. The replacement of the rigid and bulky batteries in wearable energy-harvesting and energy-storage devices has received tremendous attention for a broad range of wearable applications.7,8 Flexible and stretchable energy-storage batteries and supercapacitors have thus been developed to provide stable performance under different levels of mechanical deformations.8–11 However, their mass adaptation has been limited by their fabrication process, low energy densities and need for frequent recharging, which is inconvenient to users.12 To address this challenge, the integration of wearable energy-harvesting with energy-storage devices represents an attractive route that opens up a new possibility for empowering wearable electronics sustainably.13 In addition to harvesting energy from sunlight, body motion, or the combination of such,14–19 scavenging sustainable alternative energy from human biofluids has stimulated a considerable interest.20–22 Since the majority of wearable devices are adapted to the human skin, epidermal biofuel cells (BFCs) have received particular attention. The widespread adaptation of wearable electronics and the development of self-powered systems await the development of innovative energy-harvesting devices and their direct integration with energy-storage devices for creating self-powered systems. Realizing such high performance of wearable devices, capable of undertaking multiple complex tasks and activating various device components (e.g., sensors, microcontrollers, or communication systems), thus requires the judicious integration of powerful energy sources onto conformal platforms.

Herein, we present the first example of an energy textile-based system that practically integrates the scavenging of biochemical energy from the wearer's sweat using a BFC with the storage of the harvested energy using a supercapacitor (SC). The successful integration of stretchable textile-based energy-harvesting and energy-storage BFC and SC devices requires the fabrication of multiple, complex, different functional components on a single piece of soft textile, without compromising the performance of the individual devices or the textile flexibility. Hybrid energy systems, integrating energy-harvesting and energy-storage modules in a single wearable platform, can enable the conversion of bioenergy into electricity, providing a sustainable power source for electronic devices.23 Wearable BFCs represent one of the most attractive on-body energy harvesting devices, as they can generate electric energy directly from the reactions with biofluids (e.g., tears, blood, interstitial fluid, and sweat).24–28 Sweat-based epidermal enzymatic BFCs offer an efficient non-invasive energy-harvesting route, thus providing significant promise for powering wearable electronics.20 For example, epidermal screen-printed BFCs based on sweat fuel have been reported.29–31 Soft BFC arrays, combining bioanodes and cathode pellets with lithographically-patterned interconnections, offer high power density (1.2 mW cm−2) along with high stretchability.12 Such new capabilities greatly enhance the prospects of using sweat-based BFCs as on-body energy harvesters. However, epidermal BFCs can generate and deliver power only in the presence of sweat. Therefore, variable levels of sweat and lactate biofuel limit a constant power output due to the irregular user perspiration. Moreover, the limited voltage BFC output requires extra electronic components such as DC/DC converters, limiting the flexibility of the entire device.32 An effective storage method for the produced biopower is thus of considerable importance to real-life applications of sweat-based BFCs.

To address this issue, efforts have been made toward the development of conformal energy-storage devices. SCs have been considered as promising alternatives to rechargeable batteries due to their safety, high-power density, fast charge–discharge rates, and long cycling life.33 The integration of SCs with textiles has been encouraging, especially for in-plane SCs, possessing favorable flexibility for shape designs.34–39 Textiles represent one of the most important and useful platforms for wearable electronics with the most human-contact and body-conformity. Lee et al. reported non-stretchable screen-printed textile-based SCs with high areal capacitance density and flexibility.40 To the best of our knowledge, there are no early reports on using stretchable in-plane SCs to address the stable and durable output limitation of the wearable BFC and on the coupling of such stretchable energy-storage devices with BFC energy-harvesting devices onto the textile platform.

The all-printable stretchable hybrid energy textile device presented in this study integrates sweat-energy harvesting and supercapacitor storage (Fig. 1). Our all-in-one energy harvesting/storage devices are capable of generating and storing the bioenergy produced by human activity. Using this dual-functional hybrid device, the sweat-based BFC scavenges energy from the lactate fuel that can then be stored and released by the integrated SC toward practical wearable applications. The textile-based hybrid energy system must be flexible and stretchable to easily conform to the body contours and withstand repeated mechanical stress. The major challenges faced by integrating two stretchable devices on two sides of the textile is that the printing of multiple layers on a single piece of the soft textile can lead to decreased textile softness and deteriorated printing quality. The fabrication of the flexible wearable dual-functional energy textile system relies on the low-cost and high-throughput screen-printing of stress enduring inks. Customized SC and BFC inks have thus been used for printing the corresponding electrodes and silver ink interconnections onto opposite sides of a stretchable textile substrate (Fig. 1e). The printed SC, based on a MnO2/carbon nanotube (CNT) ink, features outstanding areal capacitance and electrochemical cycling and mechanical stability. The energy harvested by sweat-based BFCs could thus be rapidly stored in wearable in-plane SC and used for wearable electronics, providing a constant output. The system, adapting to physiological sweat lactate concentrations, can deliver stable output over long charging periods, boost the voltage output of the BFC, and exhibit favorable cycling ability. Upon mounting the printed hybrid system directly on the arm of the human subject, the SC can be charged to 0.4 V by the BFC and remains stable. The attractive performance and unique architecture of the new hybrid textile-based energy harvesting/storage device are illustrated in the following sections. Such successful integration of textile-based SC and BFCs promotes the development of self-sustained on-body energy systems for wearable electronics toward self-powered wearable electronics.

image file: c8ee02792g-f1.tif
Fig. 1 Schematic illustration of the wearable BFC–SC energy harvesting-storage hybrid device: (a–c) Schematic illustration of the energy generation by the BFC from the lactate in sweat to charge the SC, along with the corresponding anode and cathode reactions. (d) Image of the designed stencil for printing the hybrid energy harvesting-storage system. Photograph of the stretchable hybrid BFC–SC device (e) printed on a wearable wristband (outside: SC (f) and inside: BFC (g)).

Results and discussion

Fig. 1a shows the structure of the wearable dual-functional energy harvesting/storage system fabricated on the stretchable garment. The two electrochemical energy modules, the lactate BFC and the in-plane SC, were integrated in the opposite sides of a piece of stretchable fabric. The textile-based hybrid energy system thus harvests biochemical energy from sweat using an enzymatic BFC and stores the generated energy in an integrated stretchable supercapacitor. The MnO2–CNT layer served as the active component, with a layer of stretchable silver as the current collector, for the in-plane SC (Fig. 1b). A neutral LiCl-PVA gel was selected as the electrolyte. The membrane-less BFC, comprised of the bioanode and cathode, was constructed to be in contact with the skin and harvest energy from sweat, and pass this bioenergy to the integrated energy-storage SC (Fig. 1c). The bioanode was modified with the lactate oxidase (LOx) enzyme and naphthoquinone (NQ) redox mediator in order to catalyze the oxidation of lactate and enhance the power density, respectively. Silver oxide was used as the active cathode material. CNTs were also used to enhance the electron transfer from the redox reactions at the anode and cathode. The lactate biofuel is enzymatically oxidized on the anode in the presence of sweat, liberating electrons that flow toward the silver-oxide cathode, and generating the electrical power.

A scalable low-cost screen-printing approach was used to fabricate the textile-based BFC and SC devices, relying on stretchable engineered inks, along with the custom-designed serpentine-shaped SC (f) and BFC (g) electrode patterns (Fig. 1d–g). The soft styrene ethylene butylene styrene block copolymer (SEBS) elastomer was first coated onto the two sides of the textile to smoothen the fabric surface. The thickness of the SEBS substrate is 100–200 μm, thus ensuring the softness of the textile, as shown in Fig. S1 (ESI). For the screen-printed devices, the stable adhesion between two layers was realized by carefully selecting the top layer solvent, in which the bottom layer polymer needs to be soluble to ensure bonding between the two layers. The SC inks can be printed directly on top of the SEBS layer, due to solvents in the inks. For BFC printing, a polyurethane (PU) interlayer is added to enhance the adhesion between the CNT electrodes and SEBS layer, because of the volatile tetrahydrofuran used in the stretchable CNT ink that cannot sufficiently dissolve the SEBS layer to form the strong bonding between the two layers. The judicious design ensures that both the SC and BFC are firmly attached to the textile, and the mechanical adhesion properties of the printed traces were assessed by the scotch tape test as shown in Fig. S2 (ESI). The unique architecture of the double-sided printed energy modules on the stretchable fabrics, along with the use of stress-enduring printing inks, provided the resiliency against mechanical strains, such as bending, twisting, and stretching, while maintaining an attractive energy harvesting and storage performance.31


MnO2–CNTs, COOH–CNTs, poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS), and PU were used to formulate the functional ink, endowing attractive electrochemical performance and strain resiliency to the printed active electrodes, as shown in Fig. 2a. In our design, the three-dimensional (3D) MnO2–CNTs, having a high active surface area, are prepared by growing the MnO2 nanoplates on CNTs. However, functionalizing CNTs with MnO2 inevitably lowers the CNT conductivity. Wrapping COOH–CNTs and PEDOT–PSS around MnO2–CNTs three-dimensionally not only lowers the resistance by providing additional electron transport paths but also actively contributes to the energy storage of the whole electrode through electric double-layer capacitance or pseudocapacitance. The interaction between MnO2–CNTs and PEDOT:PSS was supported via π–π stacking. For the printed stretchable supercapacitor electrodes, both CNTs and PEDOT:PSS are needed. CNTs contribute primarily to the high conductivity, while PEDOT:PSS can increase the mechanical stability of the electrodes without compromising the energy storage performance. MnO2 was considered as one of the promising transition metal-oxide electrode materials for pseudocapacitors due to its low cost, abundance, high theoretical specific capacitance, and environmental friendliness.41 However, low conductivity and crystal expansion/contraction during the cycling result in the relatively low specific capacitance and cycling stability, limiting the practical workability of MnO2. Directly growing MnO2 on the highly conductive CNT surface is an effective way to address these drawbacks.42,43 The resulting MnO2–CNT surface, synthesized through a facile one-pot wet chemistry method, was composed of a 50 nm thin MnO2 layer on the CNT surface (Fig. S4, ESI).43 The composition of the resulting MnO2–CNT composites was characterized by XRD, as shown in Fig. S5 (ESI). Compared with the pristine CNTs, four peaks of the MnO2–CNTs are observed at 12, 24, 37, and 66°, in agreement with the (001), (002), (111), and (312) crystal planes of birnessite-type MnO2 (JCPDS 421317), respectively.44 Carboxylic acid-functionalized CNTs (COOH–CNTs) were used to improve the electrical conductivity of the ink, while PEDOT:PSS enhanced the softness and capacitance. The capacitance of the SCs for electrodes with pure COOH–CNTs, COOH–CNTs/MnO2–CNTs, and COOH–CNTs/MnO2–CNTs/PEDOT:PSS was investigated using cyclic voltammetry (CV). The results shown in Fig. S6 (ESI) clearly indicate that the electrodes with MnO2–CNT exhibit much higher areal capacitance than CNT electrodes, suggesting the high pseudocapacitance of MnO2–CNT. The PEDOT:PSS further improved the areal capacitance by two-fold. This added areal capacitance, while maintaining the capacitive behavior of the SC, can be related to the high conductivity as well as the capacitive properties of the PEDOT:PSS material.45 The mechanical resiliency of the electrodes reflects the well-dispersed CNTs in the elastomer PU matrix due to hydrogen bonds from PU and COOH-functionalized CNTs.30 The SEM image of the printed MnO2–CNT electrode material shows the good dispersion of the composites. As shown in Fig. 2b and Fig. S7 (ESI), COOH–CNTs and MnO2/CNTs were dispersed uniformly within the printed electrode, facilitating the effective electron transfer between the active components in the electrode and the current collector. In addition, the porous structure of the electrode provides highly active sites for contact with ions in the electrolyte.46 The cross-sectional SEM image of the SC electrode printed on a textile/SEBS, shown in Fig. 2c, illustrates the good bonding among the different layers.
image file: c8ee02792g-f2.tif
Fig. 2 (a) Illustration of the compositions of the MnO2–CNT/CNT/PEDOT:PSS ink. (b) SEM images of the printed MnO2–CNT/CNT/PEDOT:PSS electrode. (c) SEM image of the cross-section of the SC electrodes printed on the textile. Electrochemical performance of the wearable SC: (d) CV curves of the SC at different scan rates (2, 5, 10, 20 and 50 mV s−1). (e) Galvanostatic charge–discharge curves of the wearable SC under different areal current densities (0.1, 0.25, 0.5, 1 mA cm−2). (f) The areal capacitance of SC at different charge–discharge current densities. (g) CV curves of SCs at a scan rate of 10 mV s−1 connected in series (black plot: in series, red plot: single SC). (h) The stability of SC for 500 CV cycles and (i) the changes in the areal capacitance based on (h). Images of the wearable SC under conditions of (j) bending, (k) twisting and (l) stretching ((1) before stretching, (2) after stretching). CVs of SC tested (black dashed lines) before and (red solid lines) after (m) bending 100 times, (n) 180° twisting for 100 repetitions, and (o) 20% stretching for 100 times. The corresponding cross-sectional SEM images of the electrode before and after 100 cycles of 20% stretching are shown in Fig. S9 (ESI).

CV and galvanostatic charge–discharge (GCD) techniques were used to characterize the electrochemical properties of the SC. Fig. 2d shows the CV curves of the SC at scan rates of 2–50 mV s−1. It is clear that all the CV curves were almost rectangular in shape without significant polarization, suggesting an ideal supercapacitor behavior. The symmetric triangular shape of the GCD curves (Fig. 2e), obtained at different current densities also confirmed the excellent capacitive behavior of the SCs.40 Fig. 2f concludes the areal capacitance of the SC as a function of the applied current density, showing a decrease in capacitance by increasing the current density from 0.1 to 1 mA cm−2. A retention rate beyond 65% was achieved, indicating the reasonable rate capability of the MnO2–CNT/CNT/PEDOT:PSS printed electrode material with high electrical conductivity. Such behavior can be attributed to the short ion diffusion path and increased electronic conductivity. The calculated highest areal energy and power density of the SC are 17.5 μW h cm−2 and 0.4 mW cm−2, respectively, which are comparable with the reported literature, as shown in Fig. S8 (ESI).

Fig. 2g demonstrates the capacitive behavior of two SCs connected in series (b), showing the ability to tune the operating voltage of the in-plane SC. As clearly shown in Fig. 2g, SCs in series can withstand up to 2 V of applied potential while maintaining the same capacitive behavior as expected from carbon-based and MnO2 electrodes, similar to the individual SC. The electrochemical longevity of the printed SC was tested by comparing multiple CV experiments, cycling up to 500 cycles at a scan rate of 10 mV s−1. Results in Fig. 2h and i show a slight increase in capacitance over time from the first cycle to the 500th cycle at a scan rate of 10 mV s−1, which is attributed to the enhanced wetting of the electrode.47 Such increasing capacitive behavior confirms further the electrochemical robustness of the printed SC (i.e., both electrode and electrolyte materials). The mechanical resiliency of the SC was tested by implementing CV tests on SCs before and after bending, twisting, and stretching for 100 times. The results indicate a negligible (4.74%) capacitance loss after a 20% stretching conditions. Additionally, after 100 repeated bending and twisting deformations, the capacitance of the SC increased by a small amount. Mechanical deformations are generally expected to increase the resistance of the supercapacitor, and hence to decrease the capacitance of the SC. However, the bending and twisting affect the electrode conductivity to a lesser extent and are expected to increase the surface area of the electrodes, leading to 9.01% and 9.51% capacitance increments following repeated bending and twisting 100 times. Fig. 2j–o show the good mechanical robustness relevant to routine wearable applications. The cross-sectional SEM images of the supercapacitor electrodes after 100 cycles of 20% stretching are shown in Fig. S9 (ESI), showing adhesion stability between the print and the substrate. Such resiliency to repeated bending and twisting is attributed to the inherent mechanical properties of the printable composite materials that are reinforced by the PU binder. In addition to such intrinsic mechanical stretchability of the printed electrodes (associated with the stress-enduring inks), an additional level of stretchability can be attributed to the serpentine design pattern of the electrodes, as discussed in our previous study.31

Biofuel cells

The electrochemical performance of the textile-based lactate BFC was examined by assembling the bioanode and cathode without a separating membrane. Fig. 3a shows the performance of the resulting fabric BFC in different levels of the lactate fuel. The plots of the power density as a function of BFC voltage display a characteristic bell-shaped curve, with the power density and open circuit voltage (OCV) increasing with the lactate concentration over the 0–15 mM range. The power density reached saturation at around 15 mM of lactate. The variation in the power density as a function of the cell voltage shows that the OCV of the BFC reached 0.49 V along with a maximum power density of 252 μW cm−2 at 0.28 V. Fig. 3b shows the polarization curves of the assembled BFC. The closed-circuit current density increased from 89 μA cm−2 (without lactate) to 910 μA cm−2 (with 10 mM lactate), reflecting the lactate oxidation and energy harvesting. It also indicates that in the presence of 10 mM lactate fuel, a maximum current output of 1.1 mA cm−2 can be generated at a cell operating voltage of 0.15 V, suggesting a favorable BFC energy harvesting performance. This performance is comparable with reported wearable textile-based or stretchable BFCs, as summarized in Table S2 (ESI). The efficient power density dependence on the cell voltage was monitored during a 10 hour period, showing a 12% decrease in the maximum power density to 195 μW cm−2 at 0.27 V (Fig. 3c).
image file: c8ee02792g-f3.tif
Fig. 3 The stretchable textile-based printed BFC energy harvester. (a) The power density versus voltage plots of the stretchable lactate BFC using different lactate concentrations (I–V: 0, 1.25, 2.5, 5, 10 and 20 mM) in 0.5 M PBS (pH 7.4). (b) Polarization curves of the lactate BFC in the absence (black) and presence of 10 mM lactate (red). (c) Long-term stability of stretchable BFC in the presence of 10 mM lactate over 10 hours. The inset depicts the change in power density at 0.28 V during the 10 hours. The performance of the stretchable BFC under mechanical deformations: (d), (e), and (f) photographs of the device showing bending, twisting, and 20% stretching, respectively. (g) and (h) Power outputs (black dash lines) before and (solid pink lines) after 100 repeated bending and twisting processes. (i) Power outputs (black dash lines) before and (solid pink lines) after 100 cycles with 20% stretching, respectively. The corresponding cross-sectional SEM images of the electrode before and after 100 cycles 20% stretching are shown in Fig. S9 (ESI).

To highlight the integration of the BFCs with wearable textiles, which needs resilience to severe mechanical distortions, the mechanical durability was evaluated. Fig. 3d–f shows photographs of the folded, twisted, and stretched textile-based BFC device. The stretchable serpentine pattern is expected to accommodate the external strain.31,48 The well-distributed CNTs within the PU matrix and quality screen printing, as shown in Fig. S10 (ESI), also contribute to the resilience to the mechanical deformations. After applying these various deformations, the BFC was found to be mechanically compliant and upheld good attachment to the textile. Furthermore, the BFC performance was examined electrochemically to observe the electrochemical functionality and mechanical resiliency of the active anode and cathode. First, a repeated straight-to-bending process was studied, as shown in Fig. 3d and g. After 100 such 180° folding cycles, the textile-based BFC maintained 98.7% of the initial BFC power. Furthermore, considering real scenarios for daily life, the wearable textile can wrinkle. Accordingly, repeated torsional twisting was applied to simulate the on-body situations (Fig. 3e and h). One hundred repeated 180° twisting led to a small decrease (4.6%) in its initial power. Other severe deformations can be expected for on-body applications. For example, a succession of stretching strains was applied repeatedly to emulate real on-body applications and the durability of the generated power performance was monitored. After numerous stretching cycles of 20% strain, the durability of the generated power performance was observed. As shown in Fig. 3f and i, no major variations in the power curve characteristics and power density were observed after 100 stretching cycles of 20% strain. The cross-sectional SEM images of the CNT electrodes after 100 cycles of 20% stretching are shown in Fig. S11 (ESI), indicating the good strain endurance of the printed electrodes. The power was maintained at over 86.4% after 100 such stretching processes. These results confirmed the robustness of the printed wearable textile-based device.

Textile-based BFC–SC integration

The conformability of both printed SC and BFC enables their integration to create a garment-based self-powered system, in which the SC is charged by energy generated from the sweat-based BFC. The notable mechanical resilience of the textile-based devices toward extreme external strains is demonstrated in Video S1 (ESI). A variety of strains was applied, including linear and biaxial stretching, twisting, and folding. These studies show that the textile-based printed devices can withstand severe and repeated multiaxial strains pertinent to the user's movement, allowing conformable wearability in real-life circumstances.

Fig. 4a illustrates the charging of the SC for 10 min by the BFC in the presence of increasing lactate concentrations. The potential outputs of the SC, charged by the sweat-based BFC, increased upon raising the lactate fuel concentration. Without lactate, the SC could only be charged to 0.08 V, stemming from the small potential difference between the bioanode and Ag2O cathode. Using the 1.25 mM lactate fuel, 0.28 V could be obtained from the SC after charging for 10 minutes. An SC potential of 0.46 V was observed upon increasing the lactate concentration to 10 mM. No augmented output of the SC could be accessed when higher lactate-fuel concentration (20 mM) was used due to the BFC saturation expected by the enzyme kinetics. Interestingly, the integrated textile system showed stable output using a wide range of lactate concentrations, from 10 mM to 20 mM. Such behavior holds promise for a variety of realistic scenarios, considering that the normal concentration of lactate in human sweat is around 14 mM and it varies in different physiological states.49 The discharging behavior of the SC showed an increase in the discharging time upon increasing the lactate concentration (Fig. 4b). A discharging current of 50 μA was used for all experiments. At 10 mM lactate, the sweat-based BFC charged SC could be discharged in the long run (∼1000 s), confirming the potential of this self-powered electrochemical energy harvesting and storage hybrid system. The energy and power harvested by the integrated system in the presence of different lactate concentrations are shown in Fig. S12 (ESI). The maximum energy and power produced by the integrated SC–BFC device after a single charge is 11.1 mJ and 11.1 μW, respectively.

image file: c8ee02792g-f4.tif
Fig. 4 (a) Charging curves of the SC, charged by the stretchable BFC, for 600 s using different lactate concentrations (I–VI: 0, 1.25, 2.5, 5, 10, and 20 mM). (b) The discharging curves of SC at a current of 50 μA after charging by the BFC in the presence of various amounts of lactate (I–VI: 0, 1.25, 2.5, 5, 10, and 20 mM). (c) The long-term charging curve of SC charged by the stretchable BFC in the presence of 10 mM lactate. (d) and (e) charging curves of SC charged by stretchable BFC connected in parallel and in series, respectively, (blue plot: single BFC, red plot: two BFCs). (f) Charging curves of the SC charged six times by one reusable BFC. The inset shows the reusability of the integrated SC–BFC device.

To demonstrate the stability of the charging process, the voltage changes in the SC under constant charging of one BFC using 10 mM lactate fuel, over a prolonged (>2 hours) period is shown in Fig. 4c. The increase in the SC potential was rapid during the first 5 minutes, reaching over 90% of the stable potential, reflecting the large voltage difference when the SC was initially connected to the BFCs. Such potential difference drove the transfer of a large number of electrons from the BFC to SC. With longer charging time, the potential difference between SC and BFC was lower, accompanied by a minimal change in the open circuit potential of SC. A stable SC open circuit potential of about 0.46 V lasted for more than two hours, illustrating the high stability of the integrated self-power energy system for long-time charging. The effect of 20% stretching on the charging process was also studied (Fig. S13, ESI). After 20% stretching, the OCV decreased slightly from 0.46 to 0.45 V and subsequently recovered to 0.46 V, indicating that the integrated device offers good stability in practical scenarios. The self-discharging behavior of the SC was monitored for 2 hours after it was fully charged by the BFC with 10 mM lactate fuel. As shown in Fig. S14 (ESI), a self-discharge rate of 0.075 V h−1 was observed.

Using the BFCs connected in parallel and in series to charge the single SC, the integrated charging process was further studied (Fig. 4d and e, respectively). When the BFCs were connected in parallel, they provided higher power output but the same range of OCV, compared to a single BFC. This phenomenon resulted in only a slightly higher rate increase of the SC OCV over the first 500 s, compared to the SC charging by a single BFC. Using two BFCs connected in series could charge the SC to a much higher OCV of 0.76 V following 10 minutes charging, indicating the considerable promise of the integrated system to reach a high voltage output upon changing the connecting circuit. In practical application, the reusability of the integrated self-power energy devices is very precious. To demonstrate that the integrated device could be used multiple times, the same BFC was applied to charge a single SC six times. During the sixth charging, the open circuit potential of the SC was maintained at 96.1% of the OCV observed in the first charging cycle, as shown in Fig. 4f, suggesting the favorable repeatability of the integrated device toward a self-powered energy system.

The feasibility of the self-powered energy SC–BFC system was demonstrated by using three in-series SCs charged by five BFCs connected in series to light three LEDs (Fig. 5a). The corresponding circuit diagrams are shown in Fig. S15 (ESI). Placing the 10 mM lactate fuel solution on the BFCs turned the LED ‘on’. After 30 minutes of charging, the SCs were disconnected from the BFC. Encouragingly, the SC could still light the LEDs for 3 minutes, illustrating the capability of the fabricated device to serve as the textile-based power source for wearable electronics even when the sweat has evaporated. To quantify the sweat required for the operation of the device in real scenarios, different volumes of the 10 mM lactate solution were added to the BFC covered with a plastic sheet, followed by examining the dependence of the performance outputs upon the biofuel volume (Fig. S16, ESI). The data indicate that the minimum sweat volume per area of ∼40 μL cm−2 can provide stable outputs. An on-body test was also performed to demonstrate the operation of the integrated device in realistic conditions. A volunteer wearing a sweatband, printed with an SC on the outside and a BFC on the inside, maintained the rigor of a stationary cycling exercise at a constant pace (Fig. 5b). The SC was connected to the BFC by conductive yarn with the stretchable Ag ink as the conductive adhesive. Real-time OCV of the SC was recorded using a potentiostat (Fig. 5c). For comparison, the OCV change in another SC charged by LOx-free BFC was also recorded as a control experiment (green curve in Fig. 5c). Compared with the control, the voltage output of the SC started to increase when sweat was secreted on the subject's skin, due to the charging of the self-generated BFC. After 37 minutes of perspiring, the SC could be powered to 0.40 V, and then kept stable, suggesting that such a wearable integrated self-power energy system holds considerable promise to provide stable energy for wearable electronics.

image file: c8ee02792g-f5.tif
Fig. 5 (a) Demonstration of the hybrid SC–BFC device. (1 and 2) Photographs showing the application of three SCs charged by five BFCs to light LEDs using the following procedure: (3) without lactate, no power; (4) with lactate, LEDs were turned on; (5) upon disconnecting BFCs and SCs, LEDs could still be turned on. (b) The integrated chemical self-powered system on one piece of textile was applied to the arm of a volunteer. The SC and BFC were printed outside and inside the textile band, respectively. (c) The real-time voltage of the printed SC charged from the on-body BFC during a constant cycling exercise for 56 minutes. The SC charged by lactate BFC immobilized with LOx (blue plot) and without LOx as a control (black plot).


The compliant integrated energy device described in this study is capable of harvesting and storing the energy generated by human activity. We have demonstrated the first example of a wearable hybrid SC–BFC self-powered energy system, screen printed on the stretchable fabric, in which the SC was used to store the bioenergy harvested by the sweat-based BFC. Our main objective was to demonstrate the successful fabrication of the stretchable BFCs and SCs on both sides of a single textile substrate (rather than optimizing the separated SC or BFC devices) and the seamless integration of such integrated BFC–SC textile devices to the human body. For the in-plane SCs, MnO2/CNTs, COOH–CNTs, conducting polymers, and PU binder were employed to formulate an engineered stress-enduring ink for the serpentine-shaped electrode. A layer of stretchable Ag ink was printed as the current collector. The MnO2/CNT composites endowed the printed SC with high areal capacitance and cycling stability, while the elastomeric binder allowed the electrode to be compliant. The textile-based BFC was used to harvest the energy from sweat. It provided a high power density of 252 μW cm−2, long-term stability, and resiliency against severe strains. The proof-of-concept charging process was demonstrated on a human subject by charging the on-textile SCs using the energy harvested from the sweat. The SC–BFC integration allows stabilized output, serving as a stable textile-based power source for wearable electronics even when the sweat amount fluctuates. Combined with the capability for operation using a wide range of sweat lactate concentrations, the flexible device delivers stable output for long charging periods, maintains good rechargeability, and can be fabricated at a low cost; the new SC–BFC hybrid system therefore has considerable potential for practical real-life applications. Further efforts should be made toward improving the performance of the SC and BFC. The stretchability of the devices can also be increased by printing the free-standing electrodes. Washability is another important concern for textile-based devices and could be addressed by enhancing the immobilization and surface adhesion of the BFC components and improving the encapsulation of the SC. The vulnerability of the enzymatic BFC toward the environmental impacts should also be investigated. Such a high level of device integration and creation of multi-layer stretchable textile energy systems thus opens up future opportunities. The present hybrid system could successfully generate and store the energy in practical exercise scenarios, representing a vital step forward in the development of wearable self-powered textile-based electronics.

Experimental section

Chemicals and reagents

Carboxylic acid-functionalized multi-walled carbon nanotubes (COOH–CNTs) and hydroxyl-functionalized multi-walled carbon nanotubes (OH–CNTs) (purity >95%, diameter = 10–20 nm, length = 10–30 μm) were purchased from Cheap Tubes Inc. Multiwall carbon nanotubes were bought from Nanocyl S. A. Potassium permanganate (KMnO4), sulfuric acid (H2SO4) (98%), polystyrene-block-polyisoprene-block-polystyrene (SIS), polyvinyl alcohol (PVA) (MW 146[thin space (1/6-em)]000–186[thin space (1/6-em)]000), toluene, chitosan, 1,4-naphthoquinone (NQ), bovine serum albumin (BSA), glutaraldehyde, Nafion (20%), L(+)-lactic acid, potassium phosphate dibasic (K2HPO4), potassium phosphate monobasic (KH2PO4), ethanol, and acetone were purchased from Sigma-Aldrich. Terpineol and tetrahydrofuran (THF) were purchased from Alfa Aesar and EMD Millipore, respectively. Polyurethane (PU) (Tecoflex SG-80A), styrene ethylene butylene styrene copolymer (SEBS), and Ecoflex® 00-30 were obtained from Lubrizol LifeSciences, Kraton and Smooth-On, Inc. PA., respectively. Lithium chloride (LiCl), silver oxide (Ag2O) and LOx were purchased from Fisher Scientific, Acros and Toyobo, respectively. All chemicals were used without further purification. Water-based polyurethane resin DispercollU-42 and Bayhydur 302 crosslinker were purchased from Covestro. PEDOT:PSS (translucent clear conductor C2100629D1) was from purchased from Gwent. Ultra-pure deionized water (18.2 MΩ cm) and was used for all of the aqueous solutions. Ecoflex® 00-30 was prepared by mixing equal volumes of pre-polymers A and B, provided by the supplier. Stretchable textile Lycra Shiny Milliskin Nylon Spandex Fabric was purchased from Spandex world. Inc, USA. Tegaderm film was purchased from 3M.

Formulation of the stretchable Ag ink

The method for making the stretchable Ag ink is similar to that reported in our previous work.50 Briefly, a resin consisting of 4 g mL−1 SIS in toluene was prepared, then 0.7 g of the SIS resin was mixed with 1.4 g of silver flakes (Aldrich, 10 μm) in a dual asymmetric centrifugal mixer (Flacktek Speedmixer, DAC 150.1 KV-K) under a speed of 1800 rpm for 5 minutes. Zirconia (YSZ) beads (Inframat Advanced materials, 3 mm) were also added for enhanced mixing.

Formulation of the SEBS and PU interlayer substrate resins

Styrene ethylene butylene styrene polymer (SEBS) beads were dissolved in toluene with a weight ratio of 2[thin space (1/6-em)]:[thin space (1/6-em)]5 to obtain the substrate resin with the desired viscosity, which was screen-printed onto the textile substrate. A water-based polyurethane resin was used to enhance the adhesion between the BFC and SEBS layer. Next, 2 g of the PU resin was mixed with 100 mg of crosslinker to endow the suitable viscosity for screen printing.

Synthesis of the MnO2–CNT composite materials

The synthetic route of the MnO2/CNT nanocomposites follows the reported work with a minor modification.43 First, 200 mg of COOH–CNTs and 2.5 g of KMnO4 were mixed and ground in an agate mortar for 10 minutes. The mixture was dispersed in 100 mL of water and then stirred for 10 minutes; 0.5 mL of concentrated H2SO4 (98%) was then added and stirred for an additional 30 minutes. The mixture was allowed to react in an 80 °C water bath for one hour with continuous stirring. Finally, the mixture was cooled down to room temperature and the product was collected by centrifugation at 3000 rpm for 5 minutes and washed with water and ethanol. The washed sample was dried in the oven at 60 °C for 12 hours then prepared for ink formulation.

Formulation of the stretchable MnO2–CNT/CNT/PEDOT:PSS ink

MnO2/CNTs, COOH–CNTs, PEDOT:PSS translucent conductor and PU were used to make the ink for the SC electrodes. Briefly, 100 mg of MnO2/CNTs were mixed thoroughly with 100 mg of COOH–CNTs by grinding in the mortar. The ground mixture, together with 100 mg of PEDOT:PSS conductor, was dispersed in 0.6 mL of terpineol and 5 mL of THF by sonication for one hour. Terpineol was used to disperse the CNTs and stabilize the viscosity of the ink.51 After the components were fully dispersed in the solvent, 100 mg of PU beads were added, and the resulting mixture was shaken for 12 hours.

Formulation of the stretchable CNT ink

The procedure for making the stretchable CNT ink is based on our previous work with modification.30 COOH–CNTs (100 mg) and 30 mg of mineral oil were dispersed in a solvent containing 0.3 mL of terpineol and 5 mL THF by sonication for one hour. Then, 100 mg of PU was added, and the mixture was shaken for 12 hours. Before printing, the solvent in the ink was allowed to partially evaporate until the suitable viscosity for screen-printing was obtained.

Screen-printing of the devices

An MPM-SPM semi-automatic screen printer (Speedline Technologies, Franklin, MA) was used to perform the screen printing. The stainless-steel stencils (Metal Etch Services, San Macros, CA) with various thicknesses were designed using AutoCAD and were laser-cut to obtain the required patterns. The printing steps for SCs are as follows: a SEBS layer was first printed on the surface of the stretchable substrate (Lycra Shiny Milliskin Nylon Spandex Fabric, Spandex world. Inc) using a 100 μm-thick stencil and cured at 65 °C for 15 minutes. Then, a stretchable Ag layer using Ag-SIS ink was printed using the 100 μm stencil as the current collector and cured in the oven at 60 °C for 20 minutes. Finally, the MnO2–CNT/CNT/PEDOT:PSS ink was used to print the active layer using a 175 μm-thick stencil and cured first at room temperature for 3 hours and then at 80 °C for 25 minutes. The printing of BFCs was conducted as follows: the substrate with SEBS layer was first prepared using the same procedure. One interlayer using water-based PU resin was printed on top of the SEBS layer and cured at 50 °C for 30 minutes. The interdigitated CNT electrodes were then printed on the interlayer using the stretchable CNT ink and cured at 80 °C for 20 minutes. Finally, a stretchable Ag ink was printed to connect the anode and cathode electrodes separately and cured at 60 °C for 20 minutes. The weights of the active materials printed for the SC and the BFC in 1 cm2 were ∼7.8 mg and 4.3 mg, respectively. The thicknesses of active materials in the cured SC and BFC electrode by SEM images were 38 μm and 95 μm, as shown in Fig. S9 and S11 (ESI), respectively. The parameters of printed SC and BFC are summarized in Table S1 (ESI).

Assembly of the supercapacitor

A gel electrolyte was prepared by heating the solution of 2.1 g of LiCl and 1 g of PVA in 10 mL of water at 90 °C for one hour with constant stirring. Ecoflex® was used to cover the exposed Ag layer and define the active area. The fabrication of the SC was completed by applying the aforementioned electrolyte to the surface of the MnO2–CNT/CNT/PEDOT:PSS layer. Then, the electrolyte was allowed to condition the SC electrodes for 24 hours. The printed active area was covered by a thin Tegaderm film, prior to performing electrochemical measurements. The packaged SC is shown in Fig. S3 (ESI).

Assembly of the stretchable BFC

The electrodes were firstly activated with saturated sodium carbonate solution at +1.2 V for 100 s. The modification of the enzymatic anode electrodes was done by consecutively drop-casting solutions onto the electrode surface. The sequence and the formulation of the solution are as follows. Step 1, 30 μL: 2 mg mL−1 OH–CNTs in 0.2 M NQ in 9[thin space (1/6-em)]:[thin space (1/6-em)]1 vol/vol ethanol/acetone solution; step 2, 36 μL: 20 mg mL−1 LOx in 10 mg mL−1 BSA; step 3, 30 μL: 1% glutaraldehyde solution; and step 4, 30 μL: 1 wt% chitosan in 0.1 M acetic acid. Each drop-casting step was performed after the previous casted solution had completely dried at room temperature. The chitosan layer was left to dry overnight. The surface of the electrodes was then rinsed with water. The modification of the cathode was done by drop casting 500 μL of 30 mg mL−1 of Ag2O/CNTs (3[thin space (1/6-em)]:[thin space (1/6-em)]1) in 1% Nafion solution onto the electrode surface. The device was left to dry overnight at room temperature.

Characterization of the materials and electrodes

The phase composition was characterized by X-ray diffraction (XRD) (Rigaku RU200B diffractometer). Scanning electron microscopy (Phillips XL30 ESEM) was used to obtain the morphological information for the synthesized materials and printed SC electrodes. The element composition of the SC electrode was analyzed using energy dispersive X-ray (EDX).

Electrochemical measurements

Electrochemical studies of the two-electrode SCs were performed using a CHI 660C potentiostat (Austin, TX) with LiCl-PVA gel as the electrolyte. Cyclic voltammetry (CV) and galvanostatic charge/discharge (GCC) techniques were used to investigate the performance of the SCs. Furthermore, the effect of mechanical deformation on the SCs’ performance was studied by implementing CVs on variously deformed samples (i.e., 180 degree bent, twisted, and 20% stretched). The geometrical area of the compartments, 1.224 cm2, was used to calculate the areal capacitance. The electrochemical properties of BFCs were analyzed using a μAutolab Type II potentiostat. Open circuit voltages (OCV) of the BFCs were measured while quantified lactate solution was added to the 0.5 M phosphate buffer solution (PBS) (pH 7.4). Polarization curves were measured by scanning the current of the BFCs between the OCV and 0 V with a scan rate of 5 mV s−1. The geometrical area of each compartment ∼0.9 cm2 was used to calculate the areal power density. The stability of the BFC was analyzed by obtaining the polarization curve of the BFC every two hours for up to 10 hours in the presence of 10 mM lactate at room temperature. The effects of deformation on the BFCs were tested by comparing the polarization curves when the same mechanical deformations as above were applied to the BFCs.

Characterization of the integrated device

The OCVs and discharging tests were conducted on the μAutolab Type II potentiostat. The on-body test was performed by attaching the textile-based device to the arm, with the BFC facing down to the skin and the SCs facing outside. The conductive yarn was used to connect the SC and BFC and the stretchable Ag ink was applied to the junctions to ensure connections, and the junctions were insulated using Ecoflex®. All on-body experiments were performed with informed consent from the human subject volunteers, and in strict compliance with the guidelines of Institutional Review Boards (IRB) and were approved by Human Research Protections Program at University of California, San Diego.

Conflicts of interest

There are no conflicts to declare.


This work is supported by the UCSD Center for Wearable Sensors and Honda Inc. J. L., I. J., C. S. L., and J. H. Jang acknowledge support from the China Scholarship Council (CSC), Thai Development and Promotion of Science and Technology Talents Project (DPST), UC MEXUS-CONACY, and Korean Government Research Fund (NRF-2015-RIA2A01005250), respectively. J. L., C. Kong and Z. Yang acknowledge the support from the World-Class Universities (Disciplines) and the Characteristic Development Guidance Funds for the Central Universities.


  1. A. J. Bandodkar, I. Jeerapan and J. Wang, ACS Sens., 2016, 1, 464–482 CrossRef CAS.
  2. W. Gao, S. Emaminejad, H. Y. Y. Nyein, S. Challa, K. Chen, A. Peck, H. M. Fahad, H. Ota, H. Shiraki, D. Kiriya, D.-H. Lien, G. A. Brooks, R. W. Davis and A. Javey, Nature, 2016, 529, 509–514 CrossRef CAS PubMed.
  3. S. Gong, W. Schwalb, Y. Wang, Y. Chen, Y. Tang, J. Si, B. Shirinzadeh and W. Cheng, Nat. Commun., 2014, 5, 3132 CrossRef PubMed.
  4. S. Choi, J. Park, W. Hyun, J. Kim, J. Kim, Y. B. Lee, C. Song, H. J. Hwang, J. H. Kim, T. Hyeon and D.-H. Kim, ACS Nano, 2015, 9, 6626–6633 CrossRef CAS PubMed.
  5. G. Cai, J. Wang, K. Qian, J. Chen, S. Li and P. S. Lee, Adv. Sci., 2017, 4, 1600190 CrossRef PubMed.
  6. L. Lipani, B. G. R. Dupont, F. Doungmene, F. Marken, R. M. Tyrrell, R. H. Guy and A. Ilie, Nat. Nanotechnol., 2018, 13, 504–511 CrossRef CAS PubMed.
  7. W. Liu, M. S. Song, B. Kong and Y. Cui, Adv. Mater., 2017, 29, 1603436 CrossRef PubMed.
  8. R. Kumar, J. Shin, L. Yin, J. M. You, Y. S. Meng and J. Wang, Adv. Energy Mater., 2017, 7, 1602096 CrossRef.
  9. Y. Zhang, Y. Zhao, J. Ren, W. Weng and H. Peng, Adv. Mater., 2016, 28, 4524–4531 CrossRef CAS PubMed.
  10. D. P. Dubal, N. R. Chodankar, D.-H. Kim and P. Gomez-Romero, Chem. Soc. Rev., 2018, 47, 2065–2129 RSC.
  11. S. Gong, Y. Zhao, Q. Shi, Y. Ang, L. Weiy and W. Cheng, Electroanalysis, 2016, 28, 1298–1304 CrossRef CAS.
  12. A. J. Bandodkar, J.-M. You, N.-H. Kim, Y. Gu, R. Kumar, A. M. V. Mohan, J. Kurniawan, S. Imani, T. Nakagawa, B. Parish, M. Parthasarathy, P. P. Mercier, S. Xu and J. Wang, Energy Environ. Sci., 2017, 10, 1581–1589 RSC.
  13. T. Chen, L. Qiu, Z. Yang, Z. Cai, J. Ren, H. Li, H. Lin, X. Sun and H. Peng, Angew. Chem., Int. Ed., 2012, 51, 11977–11980 CrossRef CAS PubMed.
  14. Z. Gao, C. Bumgardner, N. Song, Y. Zhang, J. Li and X. Li, Nat. Commun., 2016, 7, 11586 CrossRef CAS PubMed.
  15. J. Xu, Y. Chen and L. Dai, Nat. Commun., 2015, 6, 8103 CrossRef CAS PubMed.
  16. Y.-H. Lee, J.-S. Kim, J. Noh, I. Lee, H. J. Kim, S. Choi, J. Seo, S. Jeon, T.-S. Kim, J.-Y. Lee and J. W. Choi, Nano Lett., 2013, 13, 5753–5761 CrossRef CAS PubMed.
  17. X. Pu, L. Li, M. Liu, C. Jiang, C. Du, Z. Zhao, W. Hu and Z. L. Wang, Adv. Mater., 2016, 28, 98–105 CrossRef CAS PubMed.
  18. X. Pu, L. Li, H. Song, C. Du, Z. Zhao, C. Jiang, G. Cao, W. Hu and Z. L. Wang, Adv. Mater., 2015, 27, 2472–2478 CrossRef CAS PubMed.
  19. Z. Wen, M.-H. Yeh, H. Guo, J. Wang, Y. Zi, W. Xu, J. Deng, L. Zhu, X. Wang, C. Hu, L. Zhu, X. Sun and Z. L. Wang, Sci. Adv., 2016, 2, e1600097 Search PubMed.
  20. A. J. Bandodkar and J. Wang, Electroanalysis, 2016, 28, 1188–1200 CrossRef CAS.
  21. A. J. Gross, M. Holzinger and S. Cosnier, Energy Environ. Sci., 2018, 11, 1670–1687 RSC.
  22. C. Agnès, M. Holzinger, A. Le Goff, B. Reuillard, K. Elouarzaki, S. Tingry and S. Cosnier, Energy Environ. Sci., 2014, 7, 1884–1888 RSC.
  23. J.-H. Lee, J. Kim, T. Y. Kim, M. S. Al Hossain, S.-W. Kim and J. H. Kim, J. Mater. Chem. A, 2016, 4, 7983–7999 RSC.
  24. M. Falk, V. Andoralov, M. Silow, M. D. Toscano and S. Shleev, Anal. Chem., 2013, 85, 6342–6348 CrossRef CAS PubMed.
  25. M. Cadet, S. Gounel, C. Stines-Chaumeil, X. Brilland, J. Rouhana, F. Louerat and N. Mano, Biosens. Bioelectron., 2016, 83, 60–67 CrossRef CAS PubMed.
  26. G. Valdés-Ramírez, Y.-C. Li, J. Kim, W. Jia, A. J. Bandodkar, R. Nuñez-Flores, P. R. Miller, S.-Y. Wu, R. Narayan, J. R. Windmiller, R. Polsky and J. Wang, Electrochem. Commun., 2014, 47, 58–62 CrossRef.
  27. H. du Toit, R. Rashidi, D. W. Ferdani, M. B. Delgado-Charro, C. M. Sangan and M. D. Lorenzo, Biosens. Bioelectron., 2016, 78, 411–417 CrossRef CAS PubMed.
  28. A. Koushanpour, M. Gamella and E. Katz, Electroanalysis, 2017, 29, 1602–1611 CrossRef CAS.
  29. W. Jia, G. Valdés-Ramírez, A. J. Bandodkar, J. R. Windmiller and J. Wang, Angew. Chem., Int. Ed., 2013, 52, 7233–7236 CrossRef CAS PubMed.
  30. A. J. Bandodkar, I. Jeerapan, J. M. You, R. Nuñez-Flores and J. Wang, Nano Lett., 2016, 16, 721–727 CrossRef CAS PubMed.
  31. I. Jeerapan, J. R. Sempionatto, A. Pavinatto, J.-M. You and J. Wang, J. Mater. Chem. A, 2016, 4, 18342–18353 RSC.
  32. W. Jia, X. Wang, S. Imani, A. J. Bandodkar, J. Ramírez, P. P. Mercier and J. Wang, J. Mater. Chem. A, 2014, 2, 18184–18189 RSC.
  33. P. Simon, Y. Gogotsi and B. Dunn, Science, 2014, 343, 1210–1211 CrossRef CAS PubMed.
  34. M. Parrilla, R. Cánovas, I. Jeerapan, F. J. Andrade and J. Wang, Adv. Healthcare Mater., 2016, 5, 996–1001 CrossRef CAS PubMed.
  35. A. M. Abdelkader, N. Karim, C. Vallés, S. Afroj, K. S. Novoselov and S. G. Yeates, 2D Mater., 2017, 4, 035016 CrossRef.
  36. H. Zhang, Y. Qiao and Z. Lu, ACS Appl. Mater. Interfaces, 2016, 8, 32317–32323 CrossRef CAS PubMed.
  37. K. Jost, D. Stenger, C. R. Perez, J. K. McDonough, K. Lian, Y. Gogotsi and G. Dion, Energy Environ. Sci., 2013, 6, 2698 RSC.
  38. M. Beidaghi and Y. Gogotsi, Energy Environ. Sci., 2014, 7, 867 RSC.
  39. X. Pu, M. Liu, L. Li, S. Han, X. Li, C. Jiang, C. Du, J. Luo, W. Hu and Z. L. Wang, Adv. Energy Mater., 2016, 6, 1601254 CrossRef.
  40. S.-S. Lee, K.-H. Choi, S.-H. Kim and S.-Y. Lee, Adv. Funct. Mater., 2018, 28, 1705571 CrossRef.
  41. K. Shimamoto, K. Tadanaga and M. Tatsumisago, Electrochim. Acta, 2013, 109, 651–655 CrossRef CAS.
  42. D. Gueon and J. H. Moon, ACS Sustainable Chem. Eng., 2017, 5, 2445–2453 CrossRef CAS.
  43. L. Li, Z. A. Hu, N. An, Y. Y. Yang, Z. M. Li and H. Y. Wu, J. Phys. Chem. C, 2014, 118, 22865–22872 CrossRef CAS.
  44. S.-B. Ma, K.-Y. Ahn, E.-S. Lee, K.-H. Oh and K.-B. Kim, Carbon N. Y., 2007, 45, 375–382 CrossRef CAS.
  45. Y. Liu, B. Weng, J. M. Razal, Q. Xu, C. Zhao, Y. Hou, S. Seyedin, R. Jalili, G. G. Wallace and J. Chen, Sci. Rep., 2015, 5, 17045 CrossRef CAS PubMed.
  46. R. Wannapob, M. Y. Vagin, I. Jeerapan and W. C. Mak, Langmuir, 2015, 31, 11904–11913 CrossRef CAS PubMed.
  47. M. F. El-Kady and R. B. Kaner, Nat. Commun., 2013, 4, 1475 CrossRef PubMed.
  48. D. S. Gray, J. Tien and C. S. Chen, Adv. Mater., 2004, 16, 393–397 CrossRef CAS.
  49. C. J. Harvey, R. F. LeBouf and A. B. Stefaniak, Toxicol. In Vitro, 2010, 24, 1790–1796 CrossRef CAS PubMed.
  50. L. Yin, R. Kumar, A. Karajic, L. Xie, J. You, D. Joshuia, C. S. Lopez, J. Miller and J. Wang, Adv. Mater. Technol., 2018, 3, 1800013 CrossRef.
  51. B. Kim, Y.-H. Lee, J.-H. Ryu and K.-D. Suh, Colloids Surf., A, 2006, 273, 161–164 CrossRef CAS.


Electronic supplementary information (ESI) available. See DOI: 10.1039/c8ee02792g
These authors contributed equally.

This journal is © The Royal Society of Chemistry 2018