PDF Quasifractal Networks as Current Collectors for Transparent ... - NJU

(1)

www.afm-journal.de

Full PaPer

Quasifractal Networks as Current Collectors for Transparent Flexible Supercapacitors

Shiqian Chen, Bibo Shi, Weidong He, Xiayan Wu, Xian Zhang, Yanbo Zhu, Shan He, Haihui Peng, Yue Jiang, Xingsen Gao, Zhen Fan, Guofu Zhou, Jun-Ming Liu,*

Krzysztof Kempa, and Jinwei Gao*

A high performance, transparent, and simultaneously mechanically flexible supercapacitor, with electrodes based on a bio-inspired, quasifractal metallic network is demonstrated. Excellent performance parameters, such as the specific capacitance of 5.6 mF cm⁻² (at 45% transparency), great long-term electrochemical and mechanical stability under stress, and low cost, make this one of the best reported, and most promising candidates for applications in the wearable, transparent integrated electronics.

DOI: 10.1002/adfm.201906618

S. Chen, B. Shi, W. He, X. Wu, X. Zhang, Y. Zhu, S. He, H. Peng, Dr. Y. Jiang, Prof. X. Gao, Prof. Z. Fan, Prof. J.-M. Liu,

Prof. K. Kempa, Prof. J. Gao

Institute for Advanced Materials and Guangdong Provincial Key Laboratory of Optical Information Materials and Technology Academy of Advanced Optoelectronics

South China Normal University Guangzhou 510006, China

E-mail: [email protected]; [email protected] Prof. G. Zhou

Guangdong Provincial Key Laboratory of Optical Information Materials and Technology & Institute of Electronic Paper Displays

South China Academy of Advanced Optoelectronics South China Normal University

Guangzhou 510006, China Prof. J.-M. Liu

Laboratory of Solid State Microstructures Nanjing University

Nanjing 210093, China Prof. K. Kempa Department of Physics Boston College

Chestnut Hill, MA 02467, USA

The ORCID identification number(s) for the author(s) of this article can be found under https://doi.org/10.1002/adfm.201906618.

conducting materials such as carbon nanotubes,^[5] graphene,^[6] and MXenes.^[7]

However, the unavoidable compro- mise between transparency and the specific capacitance for these carbon-based materials, leads to disappointing performance,^[7] and restricts their application.

Typically, the supercapacitor electrodes are precoated with ultrathin, insulating pseu- docapacitive materials, such as MnO₂,^[8]

polypyrrole (PPy),^[9] NiCo₂S₄.^[10] These are inexpensive, but their transparency per unit area is not satisfac- tory. Wang and co-workers^[4] tried to solve this problem with an island-like deposition of MnO₂ on an indium tin oxide (ITO) electrode, and obtained the specific capacity (areal capacitance) Csp = 4.73 mF cm⁻² at current density J = 50 µA cm⁻², with transmittance T = 44% at the light wavelength λ = 550 nm.

However, ITO in addition to being expensive, is also brittle and thus incompatible with flexible devices.^[11] As a result, flexible and transparent supercapacitors based on metallic networks have been proposed and studied,^[12] such as the Au mesh electrode,^[11,13] Ag nanowire (NW) electrode,^[14] and the metal grid electrode.^[15]

In an important development, Gao et al. have made a highly conducting and transparent microscaffold metallic network, by copying the structure of the leaf venation (LV).^[16]

This quasifractal electrode has been shown to outperform (in transparency and conduction) any regular or random metallic network.^[17] Here, we developed a flexible transparent supercapacitor, based on this bioinspired, quasifractal LV structure.

As expected, performance of this supercapacitor is excellent as demonstrated below.

2. Results and Discussion

Figure 1a shows a schematic of the procedure for making our LV network electrode. It consists of six steps: leaf venation extraction by chemical etching (step II), ultraviolet lithography (steps III and IV), Au metal deposition by sputtering (V), photoresist lift-off (VI), and electrodeposition of PPy (VII).

A photograph of the completed network electrode (called PPy/Au LV) is shown in the step VIII. For more information see Figures S1 and S2 in the Supporting Information.

Scanning electron microscopy (SEM) images shown in Figure 1b,c–g clarify morphologies of the uncoated and PPy- coated networks. The quasifractal structure of the LV network is

1. Introduction

With the increasing demand for portable and wearable electronics, such as smart sensors,^[1] electronic skin,^[2] and flexible health-monitoring devices,^[3] the energy storage components have been quickly transforming to emphasize light weight, integration, mechanical flexibility, and transparency. In that class, simultaneously flexible and transparent supercapacitors have been attracting attention as the new generation energy storage devices.^[4] One way to achieve high transparency, is to make the supercapacitor electrodes ultrathin, and use highly

(2)

www.afm-journal.de www.advancedsciencenews.com

clearly visible. The single Au line/ribbon has width ranging from 20 to 100 µm, and thickness of ≈200 nm (see also Figure S3 in the Supporting Information). The sheet resistance R_sh of the network is low (R_sh ≈ 1 Ω sq⁻¹), and its transmittance high (T ≈ 65%). Figure 1c–g shows images of the completed PPy/Au LV network electrodes, for different PPy deposition times, t_d, as characterized by the field emission (FE)-SEM . At t_d= 70 s, PPy clusters completely cover the Au ribbons. At an excessively long t_d= 90 s, PPy begins to pill-off from the Au ribbons. This can be seen already in Figure 1g, in the area marked by larger red circle. The average thickness of PPy at t_d= 30, 60, 70, 80, and 90 s is 2.4, 3.8, 4.4, 6.3, and 7.05 µm, respectively, and the corresponding transmittance of the electrodes (at 550 nm wavelength) is T = 61.1%, 60.4%, 59.87%, 57,64%, and 55.2%. For more information, including confirmation of the chemical com- position from Fourier transform infrared (FTIR) and energy- dispersive X-ray spectroscopy (EDX) spectra, see Figures S4–S8 in the Supporting Information.

To demonstrate the electrochemical performance of our completed PPy/Au LV network electrodes, cyclic voltammetry

(CV) was employed, in a three-electrode system utilizing 1 m

LiCl electrolyte. As shown in Figure 2a, the CV curves exhibit the desired, symmetrical quasirectangular shape, indicating excellent capacitive performance.^[18] In contrast, a solid Au electrode has negligible capacitance. With longer PPy deposition times capacity increases as expected, but saturates for t_d > 70 s.

For larger deposition times, an asymmetric distortion and appearance of oxidation peaks (for t_d > 90 s) indicate pres- ence of irreversible processes, occurring in the PPy films.^[9a]

The total loading mass of a PPy-70 s/Au LV network electrode is 0.3 mg.

The galvanostatic discharge (GD) curves at a current 0.1 mA were also measured for our LV network electrodes at various t_d (see Figure 2b). The corresponding plot of the areal capacitance (calculated from GD curves) versus t_d, is given in Figure 2c, and shows that the electrode for t_d= 70 s is the best. To assess the electronic conductivity and charge transport property, we used the electrochemical impedance spectroscopy (EIS). The resulting Nyquist impedance plots are shown in Figure 2d. Again, the electrode for t_d= 70 s has the largest slope, which confirms the Figure 1. a) Schematic of the PPy/Au LV network electrode fabrication process. b) SEM image of the Au LV network, before PPy coating. Inset shows enlarged area of the network, marked with small red circle. c–g) SEM images of the electrode at various PPy electrochemical deposition times t_d. Insets show enlarged areas, marked with small red circles on each figure. The white scale bars are 100 µm, and the black (in the insets) 1 µm.

(3)

optimal loading of this electrode. Fitting results and equivalent circuit are provided in Figure S9 (Supporting Information).

Figure 3 shows additional tests for the optimal (t_d = 70 s) PPy/Au LV network electrode. Figure 3a shows the CV curves at different scan rates. The quasirectangle shape persists for all scan rates, demonstrating excellent, fast capacitive response. The galvanostatic charge-discharge (GCD) curves at various currents, shown in Figure 3b, confirm this electrode excellent supercapac- itive property, including high and stable coulombic efficiency, shown in Figure 3c. The value of the specific areal capacitance, calculated from the GCD curves at different discharge current, also shown in Figure 3c, further confirms the excellent quality of the electrode. While the areal capacitance, drops only mod- erately (from 13.02 to 8.04 mF cm⁻²) with discharge current changing from 0.1 to 1 mA, the coulombic efficiency remains large, and essentially unchanged ≈99.56%. Figure 3d further confirms the high quality of the electrode, with only ≈7% loss of the capacitance after more than 1000 charge/discharge cycles.

The comparison of the CV curves before and after cycling (inset in Figure 3d), and the SEM images in Figure S10 (Supporting Information) further confirm the stability of this electrode.

In addition, to compare the performance between quasifractal and regular networks, we hence fabricated a regular network (grid) electrode with same working area, and inves- tigated the electrochemical performance of the single electrode with same PPy deposition time t_d = 70 s (same load mass) as shown in Figure S11 (Supporting Information).

The two network electrodes show comparable transmittance T ≈ 64% at λ = 550 nm, shown in Figure S11a (Supporting Information). EIS measurements of two electrodes were conducted and shown in Figure S11b (Supporting Informa- tion), apparently the leaf venation network electrode demon- strates a lower inside resistance and better charge transport property, compared to regular one. This could be attributed to the quasifractal nature of the leaf venation structure.^[19]

Moreover, as shown in Figure S11c,d (Supporting Informa- tion), the CV and GCD investigations of the two electrodes proved that the PPy/LV electrode provided a higher capacity density than the regular one. In detail, the areal capacitances for PPy/LV electrode and PPy/regular network electrode are 13 and 3.2 mF cm⁻², respectively, showing obvious advantage in capacitance for the LV network electrode.

Figure 2. Electrochemical performance of the PPy/Au LV electrodes. a) CV-curves taken at 20 mV s⁻¹. b) GD-curves taken at the current 0.1 mA.

c) Csp versus td, obtained from the GD curves. d) Nyquist impedance plots, inset shows the magnified section of the plots (frequency range:

0.01–100 kHz, amplitude: 5 mV).

(4)

The excellent quality of the optimal (t_d = 70 s) PPy/Au LV network electrode, promises a corresponding excellent performance of the assembled supercapacitor. As shown in Figure 4a it involves two such electrodes, with a gel electrolyte (PVA/LiCl) of thickness of 0.78 mm sandwiched between. To avoid possible electrolyte leakage problem in our devices, we introduced a frame made of acrylic double-sided adhesive to wrap the PVA/LiCl electrolyte. This structure can not only avoid electrolyte leakage and a short circuit during the bending state, but also maintain the original transmittance of the devices, which is barely possible for simply adding an opaque separator between two electrodes.^[10]

The active surface area of the supercapacitor is 2.5 × 1.5 cm². The device maintains good transparency T ≈ 45% at λ = 550 nm (see Figure 4b,c). As shown in Figure 4d, the supercapacitor device displays excellent CV performance, with the retention of an excellent quasisquare shape for different scan rates. Figure 4e shows good symmetry of the GCD curves at different currents. The corresponding C_sp = 5.6 mF cm⁻² at 0.1 mA, and C_sp = 4.4 mF cm⁻² at 1 mA (the full curve is shown in Figure S12 in the Supporting Information). To further understand the supercapacitor behavior, EIS measurements (see Figure S13, Supporting Information)

of the device were also provided. Figure 4f shows GCD curves (at I = 0.3 mA) of two supercapacitors in series, compared to the single device. The essentially identical time behavior of the two devices confirm an insignificantly small inside resistance in each, allowing for connections. Two devices connected in series show a higher potential, which will be useful in application involving wearable integrated electronics.

In addition, we have tested the long-term stability of the supercapacitor. Figure 4g shows electrochemical GCD cycling of the device at I = 1 mA. Impressively, our supercapacitor maintains 83% of the initial capacitance, and the almost unchanged symmetrical shape of GCD curves after 2500 cycles.

This is also confirmed by the less EIS tests shown in Figure S14 of the Supporting Information.

Flexibility is another desired feature, and the bending test is the best to asses this. Figure 5a shows CV cycling (taken at the scan rate of 100 mV s⁻¹), at different bending radi. Clearly, the changes are negligible, demonstrating excellent flexibility of our device. The capacitive performance under constant bending is shown in Figure 5b, demonstrating a negligible decrease in capacitance after 500 bending events.

Figure 3. Performance of the optimal (with td = 70 s) PPy/Au LV electrode. a) CV curves at different scan rate. b) GCD curves at different current density.

c) Change of the areal capacitance (specific capacity) calculated from the GCD measurements. d) Stability test: capacitance retention versus number of cycles. Inset: CV curves before and after 1000 cycles.

(5)

Figure 4. Performance of our flexible transparent supercapacitor. a) A schematic of the device structure. b) Photograph of supercapacitor, demonstrating high transparency. c) Optical transmittance spectra of the Au LV network, optimal PPy/Au LV electrode, and the flexible transparent supercapacitor. d) CV curves of the supercapacitor at different scan rates. e) GCD curves at different currents. f) GCD curves for a single device, and two connected in series (current I = 0.3 mA). g) Device stability for 2500 cycles at I = 1 mA. Inset: first ten, and the last ten cycles of the cycling curve.

(6)

Finally, we compare our LV network-based flexible and transparent supercapacitor devices, with other reported in the literature. Figure 5c and Table S1 (Supporting Information) demonstrate clearly, that our device has best, desired combination of good transmittance (≈45%), with the largest specific capacity (5.6 mF cm⁻²). For more details see Table S1 (Supporting Infor- mation). We also plot the energy density (E_d) versus the power density (P_d) for all devices in Figure 5d. Clearly, the best performance requires simultaneously large E_d and P_d, since the supercapacitor must not only accumulate large amount electric energy (E_d), but also this energy must be “mobile,” i.e., quickly delivered (fast charging) and extracted if needed, which is measured by P_d.^[21] Our devices delivered a maximum energy density of 0.28 µWh cm⁻² and maximum power density of 80 µW cm⁻², which clearly outperform most of the reported in the literature.

3. Conclusion

In summary, we have successfully fabricated a leaf venation-based network, with a specific capacity, measured in a

three-electrode system, as high as 13 mF cm⁻², at the current I = 0.1 mA, and simultaneously high optical transmissivity of 60%. The corresponding transparent supercapacitor device, assembled from these electrodes has transmittance of 45%, and the areal capacitance of 5.6 mF cm⁻², at I = 1 mA. This performance is one of the best among the reported so far. Moreover, our device shows an excellent electrochemical stability and mechanical flexibility, promising applications in the integrated, transparent wearable electronics.

4. Experimental Section

LV Network: First, a leaf venation was obtained via an alkali solution etching. A Magnolia alba leaves were immersed in 0.1 g mL⁻¹ NaOH solution at 50–70 °C for 3 h. By continually tapping the softened leaf, the mesophyll was removed leaving the vein structure of the leaf intact.

Second, the leaf venation was tiled on the polyethylene terephthalate (PET) film which was coated with negative photoresist. Followed by an ultraviolet lithography process, the leaf venation structure was transferred onto the PET film successfully. Finally, a uniform Au thin film was deposited on the template by magnetron sputtering with 100 W Figure 5. Flexibility study of our supercapacitor, and comparison with other reported devices. a) CV curves for our device at different bending radius (scan rate = 100 mV s⁻¹). Inset: photographs of the flexed device. b) Capacitance retention of our device under bending (500 cycles). Left inset:

images of the flexed device. Right inset: GCD curves. c) Comparison of the areal capacitance (specific capacity, in two-electrode configuration) C_sp, and transparency T (at λ= 550 nm), and d) the Rogon plot for ours, and other reported devices (indicated by a reference number).^[20]

(7)

for 7 min. Then the electrode was immersed in acetone to remove the photoresist. A transparent flexible Au LV network electrode was obtained.

PPy Coating: PPy was electrodeposited on the Au LV electrode by a potentiostatic method (at potential of 0.7 V),^[22] with the solution composed of 0.1 m HClO4 and an aqueous solution of the 0.1 m Pyrrole monomer.

Supercapacitor Assembly: Two identical PPy/Au LV network electrodes sandwiched a layer of the 1 m LiCl/PVA gel electrolyte. A frame made of acrylic double-sided adhesive with a thickness of 0.78 mm was introduced to wrap the PVA/LiCl electrolyte and package the device.

Characterization and Electrochemical Measurements: The morphologies and the EDX mapping were obtained using the FE-SEM (ZEISS Gemini 500, Germany). The FTIR (Nicolet 6700) was used to analyze PPy. The sheet resistance of electrodes was measured by the Keithley 2400 Source meter, and the transmittance obtained using UV–vis spectrophotometer.

The Electrochemical Test of the Single Electrode: CHI 660E electrochemical workstation (CH, Shanghai). A typical three electrode system with the LV network (PPy coated or as is) as the working electrode, Pt wire as the counter electrode, a saturated calomel electrode as the reference electrode, and 1 m LiCl aqueous solution as the electrolyte.

Supporting Information

Supporting Information is available from the Wiley Online Library or from the author.

Acknowledgements

The authors thank the financial support from National Key R&D Program of China (No. 2016YFA0201002) and NSFC-Guangdong Joint Fund (No. U1801256), NSFC (Nos. 51803064, 51571094, 51431006, and 51561135014), Guangdong Provincial Foundation (No. 2016KQNCX035), Program for Chang Jiang Scholars and Innovative Research Teams in Universities (No. IRT_17R40), and Guangdong Innovative Research Team Program (No. 2013C102). The authors also thank the support from the Guangdong Provincial Engineering Technology Research Center for Transparent Conductive Materials, National Center for International Research on Green Optoelectronics (IrGO), MOE International Laboratory for Optical Information Technologies and the 111 Project, and SCNU University Foundation (16KJ06).

Conflict of Interest

The authors declare no conflict of interest.

Keywords

current collector, energy storage device, leaf vein network electrode, transparent flexible supercapacitor

Received: August 13, 2019 Revised: September 2, 2019 Published online: September 19, 2019

[1] 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, A. Javey, Nature 2016, 529, 509.

[2] D. H. Ho, Q. Sun, S. Y. Kim, J. T. Han, D. H. Kim, J. H. Cho, Adv. Mater. 2016, 28, 2601.

[3] K. Takei, W. Honda, S. Harada, T. Arie, S. Akita, Adv. Healthcare Mater. 2015, 4, 487.

[4] Y. Wang, W. Zhou, Q. Kang, J. Chen, Y. Li, X. Feng, D. Wang, Y. Ma, W. Huang, ACS Appl. Mater. Interfaces 2018, 10, 27001.

[5] a) T. Chen, H. Peng, M. Durstock, L. Dai, Sci. Rep. 2015, 4, 3612;

b) R. Yuksel, Z. Sarioba, A. Cirpan, P. Hiralal, H. E. Unalan, ACS Appl. Mater. Interfaces 2014, 6, 15434.

[6] a) T. Chen, Y. Xue, A. K. Roy, L. Dai, ACS Nano 2014, 8, 1039;

b) H. Y. Jung, M. B. Karimi, M. G. Hahm, P. M. Ajayan, Y. J. Jung, Sci. Rep. 2012, 2, 773; c) K. Lee, H. Lee, Y. Shin, Y. Yoon, D. Kim, H. Lee, Nano Energy 2016, 26, 746.

[7] C. J. Zhang, B. Anasori, A. Seral-Ascaso, S. H. Park, N. McEvoy, A. Shmeliov, G. S. Duesberg, J. N. Coleman, Y. Gogotsi, V. Nicolosi, Adv. Mater. 2017, 29, 1702678.

[8] a) Z. H. Huang, Y. Song, D. Y. Feng, Z. Sun, X. Sun, X. X. Liu, ACS Nano 2018, 12, 3557; b) Y. Liu, Z. Zeng, B. Bloom, D. H. Waldeck, J. Wei, Small 2018, 14, 1703237; c) S. Zhu, L. Li, J. Liu, H. Wang, T. Wang, Y. Zhang, L. Zhang, R. S. Ruoff, F. Dong, ACS Nano 2018, 12, 1033.

[9] a) H. An, Y. Wang, X. Wang, L. Zheng, X. Wang, L. Yi, L. Bai, X. Zhang, J. Power Sources 2010, 195, 6964; b) M. Wang, L. Zhang, Y. Zhong, M. Huang, Z. Zhen, H. Zhu, Nanoscale 2018, 10, 17341;

c) Q. Wang, Y. Ma, X. Liang, D. Zhang, M. Miao, J. Mater. Chem. A 2018, 6, 10361.

[10] a) L. Shen, J. Wang, G. Xu, H. Li, H. Dou, X. Zhang, Adv. Energy Mater. 2015, 5, 1400977; b) B. Shi, B. Saravanakumar, W. Wei, G. Dong, S. Wu, X. Lu, M. Zeng, X. Gao, Q. Wang, G. Zhou, J.-M. Liu, K. Kempa, J. Gao, J. Alloys Compd. 2019, 790, 693;

c) X. L. Lu, Q. F. Zhang, J. Wang, S. H. Chen, J. M. Ge, Z. M. Liu, L. L. Wang, H. B. Ding, D. C. Gong, H. G. Yang, X. Z. Yu, J. Zhu, B. G. Lu, Chem. Eng. J. 2018, 358, 956.

[11] T. Qiu, B. Luo, M. Giersig, E. M. Akinoglu, L. Hao, X. Wang, L. Shi, M. Jin, L. Zhi, Small 2014, 10, 4136.

[12] J. Gao, Z. Xian, G. Zhou, J. M. Liu, K. Kempa, Adv. Funct. Mater.

2018, 28, 1705023.

[13] a) B. Han, K. Pei, Y. Huang, X. Zhang, Q. Rong, Q. Lin, Y. Guo, T. Sun, C. Guo, D. Carnahan, M. Giersig, Y. Wang, J. Gao, Z. Ren, K. Kempa, Adv. Mater. 2014, 26, 873; b) S. Kiruthika, C. Sow, G. U. Kulkarni, Small 2017, 13, 1701906.

[14] a) Y. H. Liu, J. L. Xu, S. Shen, X. L. Cai, L. S. Chen, S. D. Wang, J. Mater. Chem. A 2017, 5, 9032; b) H. Lee, S. Hong, J. Lee, Y. D. Suh, J. Kwon, H. Moon, H. Kim, J. Yeo, S. H. Ko, ACS Appl.

Mater. Interfaces 2016, 8, 15449; c) H. Sheng, X. Zhang, Y. Ma, P. Wang, J. Zhou, Q. Su, W. Lan, E. Xie, C. J. Zhang, ACS Appl.

Mater. Interfaces 2019, 11, 8992.

[15] a) G. Cai, P. Darmawan, M. Cui, J. Wang, J. Chen, S. Magdassi, P. S. Lee, Adv. Energy Mater. 2016, 6, 1501882; b) S. B. Singh, T. I. Singh, N. H. Kim, J. H. Lee, J. Mater. Chem. A 2019, 7, 10672;

c) J. L. Xu, Y. H. Liu, X. Gao, Y. Sun, S. Shen, X. Cai, L. Chen, S. D. Wang, ACS Appl. Mater. Interfaces 2017, 9, 27649.

[16] B. Han, Y. Huang, R. Li, Q. Peng, J. Luo, K. Pei, A. Herczynski, K. Kempa, Z. Ren, J. Gao, Nat. Commun. 2014, 5, 5674.

[17] B. Han, Q. Peng, R. Li, Q. Rong, Y. Ding, E. M. Akinoglu, X. Wu, X. Wang, X. Lu, Q. Wang, G. Zhou, J. M. Liu, Z. Ren, M. Giersig, A. Herczynski, K. Kempa, J. Gao, Nat. Commun. 2016, 7, 12825.

[18] a) M. Winter, R. J. Brodd, Chem. Rev. 2004, 104, 4245; b) W. Cao, E. Zhang, J. Wang, Z. Liu, J. Ge, X. Yu, H. Yang, B. Lu, Electrochim.

Acta 2019, 293, 364.

[19] J. Gao, K. Kempa, M. Giersig, E. M. Akinoglu, B. Han, R. Li, Adv. Phys. 2016, 65, 553.

[20] a) R. T. Ginting, M. M. Ovhal, J.-W. Kang, Nano Energy 2018, 53, 650; b) C. Zhang, T. M. Higgins, S. H. Park, S. E. O’Brien, D. Long, J. N. Coleman, V. Nicolosi, Nano Energy 2016, 28, 495.

[21] L. Fan, R. Ma, J. Wang, H. Yang, B. Lu, Adv. Mater. 2018, 30, 1805486.

[22] C. Zhang, J. Xiao, L. Qian, S. Yuan, S. Wang, P. Lei, J. Mater. Chem. A 2016, 4, 9502.