Foldable supercapacitors from triple networks of ...

Nano Energy (]]]]) ], ]]]–]]]

1

Available online at www.sciencedirect.com

63

3

65

5

67 journal homepage: www.elsevier.com/locate/nanoenergy

7 9

71

RAPIDCOMMUNICATION

11

73

Foldable supercapacitors from triple networks of macroporous cellulose fibers, single-walled carbon nanotubes and polyaniline nanoribbons

13 15 17 19 21

27

Q1

Dengteng Gea, Lili Yanga,b,n, Lei Fanc, Chuanfang Zhangc, Xu Xiaoc, Yury Gogotsic, Shu Yanga,nn a

33

Department of Materials Science and Engineering, University of Pennsylvania, 3231 Walnut Street, Philadelphia, PA 19104, USA b School of Transportation Science and Engineering, Harbin Institute of Technology, Harbin, Heilongjiang Province 150090, P.R. China c A.J. Drexel Nanomaterials Institute and Department of Materials Science and Engineering, Drexel University, Philadelphia, PA 19104, USA

35

Received 3 August 2014; received in revised form 7 September 2014; accepted 8 November 2014

29 31

37 39 41 43 45 47 49

75 77 79 81 83

23 25

69

85 87 89 91 93 95 97 99

Q2

Abstract We report foldable supercapacitor electrodes using a macroporous cellulose fiber network, Kimwipess, as the scaffold through a simple “dip-absorption-polymerization” method. Singlewalled carbon nanotubes (SWCNTs) wrapped around the cellulose fibers as the conductive skin, while ultrathin ( 50 nm) and ultralong (tens of microns) polyaniline (PANI) nanoribbons were synthesized in situ between macroporous cellulose fibers and interpenetrated within the SWCNT network. The hybrid material showed good volumetric (40.5 F/cm3) and areal capacitance (0.33 F/cm2), which could be attributed to the synergistic effect between electron transport within the SWCNTs network and fast charge transfer of the PANI nanoribbons. The paper-based hybrid electrode was highly flexible and compliant; it could be folded back and forth as an origami crane up to 1000 times without mechanical failure or loss of capacitance. We believe that the combination of triple networks and the unique morphology of PANI

101 103 105 107 109 111

51

113

53

115

55 57

n Corresponding author at: Department of Materials Science and Engineering, University of Pennsylvania, 3231 Walnut Street, Philadelphia, PA 19104, USA, Tel./fax: +086 451 86282191. nn Corresponding author. Tel.: +215 898 9645; fax: +215 573 2128. E-mail addresses: [email protected] (L. Yang), [email protected] (S. Yang).

59 61

117 119 121

http://dx.doi.org/10.1016/j.nanoen.2014.11.023 2211-2855/& 2014 Elsevier Ltd. All rights reserved.

Please cite this article as: D. Ge, et al., Foldable supercapacitors from triple networks of macroporous cellulose fibers, single-walled carbon nanotubes and polyaniline nanoribbons, Nano Energy (2014), http://dx.doi.org/10.1016/j.nanoen.2014.11.023

123

2

D. Ge et al.

1

nanoribbons played critical roles to the repeated foldability. Finally, we assembled six all-solidstate supercapacitors based on the SWCNT/PANI nanoribbon paper electrodes connected in series, which lighted LED before, during and after folding for 500 cycles. & 2014 Elsevier Ltd. All rights reserved.

3 5 7

63 9

Introduction 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61

The growing interests in portable and wearable electronic devices have driven the search for low-cost, flexible, lightweight, and environmentally friendly energy storage devices such as supercapacitors (SCs), which bridge the gap between rechargeable batteries and conventional high power density electrostatic capacitors [1–5]. Typically, the capacitance and cyclability of SCs for such applications should be comparable to conventional SCs. Furthermore, because the devices will be wrapped around an object (e.g. body) or folded into a small piece, which can be reopened later when needed, the SC electrodes should be lightweight and flexible, and more importantly, bendable to a large extent (e.g. with sharp folds) without losing their electrical performance. This has been proved to be a major challenge. Although free-standing graphene and carbon nanotube (CNT)-based composite films have been studied extensively in literature [2,6–14], many suffer poor mechanical reliability [13,15] in addition to complicated preparation procedures and high material cost. Recently, much attention has been paid to the use of low-cost, flexible substrates, such as paper [16–22], fabrics [3,23–27], and cellulose nanofibers [28] as templates or scaffolds for depositing active conducting materials. Paper, consisting of cellulose fibers (diameter ranging from tens to hundreds of microns) interconnected with each other, is by far the cheapest and most exploited substrate for flexible SCs. There are two types of paper investigated in literature for SCs, including printing paper, which is rather nonporous, and filter paper and wiping paper, which are highly porous with macron-sized pores interconnected. Printing paper has been used as a supporting substrate to draw or deposit conducting materials [18–21], followed by in situ polymerization of nanostructures, for example, polyaniline (PANI) nanowires [18] and polypyrrole nanoparticles [19]. Filter paper, on the other hand, has been used to absorb active materials, for example, graphene nanosheets, which fill the voids between the cellulose fibers [16]. Ethylene-vinyl acetate (EVA) copolymer/CNT has been coated on the surface of lens cleaning paper, followed by electrodeposition of MnO2 [29]. However, very few have reported the capacity of the flexible electrodes after bending. Weng et al. bent graphene-cellulose paper one time [16] and Yuan et al. bent polypyrrole-coated paper 100 cycles [19], showing almost no change of the electrical performance. It has also been shown that 85% of initial capacitance is retained after 800 bending cycles from MnO2/ EVA/CNT paper with moderate curvature [29]. It remains unclear, however, whether these devices can survive repeated folding/unfolding, and sometime twisting, where the active materials could be detached or even destroyed at the interface if they are only deposited on the surface of paper [30].

Conducting polymers, such as polyaniline (PANI), are promising electrode materials because of its high specific pseudocapacitance, good environment stability, redox reversibility and low cost [31–33]. Nanostructured PANIs of different morphologies (e.g., nanofibres [34], nanorods [35], and nanowire arrays [36]) and their composites with carbon based materials (e.g. CNTs [37] and graphene [7]) have been synthesized. It has been suggested that the morphologies of nanostructured PANIs and their aggregations play important roles in energy storage, including material utilization, electron transport/ion diffusion pathways, and mechanical properties [38–40]. It is known that networks of high-aspect-ratio nanostructured materials (e.g., nanofibers, nanowires, and nanotubes) can be easily percolated, thus, offering improved mechanical strength and electrical conductivity.[6,41,42] Recently, we showed that dual-interpenetrating networks of single-walled carbon nanotubes (SWCNTs) and ultrathin PANI nanoribbons can be prepared as free-standing, flexible lithium ion battery (LIB) electrodes with high capacity and good cyclability.[40] The SWCNT aerogel and surfactants assembled on SWCNTs are critical to the formation of PANI nanoribbons. It will be attractive to exploit the macroporous network of cellulose fibers to grow SWCNT/PANI nanoribbons, and study the resulting electrical and mechanical performance, specifically, foldability and cyclability. Here, we coated the macroporous cellulose fibers (Kimwipess) with SWCNTs, followed by infiltration and in situ polymerization of aniline monomers. The SWCNTs were found wrapping around the interpenetrating cellulose fibers, and ultrathin and ultralong PANI nanoribbons were formed within the network of SWCNTs. The prepared SC electrodes were lightweight (0.34 g/cm3), and exhibited fairly high capacitance (40.5 F/cm3 or 0.33 F/cm2). The electrodes could be folded/unfolded repeatedly up to 1000 times without mechanical failure or loss of capacitance, which was in sharp contrast to SWCNT/PANI nanoparticle paper SCs with 35% loss of capacitance.

Experimental

65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 101 103 105

Materials 107 All reagents were used without further purification. Commercial Kimwipess (KIMTECH Sciencen brand delicate task wipers, Kimberly-Clark Corporation) with density of 0.25 g/cm3 or 2 mg/cm2 was used. Single walled carbon nanotubes (SWCNTs) were purchased from Cheap Tubes Inc. (purity490% with asheso1.5 wt%) with diameter of 12 nm and length of 530 mm. Sodium dodecylbenzene sulfonate (SDBS), aniline (An), camphorsulfonic acid (CSA), ammonium peroxydisulfate (APS), isopropyl alcohol (IPA), and ethanol were obtained from Sigma-Aldrich.


109 111 113 115 117

Foldable supercapacitors from triple networks

3

1

Preparation of SWCNT/PANI nanoribbon paper

Preparation of PANI nanorods/nanosheets

63

3

First, SDBS (20 mg) was dissolved into deionized (DI) water (1 mL). Then SWCNTs (1 mg) were added into the SDBS solution, followed by ultrasonic bath (Cole-Parmer Co., 12 W) for 12 h. A piece of Kimwipes tissue was immersed into the SWCNT solution for 10 s, followed by drying in an oven at 60 1C for 10 min. Due to the infiltration of SWCNTs, the originally white paper turned grey. The content of SWCNTs adsorbed on the paper was calculated as 0.1 mg/cm2 by comparing the weight difference before and after dipping, washing, and drying. The SWCNT infiltrated paper was immersed into a CSA-An aq. solution, where [CSA] = [An] =0.25 M, and kept for 12 h at room temperature. After absorption of aniline, the paper was then transferred into the CSA-APS aq. solution with [CSA] = 0.25 M and [APS]= 0.05 M in an ice water for different durations to polymerize aniline. The paper turned to dark green after the growth of polyaniline. The as-prepared paper was then washed by the ethanolwater (1:1 v:v) mixture and anhydrous ethanol, respectively, for 3–5 times. Finally, the samples were dried by a supercritical dryer (SAMDRIs–PVT-3D), followed by heating at 60 1C for 24 h.

The preparation procedure was similar to that of pristine PANI nanoribbon paper except that Kimwipess tissue was not used. SDBS solution, CSA-Aniline and CSA-APS solution were mixed in an ice water bath for 24 h. PANI was collected after wash by water, followed by centrifugation and air-drying.

65

Preparation of PANI paper from office copy paper

71

Preparation of neat SWCNT film

Characterization

5 7 9 11 13 15 17 19 21 23

The procedure was the same as that for preparation of PANI nanoribbon paper except that the office copy paper (Xerox Corporation) was used instead of Kimwipes.

Preparation of SWCNT/PANI nanowire paper

29 31 33 35 37 39 41 43

SDBS (64 mg) was dissolved into DI water (1 mL) followed by the addition of 8 mg SWCNTs. After ultrasonic bath for 18 h, the solution was poured onto the glass slides and formed gels after a few min. The gel was then immersed into the ethanol/water (1:1 v:v) mixed solution for 12 h, followed by washing and supercritical drying. The final thickness of the CNT film was 100 μm.

Preparation of pristine PANI nanoribbon paper Kimwipers were soaked into the SDBS aq. solution (20 mg/mL) for 4 h, and then immersed into a CSA-An aq. solution and a CSA-APS aq. solution, respectively, following the same procedure as that of preparation of SWCTN/PANI nanoribbon paper for in situ polymerization of aniline for 24 h.

Preparation of PANI nanofibers 47

51 53 55

PANI nanofibers were prepared by directly mixing the CSAAn solution and CSA-APS solution, both of which were with the same concentration as that in the preparation of SWCNT/PANI nanoribbon paper, in an ice water bath and stirred for 24 h. The solution turned green immediately, and became dark green after 24 h. The PANI solution was diluted by ethanol and then dropped onto the silicon water, followed by drying in the oven at 90 1C.

57

Preparation of PANI nanowire paper 59 61

73 75 77

81 83 85 87

45

49

69

79 The procedure was similar to the preparation of SWCNT/ PANI nanoribbon paper except that SDBS was carefully removed after SWCNTs were soaked into the Kimwipes by rinsing the SWCNT paper with DI water for many times. PANI polymerization on SWCNT paper was carried out under stirring at room temperature for 2 h. Finally, the SWCNT/ PANI nanowire paper was dried in the oven at 90 1C.

25 27

67

The synthesis followed the same procedure as that of PANI nanoribbon paper with the exception of no soaking of SDBS solution at the beginning.

89 The morphology of samples was characterized by a fieldemission high-resolution scanning electron microscope (FE HRSEM JEOL 7500 F SEM) at 20.0 kV.

Electrochemical testing A three-electrode setup in a Swagelok cell was used for all the electrochemical tests. Cyclic voltammetry (CV), galvanostatic charge/discharge (GCD), and electrochemical impedance spectroscopy (EIS) were conducted on a VMP3 potentiostat/ galvanostat (Biologic, France). SWCNT/PANI nanoribbon paper composite was used as the working electrode, Ag/AgCl and a piece of overcapacitive YP-50 F electrode (Kuraray Chemical Co. LTD) were used as the reference and counter electrode, respectively. These electrodes were kept in the vacuum furnace overnight at 60 1C before testing. The working electrode and counter electrode were separated with Celgard 2325 separator. The CV and GCD tests were conducted in 1 M H2SO4 electrolyte between 0.2 V to 0.6 V (vs Ag/AgCl). Electrochemical impedance spectroscopy (EIS) was performed between 10 mHz and 200 kHz at the cell open circuit potential with a signal peak to peak amplitude of 10 mV.

91 93 95 97 99 101 103 105 107 109 111 113

Mechanical tests of SWCNT/PANI nanoribbon paper 115 The SWCNT/PANI nanoribbon paper was cut into 4 cm 4 cm, and then subjected to two kinds of bending/folding tests, including 1) wrapping around a stick (diameter, 2 mm) and released up to 1000 times, and 2) folding into a crane and unfolding up to 1000 times along the creases (see white dashed lines in Fig. S6a). After the test, the center of the paper electrode (see Fig. S6b) was punched into a pellet (diameter, 7 mm) as the working electrode to test the electrochemical stability.


117 119 121 123

4

D. Ge et al.

1

Fabrication of all-solid-state supercapacitors

3

The fabrication of all-solid-state paper-based supercapacitors followed previous report [19]. A H3PO4/polyvinyl alcohol (PVA) gel was used as the electrolyte. Pristine Kimwipess was used as the separator between two SWCNT/PANI nanoribbon paper electrodes (1 4 cm). Six supercapacitors were connected in series and charged for 30 s to reach voltage 2.35 V. To check the stability before and after folding, six supercapacitors were each folded for 500 cycles in the middle section of the electrodes, then a green LED (the lowest potential is 1.9 V) was connected in series to the device.

5 7 9 11 13 15

Results and discussion 17 19 21 23 25 27 29 31 33 35 37

The SWCNT/PANI nanoribbon paper was prepared in three steps, including dipping the commercially available Kimwipess into the aqueous solution of SWCNT and sodium dodecylbenzene sulfonate (SDBS), adsorption of aniline monomer onto SWCNTs, and in situ polymerization of PANI nanoribbons (see Scheme 1). The Kimwipess delicate task wipers we used here are lint-free cellulose fibers interweaved together (see Fig. 1a). They have little chemical additives and are commonly used in laboratories for cleaning. After soaking with SWCNT/SDBS aqueous solution, the cellulose fibers were wrapped with SWCNTs as the conductive skin (Fig. 1b - c). The SWCNT paper was then immersed into an aqueous solution of aniline and ammonium peroxydisulfate (APS, oxidant) for in situ polymerization. As seen in Fig. 1d and e, PANI nanoribbons (50 nm thick and tens of microns long) were formed between networks of cellulose fibers and SWCNTs with an overall thickness of 80710 μm (Fig. 1f). PANI nanoribbons could be seen throughout the paper, suggesting uniformity in monomer infiltration and growth. The white paper turned grey and then dark green with the growth of PANI nanoribbons (Fig. 1g).

Here, the interweaving cellulose fibers in Kimwipess offered a macroporous network to host the conformal coating of SWCNT conductive ink. This is completely different from paper-based SC electrodes reported in literature, where the paper provides a flexible support or template to deposit conductive ink either on the surface of the paper [18–21], or between the pores [16,22]. Physical absorption of conductive ink only on the outer surface of the template or filling the gaps could make the electrodes vulnerable to delamination or crack formation upon bending and folding. After the paper was infiltrated with SWCNT/SDBS solution, the surface and interspace of cellulose fibers was enriched with anionic surfactant, SDBS. Our prior study [40] suggested that SDBS could form complexation with cationic aniline through Coloumbic attractions, leading to the formation of PANI nanoribbons within a porous network. As seen in Fig. S1, when pure SDBS solution (20 mg/mL) was introduced, ultralong PANI nanoribbons were observed. When neither SDBS nor Kimwipess was present, long PANI nanofibers were formed when mixing CAS-aniline solution and CSA-APS solution together (Fig. S1b). When either SDBS or Kimwipess was present in the solution, short PANI nanowires (Fig. S1c), nanorods or nanosheets (Fig. S1d) were formed on the surface of cellulose fibers. For comparison, a low-porosity Xerox office copy paper (see Fig. S2a) was used. Only a limited number of PANI nanoribbons (Fig. S2b) were observed in the interstices of the interconnected fibers, however, they were much shorter ( 8 μm) and thicker ( 150 nm) than those obtained when using Kimwipess. These results undoubtedly confirmed the synergetic effect of SDBS and highly porous networks (macroporous cellulose fibers and microporous SWCNT network) on the formation of ultrathin and ultralong PANI nanoribbons, which played a key role in the foldability of the electrode, as discussed later. To study the capacitive behavior of SWCNT/PANI nanoribbon electrodes, we carried out cyclic voltammetry (CV) measurements (Fig. 2a) using H2SO4 electrolyte in comparison

63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99

39

101

41

103

43

105

45

107

47

109

49

111

51

113

53

115

55

117

57

119

59

121

61 Scheme 1

Illustration of the preparation of SWCNT/PANI nanoribbon paper using Kimwipess as the scaffold.


123

Foldable supercapacitors from triple networks

5

1

63

3

65

5

67

7

69

9

71

11

73

13

75

15

77

17

79

19

81

21

83

23

85

25

87

27 29 31

89 Fig. 1 (a) SEM image of Kimwipes, showing the macroporous cellulose fiber network. (b) SEM image of SWCNT infiltrated paper and (c) high resolution SEM image of twisted SWCNT on the surface of cellulose fibers. (d) SEM image of PANI nanoribbons grown in situ within the macroporous network and (e) the corresponding high resolution SEM image of the translucent PANI nanoribbons. (f) Crosssectional SEM image of SWCNT/PANI nanoribbon-decorated paper. (g) Photographs of the original Kimwipes cut in a butterfly shape (left), SWCNT infiltrated paper (middle) and SWCNT/PANI nanoribbon paper (right).

33

39 41 43 45 47 49 51 53 55 57 59 61

93 95

35 37

91

97 with pure SWCNT film and pristine PANI nanoribbon paper prepared from the wiper without SWCNTs. SWCNT film presented a nearly rectangular CV curve with no obvious peaks, characteristic of electric double-layer capacitance of carbon-based materials [7]. PANI nanoribbon paper and SWCNT/PANI nanoribbon paper exhibited two groups of distinct redox peaks, indicating the redox transition of PANI from leucoemeraldine to the polaronic emeraldine form (C1/A1) and from emeraldine to pernigraniline (C2/A2) [43]. Fig. 2b shows the CV curves of SWCNT/PANI nanoribbon paper at different scan rates ranging from 2 to 100 mV/s, where the peaks (C1/A1) current density increased with the potential scan rates. It is reported that the current varies linearly with the scan rate v for the surface-controlled faradaic redox process [44], while the current has linear relationship with the square root of scan rate, v1/2, in the diffusion-controlled process [45]. By data fitting, we confirmed that the charge propagation was a combination of surface-controlled faradaic process and diffusion-controlled process in the SWCNT/PANI nanoribbon paper (see Fig. S3 and related discussion in Supporting Information). Both volumetric and areal capacitance have been used to evaluate the total energy storage capability of thin-film electrodes [46], while gravimetric capacitance could be used to characterize the contribution of individual components. The volumetric and areal capacitances were calculated from the galvanostatic cycling based on the last cycle after 10 cycles,

It Cvol ¼ V ΔU

ð1Þ

99

ð2Þ

101

where Cvol (F/cm3) and Cs (F/cm2) represent the volumetric and areal capacitance of the paper-based electrode. I, t, ΔU, V and S indicate the current (A), discharge time (s), the cell voltage (V), volume and surface area of paper-based electrode, respectively. Since the loading of SWCNTs in our system was very low, 3.6 wt% versus 26 wt% of the PANI nanoribbons, we believe that the contribution of SWCNTs to overall capacitance could be negligible. Here, we estimated the gravimetric capacitance of the electrode from PANI nanoribbons, CPANI, only

103

CS ¼

It S ΔU

It W PANI ΔU

105 107 109 111

ð3Þ

113

where WPANI is the weight of PANI nanoribbons in the paperbased electrode. To evaluate the contribution of PANI nanoribbons to energy storage, we measured the gravimetric capacitance of SWCNT/PANI nanoribbons obtained at different polymerization time, 4 h to 48 h, at a current density of 0.2 mA/cm2 (see Fig. 2c), where the PANI content increased from 10.6 wt.% to 26.4 wt.%. There was a slight increase of gravimetric capacitance from 4 h to 24 h, followed with an unexpected decrease when PANI polymerization time was

115

CPANI ¼


117 119 121 123

6

D. Ge et al.

1

63

3

65

5

67

7

69

9

71

11

73

13

75

15

77

17

79

19

81

21

83

23

85

25

87

27

89

29 31 33

91 Fig. 2 (a) CV curves of pure SWCNT film, PANI nanoribbon paper and SWCNT/PANI nanoribbon paper at 2 mV/s in 1 M H2SO4. (b) CV curves of SWCNT/PANI nanoribbon paper at different scan rates. The PANI polymerization time was 24 h. (c-d) Changes of gravimetric capacitance (c) and areal capacitance (d) of the SWCNT/PANI nanoribbon paper as a function of the weight content of PANI nanoribbons. The corresponding polymerization time with the increase of PANI content was 4 h, 12 h, 24 h, and 48 h, respectively. The capacitance of different samples here was all tested at a current density of 0.2 mA/cm2.

35 37 39 41 43 45 47 49 51 53 55 57 59 61

93 95 97

increased up to 48 h. As seen in Fig. S4, the morphology of PANI nanoribbons varied at different polymerization time. PANI nanoribbons appeared similar for polymerization time of 4 h to 24 h, with the density of PANI nanoribbons and their percolation increasing with time. However, at 48 h, thicker nanoribbons ( 100 nm thick) appeared and became more and more aggregated and agglomerated (Fig. S4d), leading to insufficient utilization of the redox active sites and deterioration of capacitance. Overall, the SWCNT /PANI nanoribbon paper exhibited a much higher capacitance 533.3 F/g (polymerization time, 24 h) compared to 220.8 F/g from the SWCNT/PANI nanowire paper we prepared (morphologies shown in Fig. S1b), 236 F/g from electrodeposited PANI on a SWCNT film [37] and 424 F/g from in situ polymerized PANI on a CNT sheet [47]. As seen in Fig. 2d, the SWCNT/PANI nanoribbon paper had an areal capacitance of 0.33 F/cm2 and a volumetric capacitance of 40.5 F/cm3 at a current density of 0.2 mA/ cm2. The areal capacitance of SWCNT/PANI nanowire paper was 72.5 mF/cm2 at a current density of 0.2 mA/cm2,22% of SWCNT/PANI nanoribbon paper. The mass loading of PANI in SWCNT/PANI nanoribbon paper could reach 0.68 mg/cm2, whereas in PANI nanowire paper, it was 0.32 mg/cm2. In the latter, PANI nanowires only appeared on the surface of the paper instead of wrapping around the cellulose fibers such as the nanoribbons. In turn, the gravimetric capacitance of SWCNT/PANI nanoribbon paper was 2.41 times of PANI

nanowires. We compared areal capacitance and volumetric capacitance of our system with other paper-based electrodes reported in literature in Table 1. It is clear that the volumetric/areal capacitance of our electrode was comparable or higher than those of paper-based electrodes reported in literature [17,19,21,29], except that of PANI/Au paper using electrodeposition method [18]. The high large areal and volumetric capacitances reported from the PANI/Au paper [18] may be attributed to the excellent electrical conductivity of the e-beam deposited Au layer, large PANI content at the long electrodeposition time, and a considerably thinner film (10 μm). To investigate the capacitance dependence on the chargedischarge rate, the SWCNT/PANI nanoribbon paper with PANI polymerization time of 24 h (Fig. 3a, Table S1) was evaluated under different current densities in comparison to pure SWCNT film and pristine PANI nanoribbon paper (i.e. no SWCNTs) (Fig. S5). As the current density increased from 0.2 to 2 mA/cm2, the SWCNT/PANI nanoribbon paper maintained 75% of initial capacitance (256 mF/cm2, Fig. 3b). In contrast, the pristine PANI nanoribbon paper and SWCNT film dropped 42% and 64% of the initial capacitance, respectively. Moreover, at the current density of 0.2 mA/cm2, the PANI nanoribbon paper and SWCNT films had smaller areal capacitance, 106 mF/cm2 and 46 mF/cm2, respectively, than the SWCNT/PANI nanoribbon paper (330 mF/cm2). The contribution of active material in the PANI nanoribbon paper


99 101 103 105 107 109 111 113 115 117 119 121 123

Foldable supercapacitors from triple networks 1

Table 1

7 63

Capacitance of different types of paper-based electrodes.

3

Materials

Areal capacitance (F/cm2)

Thickness (μm)

Volumetric capacitance (F/cm3)

References

65

5

MnO2/EVA/CNT paper PPy-coated paper PANI/Au paper PANI/graphite paper Graphite paper SWCNT/PANI nanoribbon paper

0.126 0.42 0.8 0.3556 0.0023 0.33

N/A 147 10n 92 N/A 80

N/A 28.6n 800 38.6n N/A 40.5

29 19 18 17 21 This paper

67

7 9 11

n

69 71 73

calculated from data reported in the reference.

13

75

15

77

17

79

19

81

21

83

23

85

25

87

27

89

29

91

31

93

33

95

35

97

37

99

39

101

41

103

43

105

45 47 49 51 53 55 57 59 61

Fig. 3 (a) Galvanostatic charge-discharge curves of SWCNT/PANI nanoribbon paper at different current densities. Polymerization time, 24 h. (b) Areal capacitance of the pure SWCNT film, PANI nanoribbon paper and SWCNT/PANI nanoribbon paper at different current densities. (c) Nyquist plots of pure SWCNT film, PANI nanoribbon paper and SWCNT/PANI nanoribbon paper. Inset: magnified high-frequency regions. (d) Cyclic stability of SWCNT/PANI nanoribbon paper, PANI nanoribbon paper and PANI nanowire paper at a current density of 0.2 mA/cm2.

107

network facilitated ion diffusion from the electrolyte to the PANI nanoribbons to maximize the utilization of PANI, leading to the improved electrochemical capacitance compared to the pristine PANI nanoribbon paper or pure SWCNT films. Electrochemical stability of electrodes during continuous charge-discharge cycles is another criterion to evaluate the performance of SCs. As seen in Fig. 3d, the SWCNT/PANI nanoribbon paper retained 79% of the initial capacity after 1000 cycles, while the pristine PANI nanoribbon paper also had 73% of capacity retention even though there were insufficient paths for electron transfer and ion diffusion as suggested by the Nyquist plot (Fig. 3d). Nevertheless, the capacity retention

113

was 160.3 F/g, far lower than that in SWCNT/PANI nanoribbon paper, 533.3 F/g. To understand their electrochemical behavior in different systems, specifically, the charge transport at the electrode/electrolyte interface, we performed electrochemical impedance spectroscopy (EIS) on the electrodes. As seen in Fig. 3c, pristine PANI nanoribbon paper showed large electronic resistance and poor capacitive behavior as expected from PANI [31]. In comparison, SWCNT/PANI nanoribbon paper showed a lower resistance and better capacitive behavior in both low-frequency and high-frequency regions of the Nyquist plot. This implies that the SWCNT network here acts as the conductive pathway, while the porosity within the


109 111

115 117 119 121 123

8

D. Ge et al.

1

63

3

65

5

67

7

69

9

71

11

73

13

75

15

77

17

79

19

81

21

83

23

85

25

87

27

89

29

91

31

93

33

95

35

97

37

99

39 41 43 45 47 49 51 53 55 57 59 61

101 Fig. 4 (a) Stability of electrode capacitance after repeated folding up to 1000 cycles. Insets: Pictures of the SWCNT/PANI nanoribbon paper before and after folding according to the creases indicated by the dashed lines. The dashed ring indicates the shape of pellet electrode punched for electrochemical testing. (b) Picture of six solid-state supercapacitors based on SWCNT/PANI nanoribbon paper electrodes after folding 500 cycles were connected in series to light a green LED. The clamp represented the folding line. Insets: picture of released supercapacitors after folding and the enlarged picture of the lighting green LED. (c-d) SEM image (c) and the corresponding illustration (d) of SWCNT/PANI nanoribbon paper after folding 1000 times. Arrows in (c) indicate the broken cellulose fiber. (e) SEM image of SWCNT/PANI nanowire paper after folding 1000 times. (f) The capacitance retention of SWCNT/PANI nanowire paper after bending and folding. The schematic illustrates the composite structure after folding.

doubled that of SWCNT/PANI nanowire (50 nm in diameter and 150 nm long) paper, which decreased to 37% after 1000 charge-discharge cycles. It is clear that the PANI nanoribbons can better accommodate the volume change during the chargedischarge cycling. We note that while relatively short lifetime remains an issue for PANI-based supercapacitors, it may be sufficient for flexible/wearable electronics, sensors, and biomedical devices, where only hundreds to thousands of charge/ discharge cycles are required. Further, it has been shown that cyclability of PANI can be dramatically improved by adding finely dispersed nanodiamond particles [39], which could be potentially applied to our system to improve the cycle life of the electrode. Moreover, we assembled all-solid-state supercapacitors (SCs) based on the SWCNT/PANI nanoribbon paper

electrodes (polymerization time, 24 h) using H3PO4/polyvinyl alcohol (PVA) electrolyte. The schematic diagram for the fabrication of all-solid-state SCs and their CV, GCD curves are presented in Fig. . The CV scans at different scan rates in the potential window of 0–0.8 V (Fig. S6a) and the charge-discharge curve at the current density of 0.5 mA/cm2 (Fig. S6b) suggest a satisfactory capacitive performance of all-solid-state SCs based on the SWCNT/PANI nanoribbon paper electrodes. Portable and wearable devices are often subjected to bending and folding. Although many have reported paperbased SC electrodes, little is known about their lifetime under repeated bending or folding, especially to a large strain. Above the critical strain, the active layer could detach from the current collector or fracture. Graphene/MnO2 has been shown


103 105 107 109 111 113 115 117 119 121 123

Foldable supercapacitors from triple networks 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61

stable after being bent to 1801 once, however, no data was reported for repeated bending [6]. Polypyrrole-coated paper can bend 100 cycles but at a rather large radius of curvature (tens of millimeters) [17]. Jost et al. have recently developed textile supercapacitors as wearable storage from knitted carbon fibers, which are screen printed with activated carbon particles [3,24,25]. The device could be bent almost in half. However, in the repeated stretching and bending up to 1801 for 11 times, it loses 20% of initial capacitance, possibly due to breakage of the conducting network between carbon particles [25]. Further, a hysteresis of capacitance retention is observed between the flat and stretched or bent states in the textile SCs. After resting the device for 6 h, it can regain some capacitance. Liu et al. recently have prepared PANI and reduced graphene oxide (PANI-rGO) paper and find that the capacitance loses only 5% after folding. No data is reported after multi-folding tests [22]. Our composite electrode showed superior compliance toward bending and folding. It could wrap around a stick (diameter of 2 mm) repeatedly up to 1000 cycles with almost no change of capacitance (Fig. S7) and did not require any resting time. We then performed origami following the direction and sequence shown in Fig. 4a. After creasing and folding/unfolding the composite electrode 1000 times, no visible damage or loss in the energy storage was observed. We then assembled six all-solid-state supercapacitors in series based on the SWCNT/PANI nanoribbon paper electrodes. The green LED lighted by the device demonstrated good stability (lighting for 2 min) before, during and after folding for 500 cycles (Fig. 4b). We believe that the unprecedented flexibility and foldability of SWCNT/PANI nanoribbon paper could be attributed to 1) the intrinsic flexibility of ultrathin and ultralong PANI nanoribbons and 2) the interpenetrating triple networks (cellulose fibers, SWCNTs, and PANI nanoribbons), which allowed for the maintenance of material and conductivity integrity during repeated folding. As seen in Fig. S8 and related discussion in Supporting Information, when a thick material is bent to a large degree, it can easily fracture due to a relatively small radius of curvature and maximum allowable stress (Fig. S8b). Likewise, a short structure has a small radius of curvature, which is proportional to the length, and therefore, will experience a large bending stress (Fig. S8c). Moreover, bending could cause the loss of connections at interfaces (Fig. S8d). High-aspect-ratio materials can accommodate the external force through slippage without the loss of connection between each other (Fig. S8e). As seen in the SEM image (Fig. 4c), even if the cellulose fiber was broken (indicated by arrows) after folding 1000 times, PANI nanoribbons remained connected within the SWCNT and cellulose fiber networks upon bending (see Fig. 4d, site A), buckling and/or slippage (see Fig. 4d, site B). Because they are ultrathin ( 50 nm) and ultralong (tens of microns), PANI nanoribbons could easily make contact with the adjacent ones, even after they were broken apart upon folding (Fig. 4d, site C). In the case of SWCNT/PANI nanowire paper, PANI nanowires were polymerized on the outer surface of the cellulose fiber network. After folding for 1000 times, PANI nanowires ruptured on the surface (Fig. 4f, site A), between the fibers (Fig. 4f, site B), or became completely exfoliated (Fig. 4e and f, site C), leading to weight loss of nanowires ( 27%). We believe

9 both the weight loss and breakage of the conductive network contributed to the overall decrease of capacitance after folding.

63

Conclusions

67

In summary, we prepared a lightweight and highly foldable supercapacitor electrode using Kimwipess with macroporous cellulose fibers as scaffold. Through a “dip-absorption-polymerization (DAP)” approach, we created triple semiinterpenetrating networks of SWCNTs and ultrathin and ultralong PANI nanoribbons, wrapping around the interconnected cellulose fibers instead of simply coating the fiber surface or filling in-between. The resulting SWCNT/PANI nanoribbon paper was lightweight (0.34 g/cm3) and exhibited good volumetric capacitance (40.5 F/cm3) and satisfactory cycling performance compared to other paper-based supercapacitor electrodes or PANI based electrodes. More importantly, they displayed superior mechanical compliance toward bending and folding (up to 1000 cycles) while maintaining electrochemical stability. We believe that the work presented here not only suggests a new way to create low-cost, foldable and efficient paper-based supercapacitor electrodes, but sheds light on the role of porosity in the preparation of ultrathin nanomaterials with different morphologies. These new insights could potentially be applied to other 1D systems to create flexible electronics, wearable textiles, nanophotonics and displays. Further, it is possible to significantly increase the cyclability of the SWCNT/PANI nanoribbon paper by introducing nanodiamond particles [39].

65

69 71 73 75 77 79 81 83 85 87 89 91 93

Acknowledgments

95

LY would like to acknowledge the partial support from the National Natural Science Foundation of China (No. 5112068). Penn Nanoscale Characterization Facility (NCF) is acknowledged for access SEM. Tianqi Li from Prof. Yury Gogotsi's group is acknowledged for the help in electrochemical testing of solid-state paper-based supercapacitors.

97 99 101 103

Appendix A.

Supporting information

Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/ j.nanoen.2014.11.023.

105 107 109 111

References 113 [1] S. Bauer, Nat. Mater. 12 (2013) 871. [2] M.F. El-Kady, V. Strong, S. Dubin, R.B. Kaner, Science 335 (2012) 1326. [3] K. Jost, G. Dion, Y. Gogotsi, J. Mater. Chem. A 2 (2014) 10776. [4] J.R. Miller, Science 335 (2012) 1312. [5] P. Simon, Y. Gogotsi, Nat. Mater. 7 (2008) 845. [6] Y.M. He, W.J. Chen, X.D. Li, Z.X. Zhang, J.C. Fu, C.H. Zhao, E.Q. Xie, ACS Nano 7 (2013) 174. [7] H.P. Cong, X.C. Ren, P. Wang, S.H. Yu, Energ. Environ. Sci. 6 (2013) 1185. [8] L.L. Zhang, R. Zhou, X.S. Zhao, J. Mater. Chem. 20 (2010) 5983. [9] S. Hu, R. Rajamani, X. Yu, Appl. Phys. Lett. 100 (2012) 104103.


115 117 119 121 123

10 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61

[10] K. Wang, Q.H. Meng, Y.J. Zhang, Z.X. Wei, M.H. Miao, Adv. Mater. 25 (2013) 1494. [11] S. Boukhalfa, K. Evanoff, G. Yushin, Energ. Environ. Sci. 5 (2012) 6872. [12] K.Z. Gao, Z.Q. Shao, X. Wu, X. Wang, Y.H. Zhang, W.J. Wang, F.J. Wang, Nanoscale 5 (2013) 5307. [13] X. Xiao, X. Peng, H. Jin, T. Li, C. Zhang, B. Gao, B. Hu, K. Huo, J. Zhou, Adv. Mater. 25 (2013) 5091. [14] J. Ge, H.B. Yao, W. Hu, X.F. Yu, Y.X. Yan, L.B. Mao, H.H. Li, S.S. Li, S.H. Yu, Nano Energ. 2 (2013) 505. [15] B.G. Choi, J. Hong, W.H. Hong, P.T. Hammond, H. Park, ACS Nano 5 (2011) 7205. [16] Z. Weng, Y. Su, D.W. Wang, F. Li, J. Du, H.M. Cheng, Adv. Energ. Mater 1 (2011) 917. [17] B. Yao, L.Y. Yuan, X. Xiao, J. Zhang, Y.Y. Qi, J. Zhou, J. Zhou, B. Hu, W. Chen, Nano Energ. 2 (2013) 1071. [18] L.Y. Yuan, X. Xiao, T.P. Ding, J.W. Zhong, X.H. Zhang, Y. Shen, B. Hu, Y.H. Huang, J. Zhou, Z.L. Wang, Angew. Chem. Int. Edit. 51 (2012) 4934. [19] L.Y. Yuan, B. Yao, B. Hu, K.F. Huo, W. Chen, J. Zhou, Energ. Environ. Sci. 6 (2013) 470. [20] X.D. Zhang, Z.Y. Lin, B. Chen, S. Sharma, C.P. Wong, W. Zhang, Y.L. Deng, J. Mater. Chem. A 1 (2013) 5835. [21] G.Y. Zheng, L.B. Hu, H. Wu, X. Xie, Y. Cui, Energ. Environ. Sci. 4 (2011) 3368. [22] L. Liu, Z. Niu, L. Zhang, W. Zhou, X. Chen, S. Xie, Adv. Mater. 26 (2014) 4855. [23] L. Yuan, X.H. Lu, X. Xiao, T. Zhai, J. Dai, F. Zhang, B. Hu, X. Wang, L. Gong, J. Chen, C. Hu, Y. Tong, J. Zhou, Z.L. Wang, ACS Nano 6 (2011) 656. [24] K. Jost, C.R. Perez, J.K. McDonough, V. Presser, M. Heon, G. Dion, Y. Gogotsi, Energ. Environ. Sci. 4 (2011) 5060. [25] K. Jost, D. Stenger, C.R. Perez, J.K. McDonough, K. Lian, Y. Gogotsi, G. Dion, Energ. Environ. Sci 6 (2013) 2698. [26] X. Xiao, T. Li, P. Yang, Y. Gao, H. Jin, W. Ni, W. Zhan, X. Zhang, Y. Cao, J. Zhong, L. Gong, W.C. Yen, W. Mai, J. Chen, K. Huo, Y.L. Chueh, Z.L. Wang, J. Zhou, ACS Nano 6 (2012) 9200. [27] L. Bao, X. Li, Adv. Mater. 24 (2012) 3246. [28] K.Z. Gao, Z.Q. Shao, J. Li, X. Wang, X.Q. Peng, W.J. Wang, F.J. Wang, J. Mater. Chem. A 1 (2013) 63. [29] Z. Zhang, W. Wang, C. Li, L. Wei, X. Chen, Y. Tong, K. Mai, X. Lu, J. Power Sources 248 (2014) 1248. [30] B.J. Kim, Y. Cho, M.S. Jung, H.A.S. Shin, M.W. Moon, H.N. Han, K.T. Nam, Y.C. Joo, I.S. Choi, Small 8 (2012) 3300. [31] D.W. Wang, F. Li, J.P. Zhao, W.C. Ren, Z.G. Chen, J. Tan, Z.S. Wu, I. Gentle, G.Q. Lu, H.M. Cheng, ACS Nano 3 (2009) 1745. [32] S. Bhadra, D. Khastgir, N.K. Singha, J.H. Lee, Prog. Polym. Sci. 34 (2009) 783. [33] H.Y. Mi, X.G. Zhang, S.D. Yang, X.G. Ye, J.M. Luo, Mater. Chem. Phys. 112 (2008) 127. [34] J.X. Huang, Pure Appl. Chem. 78 (2006) 15. [35] S.H. Kim, K.W. Oh, J.H. Choi, J. Appl. Polym. Sci. 116 (2010) 2601. [36] K. Wang, J.Y. Huang, Z.X. Wei, J. Phys. Chem. C 114 (2010) 8062. [37] Z.Q. Niu, P.S. Luan, Q. Shao, H.B. Dong, J.Z. Li, J. Chen, D. Zhao, L. Cai, W.Y. Zhou, X.D. Chen, S.S. Xie, Energ. Environ. Sci. 5 (2012) 8726. [38] H.H. Zhou, H. Chen, S.L. Luo, G.W. Lu, W.Z. Wei, Y.F. Kuang J. Solid State Electr. 9 (2005) 574. [39] J. Benson, I. Kovalenko, S. Boukhalfa, D. Lashmore, M. Sanghadasa, G. Yushin, Adv. Mater. 25 (2013) 6625. [40] D. Ge, L. Yang, A. Honglawan, J. Li, S. Yang, Chem. Mater. 26 (2014) 1678. [41] M.N. Hyder, S.W. Lee, F.C. Cebeci, D.J. Schmidt, Y. Shao-Horn, P.T. Hammond, ACS Nano 5 (2011) 8552. [42] P.J. King, T.M. Higgins, S. De, N. Nicoloso, J.N. Coleman, ACS Nano 6 (2012) 1732. [43] T.F. Otero, I. Boyano, J. Phys. Chem. B 107 (2003) 4269.

D. Ge et al. [44] Y.F. Xu, M.G. Schwab, A.J. Strudwick, I. Hennig, X.L. Feng, Z.S. Wu, K. Mullen, Adv. Energ. Mater 3 (2013) 1035. [45] T. Brezesinski, J. Wang, S.H. Tolbert, B. Dunn, Nat. Mater. 9 (2010) 146. [46] Y. Gogotsi, P. Simon, Science 334 (2011) 917. [47] C.Z. Meng, C.H. Liu, S.S. Fan, Electrochem. Commun. 11 (2009) 186.

63

Dengteng Ge received his B.S. (2005), M.S. (2007) and Ph.D. degree (2011) in Materials Science and Engineering from Harbin Institute of Technology (HIT), China. He joined in Prof. Shu Yang's group at University of Pennsylvania (USA) as a post-doctoral researcher from May 2012 to now. His current research interests include flexible electrical device, structure color, and surface wettability.

71

Lili Yang received her B.S. degree (2003) in Polymer Science and Engineering, M.S. degree (2005) and Ph.D. degree (2009) in Materials Science and Engineering from Harbin Institute of Technology (HIT), China. She is now an associate professor in School of Transportation Science and Engineering at HIT. Her research focuses on novel functional materials and their applications in transportation. She visited Prof. Shu Yang's group at University of Pennsylvania in 2012 and worked on the design and fabrication of flexible energy storage systems.

81

65 67 69

73 75 77 79

83 85 87 89 91 93

Lei Fan received his Ph.D. degree from Yangzhou University in June 2009, China. He now works at Yangzhou University on surface and colloid interface science. During 2013 to 2014, he was a visiting scholar at Drexel University (USA), working with Prof. Yury Gogotsi. His research interests include fabrication of metal oxide nanomaterials in colloid systems and their applications in supercapacitors and batteries.

95 97 99 101 103

Chuanfang Zhang received his B.S and M.S. from East China University of Science and Technology (ECUST), Shanghai, China in 2011 and now a Co-Ph.D. candidate of ECUST and Drexel University, Philadelphia, USA under the supervision of Prof. Yury Gogotsi. His research interests mainly lie in the transition metal oxide/carbon composites for supercapacitor and Li-ion battery. He developed various advanced electrode fabrication techniques and optimized fundamental testing methods for supercapacitor. Currently he works on two-dimensional carbides for promising energy storage application.

105 107 109 111 113 115 117

Xu Xiao received his B.S. degree in School of Physics from Huazhong University of Science and Technology (HUST), PR China in Jun, 2011. Now he is a Ph.D. candidate in Wuhan National Laboratory for Optoelectronics (WNLO) and School of Optical and Electronic Information at HUST. His research interests include flexible


119 121 123

Foldable supercapacitors from triple networks 1

electronics and flexible solid-state supercapacitors for self-powered systems.

3 5 7 9 11 13 15 17 19

Yury Gogotsi is Distinguished University Professor and Trustee Chair of Materials Science and Engineering at Drexel University. He is also the founding Director of the A.J. Drexel Nanomaterials Institute and Associate Editor of ACS Nano. His Ph.D. is in Physical Chemistry from Kiev Polytechnic and D.Sc. in Materials Engineering from Ukrainian Academy of Sciences. He works on nanostructured carbons and other nanomaterials for energy related and biomedical applications. He has coauthored 2 books, more than 400 journal papers and obtained more than 50 patents. He has received numerous national and international awards for his research, was recognized as Highly Cited Researcher by Thomson-Reuters in 2013 and elected a Fellow of AAAS, MRS, ECS and ACerS and a member of the World Academy of Ceramics.

11 Shu Yang is Professor in the Department of Materials Science and Engineering at University of Pennsylvania (Penn). She received her B. S. degree in Materials Chemistry from Fudan University, China in 1992, and Ph. D. degree in Chemistry and Chemical Biology from Cornell University in 1999. She worked Bell Laboratories, Lucent Technologies as a Member of Technical Staff before joining Penn in 2004. Her research interests include synthesis and engineering of well-‐defined polymers and inorganic materials with controlled size, shape, and morphology over multiple lengthscales, study of their directed assembly and unique surface, optical, and mechanical properties, as well as dynamic tuning.

23 25 27 29 31 33 35 37 39 41

21
