Shuihua Tang*ab,
Leping Suiab,
Zhen Daiac,
Zhentao Zhuab and
Haixin Huangfuab
aState Key Lab of Oil and Gas Reservoir Geology & Exploitation, Southwest Petroleum University, Chengdu 610500, China. E-mail: shuihuatang@swpu.edu.cn; Fax: +86-2883032879; Tel: +86-2883032879
bSchool of Materials Science and Engineering, Southwest Petroleum University, Chengdu 610500, China
cSchool of Chemistry and Chemical Engineering, Southwest Petroleum University, Chengdu 610500, China
First published on 29th April 2015
A 60 wt% Ni(OH)2/XC-72 composite was synthesized via a facile and rapid microwave-assisted method in an ethylene glycol medium. Transmission electron microscopy images show that Ni(OH)2 is uniformly dispersed on XC-72 and that when carbon nanotubes (CNTs) are added as a conductive agent, they construct a three-dimensional (3-D) network together with the Ni(OH)2 particles and XC-72. The composite demonstrates a specific capacitance of 1296 F g−1 at a scan rate of 2 mV s−1 and 1560 F g−1 at a current density of 1 A g−1 in 6 M KOH aqueous solution, and its retention of specific capacitance is 71% after 1000 cycles. This can be attributed to XC-72 being an excellent support and CNTs with high aspect ratio and good conductivity to construct a 3-D conductive network. This excellent electrochemical performance makes the Ni(OH)2/XC-72 composite promising as an electrode material for a supercapacitor.
In general, a supercapacitor can be divided into two types based on its energy storage principle. One is defined as an electrical double layer capacitor (EDLC) in which electrical energy is electrostatically stored at an interface between a conductive electrode and an electrolyte. Typically, commercial EDLC electrode materials are porous carbon nanomaterials, such as activated carbon,4 carbon nanotubes5,6 and carbon aerogels,7,8 which have large specific surface areas, good electron conductivities and stabilities. They are, however, limited by low specific capacitances and energy densities. The other energy storage principle is defined as a pseudocapacitor in which electrical energy is stored by Faradaic reactions such as redox reactions, intercalation, or electrosorption.9 The most studied materials for pseudocapacitors are conducting polymers such as polyaniline10 and transition metal oxides/hydroxides such as RuO2,11 MnO2,12,13 Co3O4,14 Ni(OH)2/NiO15 and α-FeOOH.16 These materials always have high specific capacitances but are limited by poor cycling stabilities.17–19
Among these many materials for pseudocapacitors, Ni(OH)2 has been extensively researched due to its low cost, high theoretical specific capacitance (2082 F g−1), simple synthesis and availability of various morphologies.1,20,21 However, there are issues because of large particle sizes, poor stability, and bad electron conductivity. Hence, composites of Ni(OH)2/carbon materials have been widely investigated.22–26
Huang et al. has synthesized a Ni(OH)2/activated carbon composite by a chemical precipitation method. When nickel hydroxide loading was 6 wt%, the composite showed a high specific capacitance of 314.5 F g−1, which was 23.3% higher than pure activated carbon (255 F g−1).4 Chen et al. synthesized α-Ni(OH)2/reduced graphene oxide/carbon nanotubes by a one-pot hydrothermal method and found that the amount of carbon nanotubes could significantly influence the electrochemical performance of the devices. The composite with a GO/CNT mass ratio of 20:2 and concentration of NiCl2·6H2O of 59.5 mg mL−1 exhibited a specific capacitance of 1320 F g−1 at 6 A g−1 and the capacitance was reduced by only 7.8% after 1000 cycles.5 Zhang et al. prepared a 3-D hierarchical composite of α-Ni(OH)2/graphite nanosheets by introducing the graphene oxide nanosheets into α-Ni(OH)2, then a 3-D hierarchical porous structure of fine α-Ni(OH)2 nanocrystals as building blocks was formed directly on the matrix of graphite nanosheets in the presence of urea. This composite exhibited a high specific capacitance of 1956 F g−1 at 1 A g−1 and capacitance retention reached 70% after 1000 cycles.22 Ji et al. synthesized a free-standing Ni(OH)2/ultrathin-graphite foam (UGF) composite and an asymmetric supercapacitor made with Ni(OH)2/UGF as the positive electrode and graphite oxide as the negative electrode, which showed a high power density of 44.0 kW kg−1.23 Wang et al. synthesized a α-Ni(OH)2/carbon nanotube composite based on carbon nanotube paper in which the α-Ni(OH)2 nanosheets were vertically grown on individual carbon nanotube to form hierarchical nanowires. This novel composite brought a high specific capacitance of 1144 F g−1 at 0.5 A g−1.24
Vulcan XC-72 is the most widely used support for electrocatalysts in low temperature fuel cells because of its good conductivity, large percentage of mesopores, and excellent stability. Its specific surface area is ca. 250 m2, but metallic platinum or platinum–ruthenium particles remain less than 5 nm in diameter even when the metal loading reaches up to 60 wt%.27 To date, no related literature of XC-72 alone as an electrode material or as a support for a supercapacitor has been reported. Compared to graphene or CNTs, XC-72 is a commercial product and much cheaper, and it is even cheaper than the most common activated carbon used for commercial supercapacitors. In this study, XC-72 is used as a support to synthesize a Ni(OH)2/XC-72 composite in an ethylene glycol (EG) medium, in which EG is used both as solvent and as a ligand for obtaining smaller Ni(OH)2 particles. Furthermore, a more effective microwave-assisted heating method is adopted to synthesize the Ni(OH)2/XC-72 composite in order to achieve smaller Ni(OH)2 particles with a more uniform distribution than the produced by commonly heating methods. Eventually, carbon nanotubes (CNTs) are added during the process of electrode slurry preparation in order to construct a 3-D conductive network that facilitates electron transfer and enhances the electrochemical stability of the composite.
Electrochemical performance was evaluated by cyclic voltammetry (CV), galvanostatic charge–discharge, and electrochemical impedance spectroscopy (EIS) using a 302 N AutoLab Potentiostat (Metrohm, Holland) in a three-electrode system. Nickel mesh with a dimension of 10 × 10 was cleaned before being used as a substrate for the working electrode, Hg/HgO electrode was used as the reference electrode, a platinum coil was used as the counter electrode and the electrolyte was 6 M KOH. A 1000 cycle life test was carried out with a CV scan rate of 100 mV s−1.
To prepare a working electrode, active materials (60 wt% Ni(OH)2/XC-72 composite), conductive agent, and binder (60 wt% polytetrafluoroethylene emulsion) were mixed with a mass ratio of 85:10:5 in ethanol, then ultrasonically stirred for 20 min before being dropped onto a nickel foam and dried in a 70 °C vacuum oven for 6 h. The working electrode was then pressed under 10 MPa for 1 min. The areal density of the active materials is 1 mg cm−2. For the conductive agent, both acetylene black (AB) and carbon nanotubes (CNTs, Chengdu Organic Chemicals Company, China) were investigated. The obtained electrode is denoted as XC-CNT or XC-AB.
The specific capacitance of the synthesized composite can be calculated from the CV curve using the following equation,
C = ∫IdV/2mvΔV | (1) |
Moreover, the specific capacitance can also be calculated from a galvanostatic charge–discharge curve based on the equation below,
C = IΔt/mΔV | (2) |
The FTIR spectra of XC-72 and the Ni(OH)2/XC-72 composite are shown in Fig. 2(a). The broad band in the 3000–3700 cm−1 range is assigned to the O–H stretching vibration of adsorbed water molecules and O–H bond in Ni(OH)2. The peaks at 1636, 1400 and 1092 cm−1 correspond to CC, C–OH and C–O–C vibration frequencies, respectively. Other absorption bands below 800 cm−1 are associated with metal–oxygen (M–O) stretching and bending modes.28 In the FTIR spectrum of the Ni(OH)2/XC-72 composite, the intensity of the band at 3490 cm−1 due to O–H stretching vibration is strengthened, and new bands at 606 and 486 cm−1 appear, which can be attributed to Ni–O–H bending and Ni–O stretching vibrations, respectively.29 Therefore, the characteristic Ni–O and O–H bands of Ni(OH)2 are all observed30 and it can be deduced that the Ni(OH)2/XC-72 composite is formed.
Raman spectroscopy was employed to investigate the vibrational properties of XC-72 and Ni(OH)2/XC-72 composite. The Raman spectra in Fig. 2(b) clearly show that a D band around 1351 cm−1 and a G band around 1590 cm−1 are present in carbon black and the composite, which correspond to the breathing modes of rings or k-point photons of A1g symmetry and the E2g vibrational mode of sp2-bonded carbon atoms, respectively.31 Moreover, the intensity ratio of the D and G bands (ID/IG) can be used as a useful measurement of the degree of disorder and average size of the sp2 domains of the graphite materials. From the spectra of XC-72 and Ni(OH)2/XC-72 composite, the IG/IG ratio of Ni(OH)2/XC-72 (0.87) is slightly higher than that of XC-72 (0.86), which is probably due to the presence of more unpaired defects and decrease in the average size of the sp2 graphitic domains caused by the incorporation of Ni(OH)2.32
Fig. 3(a) implies that Ni(OH)2 nanoparticles are well distributed onto the surface of XC-72 nanospheres. XC-72 has a large specific surface area and rich mesopores, which are regarded to be an excellent support for fuel cell catalysts because supported Ni(OH)2 particle sizes are supposed to be small. Fig. 3(b) demonstrates that the electrode with added CNTs has a unique nanostructure in which CNTs are cross-linked with each other to form a 3-D conductive network together with Ni(OH)2 and XC-72 nanoparticles. This unique structure facilitates electron transfer and helps electrolyte ions to gain access to active sites, thus reducing equivalent series resistance and transfer resistance.
Typical HRTEM images of the Ni(OH)2/XC-72 composite are shown in Fig. 4(a) and (b). Ni(OH)2 is uniformly distributed onto the surface of XC-72 nanospheres. From Fig. 4(c) and (d), it is observed that Ni(OH)2 and XC-72 nanospheres are connected by CNTs, similar to what is implied in Fig. 3. HRTEM images of XC-72 nanospheres with diameters of 30 nm and CNTs are shown in Fig. 4(e) and (f), respectively. This unique 3-D conductive network is expected to facilitate ion and electron migration, thus reducing charging/discharging impedance and leading to better electrochemical performance.
Fig. 4 HRTEM images of Ni(OH)2/XC-72 composite (a and b), Ni(OH)2/XC-72 electrode slurry with CNTs as a conducting agent (c and d), XC-72 (e) and CNTs (f). |
From the SEM images in Fig. 5(a), it can be discerned that XC-72 spheres are uniformly covered by Ni(OH)2, while the CNTs exist in a network among the composite. The experimentally observed structure of Ni(OH)2/XC-72/CNTs is in good agreement with the schematic one. After 1000 CV cycles with a scan rate of 100 mV s−1, Ni(OH)2 particles can be observed to agglomerate slightly, but the composite still maintains an effective 3-D conductive network with Ni(OH)2 and XC-72 particles connected by CNTs, indicating the stability of the composite is good.
Fig. 5 SEM images of 60 wt% Ni(OH)2/XC-72/CNTs electrode before (a) and after (b) 1000-cycle CV life test. |
Fig. 6 Electrochemical performance of XC-AB and XC-CNT electrodes. CV curves at a scan rate of 2 mV s−1 (a) and typical discharge curves at a current density of 1 A g−1 (b). |
Electrode | Current density, A g−1 | |||
---|---|---|---|---|
1 | 2 | 5 | 10 | |
XC-AB | 951 | 873 | 715 | 645 |
XC-CNT | 1560 | 1205 | 969 | 840 |
From Table 1, the capacitances with CNTs as the conductive agent are all better than those with AB as the conductive agent while discharging at the same current density; with an increase in current density, the capacitance decreases with either CNTs or AB as conductive agent. Therefore, we can draw the conclusion that for the 60 wt% Ni(OH)2/XC-72 composite, CNTs are a better conductive agent than AB because of its lower resistance, easier diffusion and transport of electrolyte ions, and more affordable electrochemical active sites involved in the charge–discharge Faradaic reaction. Then, XC-CNT is investigated by CV with various scan rates and galvanostatic charge–discharge at different current densities.
Fig. 7 shows the electrochemical performance of XC-CNT. CV curves of XC-CNT at various scan rates are shown in Fig. 7(a). The pair of redox peaks at 0.41 V and 0.25 V indicate that the capacitance characteristics are mainly governed by Faradaic reactions, which is very distinct from a rectangular shape, and that is produced by a mechanism in which electrical energy is stored via a static double electric layer. The two anodic peaks occurred around 0.36 V and 0.41 V (vs. Hg/HgO) at a scan rate of 2 mV s−1. They originate from the two oxidation reactions of α-Ni(OH)2 and β-Ni(OH)2 to NiOOH. The cathodic peak occurring at around 0.25 V (vs. Hg/HgO) should correspond to the reduction reaction of NiOOH to Ni(OH)2. With an increase in scan rate, anodic and cathodic peak potentials are almost unchanged and the Ia/Ib ratio is closed to 1, indicating that the electrochemical redox reaction Ni(OH)2 + OH− ↔ NiOOH + H2O + e− is reversible. The specific capacitances are calculated to be 1296, 1112 and 952 F g−1, corresponding to CV scan rates of 2, 5, and 10 mV s−1 based on eqn (1).
Fig. 7 CV curves of XC-CNT at different scan rates (a). Galvanostatic charge–discharge curves of XC-CNT at various current densities (b). Discharge curves of XC-CNT at various current densities (c). |
Galvanostatic charge–discharge curves of XC-CNT at various current densities with a potential window between 0.15 and 0.55 V are shown in Fig. 7(b), and the discharging curves are separately displayed in Fig. 7(c) for a more direct evaluation of the specific capacitance. The specific capacitances are calculated to be 1560, 1205, 969, and 840 F g−1 at discharging current densities of 1, 2, 5, and 10 A g−1 according to eqn (2), respectively. The reproducibility experiments for XC-CNT are shown in Fig. S2–S4.† Three batches of samples were prepared under the same parameters, and the standard deviation (SD) and relative standard deviation (RSD) are calculated via the specific capacitances at a scan rate of 2 mV s−1. The SD and RSD are 47.7 F g−1 and 3.7%, respectively, indicating that the reproducibility is acceptable. Furthermore, the results of this study are compared with previously reported data based on Ni(OH)2 and carbon materials. The details are shown in Table 2.
Composite | Ni(OH)2 content | Scan rate or current density | Specific capacitance, F g−1 | Reference |
---|---|---|---|---|
XC-CNT | 60 wt% | 1 A g−1 | 1560 | This work |
XC-CNT | 60 wt% | 2 mV s−1 | 1290 | This work |
Ni(OH)2/AC | 6 wt% | 2 mV s−1 | 314.5 | Ref. 4 |
α-Ni(OH)2/CNT | 66 wt% | 0.5 A g−1 | 1144 | Ref. 24 |
NiAl-LDH/GNS | 89.7 wt% | 1 A g−1 | 1255.8 | Ref. 28 |
Ni(OH)2/CNT | 98.2 wt% | 1 A g−1 | 720 | Ref. 36 |
Ni(OH)2/MWCNT | 10 wt% | 10 mV s−1 | 432 | Ref. 37 |
Cycling stability is crucial for a supercapacitor electrode material. 60 wt% Ni(OH)2/XC-72 was continuously tested over 1000 cycles with a CV scan rate of 100 mV s−1 and the specific capacitance was calculated at various points before its cycling stability was recorded, as shown in Fig. 8(a). The specific capacitance of the composite with CNTs as the conductive agent rapidly decreases to ca. 90% after the first 100 cycles, then it is relatively stable and shows 71% capacitance retention after 1000 CV cycles. Therefore, we can draw the conclusion that the 60 wt% Ni(OH)2/XC-72 composite demonstrates relatively good stability, mainly because carbon nanomaterial XC-72 possesses a relatively large specific surface area that assures Ni(OH)2 particles of smaller size are distributed uniformly. In addition, it has a large amount of mesopores and good conductivity, which reduces the transportation impedance of electrolyte ions and electrons during the charging and discharging processes. The composite with AB as the conductive agent shows 66.9% capacitance retention after 1000 CV cycles. Compared to different conductive agents, CNTs exhibit better specific capacitance performance and cycling stability than AB because CNTs is highly graphitized with better electron conductivity and much larger aspect ratio, which facilitates the redox reaction.
In the end, the stability decreases nearly 30% after 1000 CV cycles. What factors contribute to the performance decay? From the CV curves of the 60 wt% Ni(OH)2/XC-72 composite before and after a 1000-cycle CV test shown in Fig. 8(b), the anodic peaks change from being two peaks to one peak after the life test; moreover, the peak at 0.36 V disappeared, which means that α-Ni(OH)2 is totally converted to β-Ni(OH)2 during the cycling test. The β-Ni(OH)2 phase is more condensed and has less water in the interlayer space,38 thus leading to slower redox reactions and smaller current density. Furthermore, the cathodic current density decays more significantly than the anodic one, most likely because it is more difficult to diffuse H+ ions into the Ni(OH)2 bulk phase than it is to diffuse them out. Another reason originates from the gradual agglomeration of Ni(OH)2 particles during long charging/discharging durations in the electrical field. This is proved by the SEM images in Fig. 5.
EIS is a useful technique to probe reaction mechanisms in electrochemistry. For the 60 wt% Ni(OH)2/XC-72 composite with CNTs as the conductive agent, the experiment was carried out at an open circuit potential and potential amplitude of 5 mV in a frequency range of 105 Hz to 10−2 Hz. The Nyquist plots obtained before and after 1000 CV cycles are displayed in Fig. 8(c). After the cycling test, Rs increases from 0.5 to 1.2 Ω. This increase is mainly ascribed to changes in the active material properties such as the increment of Ni(OH)2 particle size and phase change as previously mentioned. Moreover, charge-transfer resistance increases because β-Ni(OH)2 has less interlayer water. However, Warburg resistance decreases slightly, perhaps because of the presence of bigger Ni(OH)2 particles with smaller diffusion resistances. The EIS result is in good agreement with the SEM images, CV curves, and cyclic stability test results.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra04301h |
This journal is © The Royal Society of Chemistry 2015 |