Published on 18 May 2016. Downloaded by Norwegian University of Science and Technology on 12/06/2016 22:59:53. Nanoscale COMMUNICATION Cite this: Nanoscale, 2016, 8, 11787 Received 11th February 2016, Accepted 24th April 2016 DOI: 10.1039/c6nr01200k www.rsc.org/nanoscale View Article Online View Journal | View Issue Amorphous MoSx thin-film-coated carbon fiber paper as a 3D electrode for long cycle life symmetric supercapacitors† Suresh Kannan Balasingam,‡a,b Arun Thirumurugan,c Jae Sung Leed and Yongseok Jun*b Amorphous MoSx thin-film-coated carbon fiber paper as a binderfree 3D electrode was synthesized by a facile hydrothermal method. The maximum specific capacitance of a single electrode was 83.9 mF cm−2, while it was 41.9 mF cm−2 for the symmetric device. Up to 600% capacitance retention was observed for 4750 cycles. Supercapacitors (also known as electric double layer capacitors or ultracapacitors) are a kind of electrochemical energy-storage device, with a high power density, fast charge–discharge rate, high capacitance, and superlong cycle life.1,2 Various carbonbased materials have been used in commercial supercapacitors with high power density owing to the electric double layer charge (EDLC)-storage mechanism. However, the energy density value of such supercapacitors is still much lower than those of commercial lithium ion batteries.3 In order to increase the energy density, various pseudocapacitive materials such as electrically conducting polymers, transition metal oxides (TMOs), transition metal carbides (TMCs), transition metal nitrides (TMNs), and transition metal dichalcogenides (TMDCs) have been extensively investigated.4–6 Molybdenum disulfide (MoS2), a class of TMDCs, is one of the most studied materials because of the following advantages: excellent electrocatalytic activity (next to noble metals, as evidenced by the volcano curve), a 2D layered crystal structure that facilitates ion intercalation (Li+ or other alkali metal ions), high surface area, and higher electronic conductivity a Department of Chemistry, School of Molecular Sciences, Ulsan National Institute of Science and Technology (UNIST), Ulsan 44919, Republic of Korea b Department of Materials Chemistry & Engineering, Konkuk University, Seoul 05029, Republic of Korea. E-mail: yjun@konkuk.ac.kr; Tel: +82-2450-0440 c Institute of Physics, Bhubaneswar 751 005, India d School of Energy and Chemical Engineering, Ulsan National Institute of Science and Technology (UNIST), Ulsan 44919, Republic of Korea † Electronic supplementary information (ESI) available. See DOI: 10.1039/ C6NR01200K ‡ Present Address: Department of Materials Science and Engineering, Faculty of Natural Sciences and Technology, Norwegian University of Science and Technology (NTNU), Trondheim 7491, Norway. E-mail: suresh.k.balasingam@ntnu.no This journal is © The Royal Society of Chemistry 2016 than that of transition metal oxides (TMOs).7 Because of these advantages, MoS2 has been utilized for various electrochemical applications such as sensors, hydrogen evolution catalysts, solar cells, and lithium/sodium ion batteries.8,9 Until 2007 when the capacitive behavior of MoS2 nanowall films was first reported by Loh’s group, researchers did not think about using TMDCs for application in supercapacitors.10 However, from 2007 onwards, a considerable number of articles have been published on MoS2 and its composite (MoS2/conducting polymer and MoS2/carbon)-based electrodes for supercapacitor applications.11 Most of the MoS2 synthesis methods are hightemperature solution-based chemical methods or they involve top-down exfoliation of bulk MoS2 crystals, forming highly crystalline MoS2 materials.4,11 In general, the amorphous phase of pseudocapacitive materials with low crystallinity shows a higher specific capacitance (Cs) than their highly crystalline counterparts.12 This phenomenon has been proved by a systematic study on the amorphous and crystalline NiO column structure and its surface and bulk contribution to the Cs value.13 Furthermore, Lu et al. experimentally verified that the bulk contribution of an amorphous NiO electrode was three times higher than that of its crystalline phase. Very recently, Zhang and co-workers synthesized an amorphous MoS2 shell over the crystalline Ni3S4 core (Ni3S4@MoS2 core– shell structure), which showed better capacitive performance than pure Ni3S4.14 In the present communication, we synthesized amorphous MoSx on a highly conductive 3-D carbon fiber paper network (CFP/a-MoSx). To the best of our knowledge, this is the first report on using a CFP/a-MoSx binder-free electrode material for supercapacitor applications. Herein, we adopted a facile hydrothermal method to deposit an amorphous MoSx thin layer on carbon fiber paper. Carbon fiber paper is hydrophobic. Hence, prior to conducting the hydrothermal growth of MoSx, the carbon fiber paper is pretreated with oxygen plasma to induce hydrophilicity (see the experimental details in the ESI†).8 The as-synthesized CFP/ a-MoSx electrode was subjected to X-ray diffraction (XRD) and Raman spectroscopic analysis (see Fig. S1 and S2 in the ESI†). The XRD pattern of CFP/a-MoSx is shown in Fig. S1.† Diffrac- Nanoscale, 2016, 8, 11787–11791 | 11787 View Article Online Published on 18 May 2016. Downloaded by Norwegian University of Science and Technology on 12/06/2016 22:59:53. Communication tion peaks appearing at 2θ = 26.5° and 54.5° correspond to the (002) and (004) planes of the carbon fiber paper (JCPDS no. 411487). Except carbon peaks, no characteristic peaks of molybdenum sulfide are observed, which infers that the as-deposited MoSx is amorphous in nature. Fig. S2† shows the Raman spectrum of CFP/a-MoSx. The typical peaks observed at 1358 cm−1 and 1587 cm−1 correspond to the D and G bands of the carbon fiber paper. Except D and G bands of the carbon material, no characteristic peaks (E2g1 and A1g modes) of molybdenum sulfide are observed. The absence of any characteristic peaks of MoS2 from XRD and Raman spectra inferred that the asdeposited material is amorphous. Fig. 1 shows the scanning electron microscopy (SEM) images of bare CFP (Fig. 1a and b) and those of MoSx-coated CFP (Fig. 1c and d), both at low and high magnifications. It is clearly evident that the high-magnification SEM image (Fig. 1d) shows a uniformly coated MoSx thin layer on the 3D-CFP network. To examine the thickness of the MoSx coating, a small portion of the sample was sliced using a focused ion beam equipped with the SEM instrument (FIB-SEM). To further confirm the structural information and thickness, the as-prepared thin slice of CFP/a-MoSx was examined using high-resolution transmission electron microscopy (HR-TEM). Fig. 1e shows the high-magnification TEM image of CFP/a-MoSx, which shows that the average thickness of the MoSx thin film was around 23 ± 1 nm. The absence of clear lattice fringes and the corresponding fast Fourier transform (FFT) image (inset of Fig. 1e) confirms the amorphous nature of MoSx, which is in agreement with the XRD and Raman measurements. Fig. 1f shows the high angle annular dark field-scanning transmission electron microscopy (HAADFSTEM) image of the CFP/a-MoSx and Fig. 1g and S6 (see the ESI†) show the corresponding energy dispersive spectrum (EDS) of a-MoSx as marked in the small (red square) region of Fig. 1f. From the magnified EDS spectrum (Fig. S7, see the ESI†), the presence of molybdenum and sulfur has been confirmed. The detection of copper might be due to the TEM grid Nanoscale and the gallium impurity originates from the FIB milling of TEM lamella sample preparation. The chemical composition and oxidation state of the elements were analyzed using X-ray photoelectron spectroscopy (XPS). Fig. 2a shows the Mo 3d spectra of amorphous MoSx. The doublets of the 3d orbital clearly resolved into two different peaks at 3d5/2 and 3d3/2 centered at 228.99 and 232.41 eV, respectively, which confirms the +4 oxidation state of Mo (i.e., Mo(IV)). The sulfur 2s orbital appears at the binding energy of 226.43 eV and a small hump of a Mo6+ peak appears at a higher binding energy of 235.8 eV; the latter may have formed because of the formation of oxides during the sample-preparation process.15 Because of the amorphous nature of MoSx, an unresolved broad hump of a sulfur 2p peak is also observed (Fig. 2b). From the atomic concentration table, the sulfur to molybdenum stoichiometric ratio was estimated at around 2.17. The excess sulfur present in the compound may be owing to the formation of MoS3(Mo4+[S2−][S22−]), rather than MoS2. Upon deconvolution, the doublet of S2− appears at lower binding energies and that of S22− at higher binding energies. The peak position of each orbital is tabulated in the ESI (Table S1 in the ESI†). The electrochemical properties of the CFP/a-MoSx binderfree electrode material were investigated for supercapacitor applications by using a symmetric two-electrode cell with an aqueous 0.5 M sulfuric acid electrolyte. Fig. 3a represents the cyclic voltammetry (CV) curves of the CFP/a-MoSx symmetric cell at various scan rates in the potential range varying from −0.3 to 0.6 V. At low scan rates, two pairs of prominent redox peaks can be clearly seen, which are attributed to the redox reaction of Mo atoms: Mo ↔ Mo(IV) ↔ Mo(VI).14 When the scan rate was increased, an obvious shift was seen in the redox peaks. Broad redox peaks are observed even at a higher scan rate, which confirmed that a predominant charge-storage reaction occurs via a redox reaction ( pseudocapacitance). Apart from pseudocapacitance, non-Faradaic processes owing to the Fig. 1 SEM images of bare CFP (a and b); CFP/a-MoSx (c and d); HR-TEM cross-sectional image of CFP/a-MoSx (e), the inset shows the FFT pattern; A HAADF-STEM cross-sectional image of CFP/a-MoSx (f); EDS spectrum of MoSx (g) (measured at the small portion of STEM image). 11788 | Nanoscale, 2016, 8, 11787–11791 This journal is © The Royal Society of Chemistry 2016 View Article Online Published on 18 May 2016. Downloaded by Norwegian University of Science and Technology on 12/06/2016 22:59:53. Nanoscale Fig. 2 High-resolution XPS spectra of CFP/a-MoSx: Mo 3d (a); and S 2p (b). adsorption and desorption of protons (or in some cases, alkali metal ions) on the surface of MoS2, which involve the double layer (capacitance) charge-storage mechanism, have also been reported.10,16–20 In some cases, the Faradaic process is only observable at a very lower scan rate; when the scan rate is increased due to the kinetic limitation of cation diffusion, the non-Faradaic process (EDLC behavior) becomes dominant.10,17 In our study, however, even at a higher scan rate of 125 mV s−1, broad redox peaks were observed with distorted rectangular CV curves, which confirmed that both the pseudocapacitance and EDLC-type charge-storage mechanism were involved. Eqn (S1) and (S2†) were used to measure the cell capacitance using CV and CD curves, and the corresponding areal specific capacitance of a cell (Cm), and a single electrode (Cs), was calculated using eqn (S3)–(S5) (see the experimental methods in the ESI†). The relationship between the scan rates and Cs is correlated (Fig. 3b). When the scan rate is increased gradually from 1 to 125 mV s−1, the Cs value decreases, probably owing to the insufficient time available for the insertion/extraction of protons into the interior (bulk) of the electrode.21 Based on this observation, it can be inferred that the electrochemical reaction kinetics is controlled by the protonic diffusion process at higher scan rates. When the scan rate was 1 mV s−1, a single electrode showed a Cs value of 83.9 mF cm−2 and the device This journal is © The Royal Society of Chemistry 2016 Communication showed a Cm value of 41.96 mF cm−2. The Cs value obtained in our study using the two-electrode configuration is still higher than that reported by Soon et al. (70 mF cm−2 @ 1 mV s−1, obtained using the three-electrode configuration).10 When the scan rate is increased to 5 mV s−1, Cm decreased to 20.4 mF cm−2; however, the Cm value obtained using our cell is much higher than that reported by a recent study on exfoliated MoS2 (1.83 mF cm−2) and graphene-exfoliated MoS2 composite (4.29 mF cm−2) based symmetric devices at the same scan rate.20 On further increasing the scan rate to 10 mV s−1, the corresponding Cs value dropped to 33.6 mF cm−2 and the Cm value decreased to 16.8 mF cm−2. However, the Cm value obtained in the present work is eight times higher than that in a previous report on exfoliated-MoS2 (2 mF cm−2 @ 10 mV s−1) and 33 times higher (0.5 mF cm−2 @ 10 mV s−1) than that obtained with bulk MoS2-based symmetric devices.19 To further evaluate the electrochemical performance of the CFP/a-MoSx electrode material, galvanostatic charge–discharge measurements were carried out for the symmetric device at current densities of 0.05–0.6 mA cm−2 (Fig. 3c). The nonlinear shape of the discharge curves further confirmed the pseudocapacitive charge-storage behavior. The relationship between Cs and the various current densities is shown in Fig. S3 (see the ESI†) and the Ragone plot of the symmetric device showing the areal energy and power density relationship is shown in Fig. S4 (see the ESI†). When the discharge rate was 0.3 mA cm−2, the corresponding Cs was 8.76 mF cm−2, which is more than five times that of exfoliated MoS2 (1.65 mF cm−2 @ 0.25 mA cm−2).20 The high specific capacitance of the CFP/ a-MoSx material may be due to the following reasons: (i) the high surface area of the 3D porous carbon fiber network with good electrical conductivity that increases the electron-transfer reaction rate, even though the as-synthesized MoSx is poorly crystalline. (ii) A thin amorphous MoSx film (thickness: approximately 23 ± 1 nm, as determined from TEM measurements) comprised of high-density grain boundaries with fine grains that act as diffusion channels for a more efficient intercalation of protons, which leads to an effective utilization of bulk electrode materials when compared to their crystalline counterparts.12 The higher areal specific capacitance of the CFP/a-MoSx electrode with the previously reported MoS2 based materials is compared in Table S2 (see the ESI†). Electrochemical impedance spectroscopy is one of the ideal tools to investigate the charge transfer mechanism at the electrode–electrolyte interface. Fig. 3d and S8 (see the ESI†) show the Nyquist and Bode plots of the symmetric device, respectively. From the Nyquist plot, a high-frequency intercept on the x-axis represents the electrochemical series resistance (ESR) of the symmetric device. The semicircle arc at the high- to medium-frequency range represents the charge transfer resistance (Rct), which is determined to be 0.5 Ω. A near vertical line (almost perpendicular to the x-axis) is observed at the medium-to-low frequency region, which represents the capacitive behavior of the device. The long-term cyclability is one of the key criteria for the real-time application of the device for commercial use. The Nanoscale, 2016, 8, 11787–11791 | 11789 View Article Online Published on 18 May 2016. Downloaded by Norwegian University of Science and Technology on 12/06/2016 22:59:53. Communication Nanoscale Fig. 3 CV curves of CFP/a-MoSx at various scan rates ranging from 1 to 125 mV s−1 (a). The relationship of specific capacitance of the electrode vs. scan rate (b). CD curves of CFP/a-MoSx at various current densities (c). EIS spectrum of the symmetric supercapacitor (d); the inset shows the highfrequency region. Long-term stability of the device measured as a function of capacitance retention vs. cycle number (e). CV curves of precycle and post-4750 cycles of the long-term stability test (f ). symmetric capacitor has been tested to investigate the continuous long-term charge–discharge process at a current density of 0.3 mA cm−2. The Cs value in the first cycle was around 8.46 mF cm−2, which increased to 50.4 mF cm−2 at the 4750th cycle. Most of the previous reports on the long-term stability of pseudocapacitive materials, including MoS2 electrode materials, showed generally either a decrease in capacitance retention or a stable value over many cycles.14,16–19,22,23 In contrast, the capacitance retention of CFP/a-MoSx synthesized in our study continuously increased up to 600% for 4750 cycles (Fig. 3e). This type of increase in capacitance is rarely observed in binder-free graphene-based supercapacitors (using KCl electrolyte) because of the “electroactivation” of electrode materials via continuous (K+) ion intercalation/deintercalation, which leads to a partial re-exfoliation process.24–26 Very recently, Bissett et al. also observed the same trend on exfoliated MoS2 thin-film-based symmetric capacitors (using Na2SO4 electrolyte), wherein the capacitance retention increased up to 800% over the first 3000 cycles.20 The authors investigated the charge storage mechanism by CV analysis before and after long-term cycling. Before cycling, the CV curves showed the EDLC-type charge-storage behavior; however, after long-term cycling for over 10 000 cycles, the shape of the post-CV curve distorted with enhanced current density. Hence, the authors concluded that the electro- 11790 | Nanoscale, 2016, 8, 11787–11791 chemical charge-storage mechanism changes from EDLC to pseudocapacitive nature because of the continuous ion intercalation/deintercalation process. Also, they proposed a mechanism involving continuous intercalation/deintercalation of relatively large Na+ ions over a long cycling process that further increased the active surface area by partial re-exfoliation of MoS2 layers, and hence, led to enhanced current density values in post-CV curves. In our case, energy storage occurs predominantly by pseudocapacitance (from the first cycle to 4750th cycle and even for high scan rate CV curves); hence, clear redox peaks are observed from the CV curves. After long-term cycling, post-CV analysis was done at the same scan rate, which shows characteristic redox peaks with an enhanced current density value (Fig. 3f ). Moreover, Bissett et al. investigated liquid-phase exfoliation of crystalline MoS2 powder, and the resultant exfoliated MoS2-flakes are also crystalline. However, in our case, the assynthesized MoSx is amorphous and has many defect sites. Also, the electrolyte used in our study is 0.5 M sulfuric acid; H+ cations act as shuttle ions, which are much smaller than Na+ ions. The observed increase in the current density value, as evident from the pre- and post-CV analyses, may be due to the “electroactivation” process. However the re-exfoliation of amorphous MoSx by H+ ions is not yet understood. In general, Li+ ion intercalation into bulk MoS2 facilitates the formation of This journal is © The Royal Society of Chemistry 2016 View Article Online Published on 18 May 2016. Downloaded by Norwegian University of Science and Technology on 12/06/2016 22:59:53. Nanoscale highly conducting 1T phase MoS2;27 however, a recent report on MoS2 and rGO hybrids showed that a heterostructure comprising 1T MoS2 formed at the MoS2/rGO interface and 2H MoS2 at the bulk.23 In our case, the as-synthesized material is amorphous. Hence, the in-depth structural analysis of electrodes, both before and after long-term measurement, does not give any detailed information (see Fig. S9,† post-mortem SEM image). The trend of increasing capacitance retention is interesting to researchers working on electrochemical energy storage devices. However, the increasing trend reaction mechanism is still unclear and presently under investigation. In summary, a thin layer of amorphous MoSx on a highly conducting 3-D network carbon fiber paper was synthesized by a facile hydrothermal method. The as-synthesized binder-free electrode material with excellent electrochemical energystorage capability and a symmetric device showed a high specific device capacitance value of 41.96 mF cm−2. More interestingly, because of “electroactivation”, the capacitance retention increases up to 600% for amorphous MoSx in the presence of a sulfuric acid medium. Although the exact mechanism of the increasing trend in capacitance retention is still under investigation, we believe that our primary observation and the obtained results may be interesting to the energy storage research community and may provide futuristic outcomes in this field. The authors gratefully acknowledge the financial support provided by the National Research Foundation of Korea (NRF) funded by the Korean government, MSIP/KEIT (2013R1A1A2059244, 2015M1A2A2056829, 2016R1A2B4007570, 10050509). References 1 S. Peng, L. Li, C. Li, H. Tan, R. Cai, H. Yu, S. Mhaisalkar, M. Srinivasan, S. Ramakrishna and Q. Yan, Chem. Commun., 2013, 49, 10178–10180. 2 A. Ramadoss, B. Saravanakumar, S. W. Lee, Y.-S. Kim, S. J. Kim and Z. L. Wang, ACS Nano, 2015, 9, 4337–4345. 3 A. Ramadoss, B. Saravanakumar and S. J. Kim, Nano Energy, 2015, 15, 587–597. 4 C. N. R. Rao, K. Gopalakrishnan and U. Maitra, ACS Appl. Mater. Interfaces, 2015, 7, 7809–7832. 5 C. N. R. Rao and U. Maitra, Annu. Rev. Mater. Res., 2015, 45, 29–62. 6 S. K. Balasingam, J. S. Lee and Y. Jun, Dalton Trans., 2016, DOI: 10.1039/C6DT00449K. This journal is © The Royal Society of Chemistry 2016 Communication 7 S. K. Balasingam, J. S. Lee and Y. Jun, Dalton Trans., 2015, 44, 15491–15498. 8 R. Bose, S. K. Balasingam, S. Shin, Z. Jin, D. H. Kwon, Y. Jun and Y.-S. Min, Langmuir, 2015, 31, 5220–5227. 9 C. Tan and H. Zhang, Chem. Soc. Rev., 2015, 44, 2713–2731. 10 J. M. Soon and K. P. Loh, Electrochem. Solid-State Lett., 2007, 10, 250–254. 11 X. Chia, A. Y. S. Eng, A. Ambrosi, S. M. Tan and M. Pumera, Chem. Rev., 2015, 115, 11941–11966. 12 Q. Lu, J. G. Chen and J. Q. Xiao, Angew. Chem., Int. Ed., 2013, 52, 1882–1889. 13 Q. Lu, Z. J. Mellinger, W. Wang, W. Li, Y. Chen, J. G. Chen and J. Q. Xiao, ChemSusChem, 2010, 3, 1367–1370. 14 Y. Zhang, W. Sun, X. Rui, B. Li, H. T. Tan, G. Guo, S. Madhavi, Y. Zong and Q. Yan, Small, 2015, 11, 3694– 3702. 15 X. Xiong, W. Luo, X. Hu, C. Chen, L. Qie, D. Hou and Y. Huang, Sci. Rep., 2015, 5, 9254. 16 M. Acerce, D. Voiry and M. Chhowalla, Nat. Nanotechnol., 2015, 10, 313–318. 17 A. Ramadoss, T. Kim, G.-S. Kim and S. J. Kim, New J. Chem., 2014, 38, 2379–2385. 18 G. Sun, J. Liu, X. Zhang, X. Wang, H. Li, Y. Yu, W. Huang, H. Zhang and P. Chen, Angew. Chem., Int. Ed., 2014, 53, 12576–12580. 19 A. Winchester, S. Ghosh, S. Feng, A. L. Elias, T. Mallouk, M. Terrones and S. Talapatra, ACS Appl. Mater. Interfaces, 2014, 6, 2125–2130. 20 M. A. Bissett, I. A. Kinloch and R. A. W. Dryfe, ACS Appl. Mater. Interfaces, 2015, 7, 17388–17398. 21 M. Lee, S. K. Balasingam, H. Y. Jeong, W. G. Hong, H.-B.-R. Lee, B. H. Kim and Y. Jun, Sci. Rep., 2015, 5, 8151. 22 G. Ma, H. Peng, J. Mu, H. Huang, X. Zhou and Z. Lei, J. Power Sources, 2013, 229, 72–78. 23 Q. Mahmood, S. K. Park, K. D. Kwon, S.-J. Chang, J.-Y. Hong, G. Shen, Y. M. Jung, T. J. Park, S. W. Khang, W. S. Kim, J. Kong and H. S. Park, Adv. Energy Mater., 2016, 6, 1501115. 24 M. Beidaghi and C. Wang, Adv. Funct. Mater., 2012, 22, 4501–4510. 25 Q. Cheng, J. Tang, J. Ma, H. Zhang, N. Shinya and L.-C. Qin, Phys. Chem. Chem. Phys., 2011, 13, 17615–17624. 26 Q. Cheng, J. Tang, J. Ma, H. Zhang, N. Shinya and L.-C. Qin, Carbon, 2011, 49, 2917–2925. 27 D. Voiry, A. Mohite and M. Chhowalla, Chem. Soc. Rev., 2015, 44, 2702–2712. Nanoscale, 2016, 8, 11787–11791 | 11791