Montakan Suksomboona,
Patarachai Srimuka,
Atiweena Krittayavathananona,
Santamon Luanwuthia and
Montree Sawangphruk*ab
aNational Center of Excellence for Petroleum, Petrochemicals and Advance Material, Department of Chemical Engineering, Faculty of Engineering, Kasetsart University, Bangkok 10900, Thailand. E-mail: fengmrs@ku.ac.th; Fax: +66-2-561-4621; Tel: +66-2-942-8555
bCenter for Advanced Studies in Nanotechnology and Its Applications in Chemical, Food and Agricultural Industries, Kasetsart University, Bangkok 10900, Thailand
First published on 24th October 2014
Although alpha-cobalt hydroxide (α-Co(OH)2) with a layered double hydroxide (LDH) structure has been widely used as a supercapacitor electrode, the effect of an alkaline electrolyte on the charge storage performance of the α-Co(OH)2 has not yet been investigated. In this work, α-Co(OH)2 was electrodeposited on reduced graphene oxide-coated carbon fiber paper (rGO/CFP) using chronoamperometry at −0.5 V vs. Ag/AgCl. The effect of alkaline aqueous electrolytes on the performance of the α-Co(OH)2/rGO/CFP electrodes was then investigated by means of scanning and transmission electron microscopies, X-ray photoelectron and absorption spectroscopies, and electrochemical techniques. It was found that the concentrated alkaline electrolytes (i.e., 3–6 M [OH−]) can strip off and/or deform the porous structure of the α-Co(OH)2 deposited on rGO/CFP leading to poor charge storage capacity. 1 M [OH−] was found to be a suitable electrolyte concentration providing high specific capacitance (1096 F g−1 at 1.8 A g−1) without the deformation of the porous α-Co(OH)2 structure after testing. Morphological and electrochemical analyses of the α-Co(OH)2/rGO/CFP electrodes suggest that the effect of the alkaline electrolyte concentration plays a major role on the charge storage performance of α-Co(OH)2-based supercapacitors.
Nanomaterials play a vital role in improving the charge storage capacity of both EDLC capacitors and pseudocapacitors. Graphene has been widely used and recognized as the high-performance electrode of EDLC capacitors.5–7 Whilst, Co(OH)2 nanostructures, known as inexpensive, nontoxic, and electrochemically active pseudocapacitor materials, have ultrahigh charge storage capacity due to a large interlayer spacing between its adjacent layers of Co(OH)2.8,9 Anions can be moved in/out through the spacing of Co(OH)2. Co(OH)2 LDHs have been produced by various techniques such as hydrothermal, solvothermal, microwave, sonochemical, and electrodeposition techniques.10,11 Among them, the electrodeposition is of great interest since it is relatively simple, cheap, and scalable. Changing deposition variables such as electrolyte, deposition potential, temperature, and time can accurately control the surface microstructure of the deposits.12–14 Importantly, the active materials produced by the electrodeposition can be directly exposed to the electrolytes leading to high specific activity at very diluted loading content.15 Additionally, there is no need of other materials e.g. polyvinylidene difluoride (PVDF) binder and conductive carbon black typically used for the fabrication of energy storage electrodes. Note, the active materials can be buried by PVDF and carbon black particles leading to inactive surface area of the materials.
In addition, there is another important issue relating to the electrolyte used for the electrochemical evaluation of Co(OH)2 supercapacitor electrodes. Different concentrations of KOH (1–6 M) were previously used without explanation in details why KOH has been used and which suitable concentration of alkali should be used leading to difficulty in the performance comparison of Co(OH)2-based supercapacitors. In this work, the effect of two alkaline electrolytes i.e., KOH and NaOH with different concentrations (1–6 M) on the charge storage capacity and the morphological structure of α-Co(OH)2 LDH was systematically investigated. Interestingly, it turned out for the first time in this work that at high concentrations e.g. 3–6 M KOH and NaOH, the electrolytes are rather corrosive, which can strip off and/or deform the porous structure of the α-Co(OH)2 LDH electrodeposited on reduced graphene oxide (rGO)-coated carbon fiber paper (CFP).6,16 On the other hand, the α-Co(OH)2 electrode tested in 1 M NaOH exhibits high charge storage performance without morphology change.
NO3− + 7H2O + 8e− → NH4+ + 10OH− | (1) |
Co2+ + 2OH− → Co(OH)2 | (2) |
When a potential of −0.5 V vs. Ag/AgCl finely tuned was applied to the working electrode (rGO/CFP), nitrate ions in the electrolyte can be reduced on the cathodic surface to produce hydroxide ions. The generation of OH− at the cathode raises the local pH, resulting in the precipitation of Co(OH)2 LDH with the anion (NO3−) interlayer at the rGO/CFP surface. The colour of Co(OH)2 deposit is green, which is an optical appearance of α-Co(OH)2 (see Fig. S2a of ESI†).13
When the reactions (1) and (2) take places, α-Co(OH)2 films with high porosity are found on the rGO/CFP (Fig. 1). The pore size of air voids is clearly decreased when increasing deposition time. Large Co(OH)2 nanosheets with open pore structure are clearly seen at a deposition time of 10 min (Fig. 1d). At longer deposition time (e.g. 20 and 40 min), the morphology of the deposit becomes denser and the nanosheets found at 10 min disappears (see Fig. 1e–f and low-magnification SEM images in Fig. S3e–f of ESI†). The denser morphology may limit the electrolyte diffusion. On the other hand at short deposition time (1–5 min), the deposits did not completely cover the rGO/CFP, which may lead to low capacity retention (see Fig. S3a–c of ESI†). EDX of all as-electrodeposited samples was also carried out and the results showed that the deposits consist of Co, O, and C elements.
Fig. 2a and b show TEM images of rGO sheets and as-electrodeposited Co(OH)2 on rGO sheets. A few layers of transparent rGO sheets overlap each other forming different surface morphologies (Fig. 2a). After the electrodeposition process, two-dimension (2D) Co(OH)2 nanosheets are found on the rGO surfaces (Fig. 2b). The interfacial interaction between two layers is most possibly due to van der Waals force. The TEM result here is in good agreement with that of other LDHs (e.g. Ni(OH)2) previously reported.20 The EDX spectrum of Co(OH)2/rGO is shown in Fig. 2c for which C, O, and Co elements are observed. Note, the Cu element is found due to the Cu grid used for the preparation of the TEM specimen. Fig. 2d shows the selected area electron diffraction (SAED) pattern of the Co(OH)2 indicating the polycrystalline phase of Co(OH)2 with two separated rings.21
![]() | ||
Fig. 2 TEM images of (a) rGO and (b) as-electrodeposited α-Co(OH)2 on rGO sheets as well as (c) EDX and (d) selected area electron diffraction (SAED) pattern of α-Co(OH)2. |
Co(OH)2 deposit was scrapped off from the electrode using a sharp doctor blade (see Fig. S2b of ESI†). The powder XRD of as-scrapped Co(OH)2 is shown in Fig. 3a for which all planes i.e., (003), (006), (012), (015), (018), and (110) are of α-Co(OH)2.22 The α-Co(OH)2 with the nitrate anions in their spacing has a larger interlayer spacing (ca. 8.4 Å) than that of other phases e.g. β-Co(OH)2 (4.6 Å).22 As the result, the electrolytes may freely diffuse through the large spacing of α-Co(OH)2.
![]() | ||
Fig. 3 (a) XRD pattern, (b) wide scan XPS, (c) narrow scan XPS spectra, and (d) XAS spectra of as-electrodeposited α-Co(OH)2 compared with Co foil, and α-Co(OH)2. |
XPS was carried out to confirm the structure of the as-electrodeposited α-Co(OH)2 for which all XPS spectra were referenced to the aliphatic carbon at a binding energy of 285.0 eV. Wide-scan XPS spectra of rGO and as-electrodeposited α-Co(OH)2/rGO/CFP are shown in Fig. 3b for which Co2p, C1s, and O1s are found on the samples confirming the EDX results. Fig. 3c shows a narrow-scan XPS spectrum of Co2p of the as-electrodeposited Co(OH)2. Co2p spectrum exhibits a spin–orbit splitting into Co2p1/2 and Co2p3/2 components with a spitting energy of 16.2 eV. Both components can give the same quantitative information; thus, in this work the higher intensity Co2p3/2 bands were only curve–fitted and the results in Fig. 3c show that a major peak at a binding energy of 780.4 eV is due to Co2+ of Co(OH)2 and minor peaks are due to shake-up satellite effect. This result is in good agreement with other previous work.23,24
The oxidation number of Co in the α-Co(OH)2 deposit was then evaluated using XAS. The XAS characteristic of the as-electrodeposited α-Co(OH)2 is in good agreement with that of the α-Co(OH)2 standard, not Co metal foil (see Fig. 3d). As the result, it can be concluded that the oxidation number of Co in the α-Co(OH)2 structure is 2+. Thermal stability of the as-prepared α-Co(OH)2 was also studied. Fig. S4 of ESI† shows the TGA and its derivative pattern of the as-scrapped α-Co(OH)2 carried out in the temperature range 30–350 °C in O2 gas for which its weight loss undergoes two steps. The first one at 86 °C is due to the dehydration of adsorbed water inside the interlayers.25 The second weight loss at 211 °C is owing to the decomposition of Co(OH)2 providing Co3O4.26
Co(OH)2 + OH− ⇆ CoOOH + H2O + e− | (3) |
There is no subsequent oxidation reaction leading to very stable phase CoO2 as previously reported.27 In addition to the sharp redox peaks observed, the peak separation is about 0.1–0.2 V indicating that the as-electrodeposited α-Co(OH)2 behaves as the redox-based electrode.28 It is necessary to note here that there is no sharp redox peak for two well-known pseudocapacitor materials e.g. MnO2 and RuO2, which store charges via the surface redox reactions.29,30
For the effect of the electrolytes and their concentrations, CVs in Fig. 4a and b show that the current densities of the α-Co(OH)2/rGO/CFP tested in both 1 M NaOH and KOH are much higher than those evaluated in 2–6 M NaOH and KOH. A potential range achieved in this work is also wider than that (−0.1 to 0.45 V) previously reported.31 At the potential over 0.55 V vs. Ag/AgCl, the oxygen evolution reaction occurs. Further observation, there is no significant difference in the current density obtained from 1 M NaOH and 1 M KOH. These results are somewhat interesting since other previous works mainly used KOH at high concentrations e.g. 6 M KOH.32,33 At high concentrations, the stripping (dissolution) of the deposits was also observed for which the colour of the electrolytes turned from no colour to light brown during the charging/discharging especially at fast charge/discharge rates. For further confirmation, the electrochemical impedance spectroscopy was also carried out. Nyquist plots, measured in NaOH and KOH using a sinusoidal signal of 40 mV over the frequency range from 100 kHz to 1 mHz in Fig. 4c and d, show that NaOH and KOH at high concentrations (3–6 M) are rather corrosive since clear semi-circles at high frequency (the lower left portion of the plots) relating to the stripping of metal hydroxide are observed.
SEM images of all electrodes after tested in 1–6 M NaOH and KOH are shown in Fig. S5 and S6,† respectively. Clearly, the strong basic electrolytes (3–6 M NaOH and KOH) can deform the porous structure of the α-Co(OH)2 becoming denser, which is not good for the electrolyte diffusion. NaOH is also commercially 2-fold cheaper than KOH. As the results, 1 M NaOH was then used as the electrolyte in the electrochemical evaluation hereafter.
The effect of the electrodeposition time to the areal-based charge storage performance of α-Co(OH)2/rGO/CFP was investigated. The weights of α-Co(OH)2 on 1 cm2 rGO/CFP electrodes produced for 1, 2, 5, 10, 20, and 40 min are ca. 0.1, 0.2, 0.8, 2.2, 2.7, and 4.7 mg, respectively. The capacitance (C) was calculated from the CV by following eqn (4) where I is a discharging current (A), t is discharging time (s), and ΔV is the working potential range (ca. 0.8 V);
![]() | (4) |
The areal capacitance (Ca) can be calculated by dividing C with the geometrical surface area of the electrode.34 The areal capacity (mA h cm−2) is calculated by multiplying Ca (F cm−2) with ΔV/3.6.34 Ca and areal capacity values of all as-fabricated electrodes at different loading contents vs. a square root of scan rates are shown in Fig. 5a for which these values are decreased as a function of square root of the scan rate indicating that the charge storage mechanism is under diffusion limit. This is a characteristic of redox-based energy storages.1,2,35,36 The 2.2 mg α-Co(OH)2/rGO/CFP produced for 10 min can provide the Ca of 2.2 F cm−2 and the areal capacity of 6.2 mA h cm−2 at a scan rate of 2 mV s−1. These two values are much higher than those produced at other deposition time and those of other materials previously reported.34,37–39 At high loading content of α-Co(OH)2, the Ca and areal capacity values are limited by the poor conductivity of α-Co(OH)2.37 At low loading content of α-Co(OH)2, EDLC, contributed by the rGO, plays a major role but faradaic current is rather low leading to low Ca and areal capacity. This is because α-Co(OH)2 did not cover the whole rGO/CFP surface (see Fig. S3a–c of ESI†).
To further investigate the charge storage mechanism, the as-scrapped α-Co(OH)2 was mixed with conductive carbon black and PVDF binder in N-methyl-2-pyrrolidone solvent at the weight ratio of 80:
10
:
10 wt% (α-Co(OH)2
:
PVDF
:
carbon black). The mixture was then casted on the rGO/CFP electrode and vacuum dried at 60 °C overnight. The CV results, carried out in 1 M NaOH at a scan rate of 1 mV s−1 (Fig. 5b), show that increasing the mass loading content of α-Co(OH)2 from 1.5 mg to 3.0 mg provides no significant difference in the peak position and area as well as faradaic current related to the redox reaction (3) of α-Co(OH)2. However, the Ca of the as-fabricated electrodes is rather low about 0.16 F cm−2 when compared with 2.2 F cm−2 achieved by the electrodeposition method.
The specific capacitance of the 10 min α-Co(OH)2/rGO/CFP finely tuned was also investigated by the galvanostatic charge/discharge at different applied specific currents (1.8–9.1 A g−1) as shown in Fig. 6a. For the charging process, a linear variation of the time dependence at the potential (below about 0.08 V vs. Ag/AgCl) indicates pure EDLC behavior. Whilst, a slope variation of the time dependence at the potential range of ca. 0.08–0.20 V vs. Ag/AgCl exhibits a typical pseudocapacitance characteristic due to the oxidation reaction of Co(OH)2 (see forward reaction 3). For the discharging process, the potential of ca. −0.1 V vs. Ag/AgCl indicates the reduction reaction of CoOOH (see backward reaction (3)). The specific C values of the electrode calculated by following equation; C = IΔt/ΔV where I is applied specific current, Δt is discharging time, and ΔV is working potential excluding iR drop, are 1096, 1076, 1060, 1044, and 1022 F g−1 at 1.8, 2.7, 3.6, 4.5 and 6.8 A g−1, respectively. For the relative comparison to other previous reports related to the Co(OH)2 supercapacitor electrodes, the specific capacitance of the 10 min α-Co(OH)2/rGO/CFP obtained in this work was also calculated and listed in Table 1. In addition, the capacity retention of the as-fabricated electrode was investigated by the galvanostatic charge/discharge in 1 M NaOH at 1.8 A g−1 over 5000 cycles (see Fig. S7 of the ESI†). The capacity retention is still rather high about 98% over 5000 consecutive cycles (see Fig. 6b).
![]() | ||
Fig. 6 (a) Charge/discharge curves of 10 min α-Co(OH)2/rGO/CFP in 1 M NaOH at different applied specific currents and (b) its capacity retention over 5000 cycles. |
Materials | Electrolytes | Testing methods and conditions | Specific capacitance (F g−1) | Ref. |
---|---|---|---|---|
a C/D is a galvanostatic charge/discharge.b CV is a cyclic voltammetry. | ||||
Co(OH)2 nanosheets | 6 M KOH | C/Da at 0.83 A g−1 | 1083 | 32 |
Co(OH)2 flakes on 3D multi-layered graphene foam | 1 M KOH | C/D at 9.09 A g−1 | 1030 | 14 |
Co(OH)2 nanoparticles | 2 M KOH | C/D at 1 A g−1 | 474 | 21 |
Co(OH)2 nanosheets | 1 M KOH | C/D at 1 A g−1 | 840 | 40 |
Whisker-like Co(OH)2 | 2 M KOH | C/D at ca. 1.3 A g−1 | 325 | 41 |
Co(OH)2 nanosheets | 1 M KOH | C/D at 2 A g−1 | 651 | 42 |
Flower-like Co(OH)2 | 3 M KOH | C/D at 1.3 A g−1 | 434 | 43 |
Co(OH)2 nanocone | 2 M KOH | C/D at 2 A g−1 | 562 | 44 |
Co(OH)2 flakes | 1 M KOH | C/D at 2 A g−1 | 693.8 | 45 |
Co(OH)2//Co(OH)2 device | 1 M KOH | CV at 5 mV s−1 | 44 | 46 |
Co(OH)2 nanosheets | 1 M Na2SO4 | CV at 8 mV s−1 | 141 | 47 |
Co(OH)2 nanowires | 6 M KOH | C/D at 0.5 A g−1 | 358 | 33 |
Flower-like Co(OH)2 | 3 M KOH | C/D at 0.5 μA g−1 | 170 | 48 |
Co(OH)2 nanocone arrays | 2 M KOH | C/D at 2 A g−1 | 562 | 44 |
Mesoporous Co(OH)2 | 1 M KOH | CVb at 5 mV s−1 | 246.7 | 49 |
α-Co(OH)2 nanosheets/rGO/CFP | 1 M NaOH | C/D at 1.8 A g−1 | 1096 | This work |
Footnote |
† Electronic supplementary information (ESI) available: Morphological characterizations, TGA, electrochemical evaluation, can be seen in ESI. See DOI: 10.1039/c4ra11727a |
This journal is © The Royal Society of Chemistry 2014 |