Yang Luabc,
Hailong Yanab,
Kangwen Qiuab,
Jinbing Chengab,
Weixiao Wangab,
Xianming Liud,
Chengchun Tang*c,
Jang-Kyo Kime and
Yongsong Luo*ab
aSchool of Physics and Electronic Engineering, Xinyang Normal University, Xinyang 464000, P. R. China. E-mail: ysluo@xynu.edu.cn
bKey Laboratory of Advanced Micro/Nano Functional Materials, Xinyang Normal University, Xinyang 464000, P. R. China
cSchool of Material Science and Engineering, Hebei University of Technology, Tianjin 300130, P. R. China. E-mail: tangcc@hebut.edu.cn
dCollege of Chemistry and Chemical Engineering, Luoyang Normal University, Luoyang 471022, P. R. China
eDepartment of Mechanical and Aerospace Engineering, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, P. R. China
First published on 8th January 2015
We report a novel, low-cost strategy to synthesize copper oxide (CuO) nanostructures as high-performance supercapacitor electrodes using an alkaline solution oxidation method. The structure, morphological features, surface area and pore size distribution of the products are tuned using different types of surfactants. The CuO electrode obtained from sodium dodecyl sulfate (SDS) presents the best electrochemical performance due to the synergies arising from the large surface area and pore volume created by the ultrathin nanoleaves constituting the flower-shape nanostructure. The electrode delivers a remarkable specific capacitance of 520 F g−1 at 1 A g−1 and a high rate capacitance of 405 F g−1 at 60 A g−1 with more than 95% Coulombic efficiency after 3500 cycles.
Recently, much effort has been devoted to exploring transition-metal oxides (TMOs) including RuO2,12 MnO2,13 NiCo2O4,14 and NiO15 because their higher SCs than carbonaceous materials and better cyclic stability than conducting polymers.16–18 However, experimental SC values are often much lower than the theoretical predictions, because only limited external surface of the electrodes takes part in the electrochemical reactions while the electrode internal region could barely contribute to the capacitive performance.19,20 It follows that a large specific surface area is one of the most important criteria of TMO electrode materials which enables efficient redox reactions through a large electrode–electrolyte interface, so as to maintain a large capacity by pseudocapacitors.
As an important low-cost TMO with a narrow band gap of ∼1.2 eV, copper oxide (CuO) has been extensively exploited for widespread applications in sensors, lithium-ion batteries, photocatalysts, solar cell devices, and so on.21–24 Among these potential applications, CuO has been considered a very promising candidate to replace commercial RuO2-based pseudocapacitive materials due to its low-cost, abundance and non-toxicity. More importantly, CuO can deliver a high theoretical pseudocapacitance of ∼1800 F g−1, which is close to that of the widely studied RuO2·nH2O (∼2200 F g−1).25 Significant research efforts have been directed towards the design and development of CuO as a supercapacitor electrode, Table 1 gives a brief summary of recent developments.26–30 Several different CuO nanostructures have been obtained using various strategies, including anodization, solvothermal growth, wet-chemical process, water bath method and potentio-dynamic electrodeposition. For example, CuO nanosheet arrays were electrochemically synthesized by anodization of a piece of copper foam in KOH aqueous solution.26 CuO nanobelts were synthesized using a solvothermal growth method,27 whereas cauliflower-like CuO was grown directly on a stainless steel substrate by potentio-dynamic mode of electrodeposition.30 These methods often require high temperatures, high pressures, sophisticated instrumentation and complicated experimental procedures, so that their mass production and the application of these CuO nanostructures remain a big challenge.
Method of synthesis | Morphology | Cs (F g−1) | Discharge current density | Electrolyte | Stability | Ref. |
---|---|---|---|---|---|---|
Electrospun | Nanowire | 620 | 5 mA cm−2 | 6 M KOH | 10% loss after 2000 cycles | 25 |
Anodization | Nanosheet | 212 | 0.41 mA mg−1 | 6 M KOH | 15% loss after 850 cycles | 26 |
Solvothermal method | Nanobelt | 392 | 2 A g−1 | 1 M KOH | 10–15% loss after 5000 cycles | 27 |
Wet-chemical process | Gear-like nanostructure | 348 | 1 A g−1 | 0.1 M KOH | 12.1% loss after 2000 cycles | 28 |
Water bath method | Nanoflower | 130 | 1 A g−1 | 6 M KOH | 29.8% loss after 7000 cycles | 29 |
Potentiodynamic electrodeposition | Cauliflower-like nanostructure | 162 | 2 mA cm−1 | 1 M Na2SO4 | 19% loss after 2000 cycles | 30 |
Alkaline solution oxidation method | Nanoflower | 520 | 1 A g−1 | 1 M KOH | 5.5% loss after 5000 cycles | This work |
With the above difficulties in mind, we have developed a facile, high-efficiency and low-cost synthetic route to prepare novel CuO nanostructures. Herein, we present an oxidation method in alkaline solution for large-scale fabrication of porous CuO nanostructures along with their growth mechanisms. The nature of crystalline structure, their shapes and pore size distributions are tuned using different surfactants and by controlling the reaction time. A strong dependence of specific charge storage capacities on the morphology and structure of CuO electrodes is demonstrated. The pseudocapacitor electrodes made from flower-shaped CuO deliver the best electrochemical performance including the highest SCs among three different nanostructures.
![]() | (1) |
Cs = IΔt/ΔVm | (2) |
Cu + 4OH− + (NH4)2S2O8 → Cu(OH)2 + 2SO42− + 2NH3↑ + 2H2O | (3) |
Cu(OH)2 + 2OH− → [Cu(OH)4]2− | (4) |
[Cu(OH)4]2− → CuO + H2O + 2OH− | (5) |
Under different reaction conditions, different nanostructures were synthesized, namely CuO flowers, CuO spheres and CuO corals (Fig. 1b–d).
The XRD spectrum of the CuO flowers is shown in Fig. 2a. All the diffraction peaks are indexed according to the standard monoclinic structure of CuO crystal (JCPDS file no. 45-0937). No peaks of impurities, such as copper hydroxide or cuprous oxides, were found, suggesting high purity of the material. The EDX analysis showed only copper and oxygen (see Fig. 2b), confirming the XRD result. Further evidence of the composition and purity of the material was obtained by XPS of the core-level Cu 2p, as shown in Fig. 2c. The peaks at 954.1 and 934.1 eV correspond to the core-level Cu 2p1/2 and Cu 2p3/2 transitions of copper, respectively.31 Moreover, the presence of satellite peaks (with asterisks in Fig. 2c) at higher binding energy sides, whose intensities are comparable to those of the core-level Cu 2p peaks, further demonstrates that the flower-like structure consisted of pure CuO crystals.32 Additional information was obtained by Raman spectroscopy in the range of 200–800 cm−1, as shown in Fig. 2d. The broad peak with a relatively high intensity at 295 cm−1 is assigned to Ag band, while the two peaks at 342.8 and 628.5 cm−1 are assigned to 2Bg. The significant intensities of these peaks indicate a single phase and high crystallinity of CuO flowers, in good agreement with the previous reports.33,34 The thermogravimetric analysis shows that there was a very small weight loss, ∼2 wt%, below 200 °C, which can be attributed to the evaporation of adsorbed moisture (Fig. S1†). There was virtually no weight loss between 200 and 600 °C, indicating the absence of ionic liquids in the material. This further confirms the high purity of the as-prepared CuO flowers.
![]() | ||
Fig. 2 XRD pattern (a), EDX spectrum (b), Cu 2p core-level XPS spectrum (c), and Raman spectrum (d) of CuO flowers. |
The morphologies and structures of CuO flowers were examined by SEM and TEM. Fig. 3a–c clearly present flower-like CuO crystals with a high yield, mostly within the size distribution ranging 1.2–1.8 μm and with an average of 1.5 μm (Fig. S2†). The single flower was composed of many nano-sized leaves of 10–20 nm in thickness arising from the flower center. The TEM images (Fig. 3d and e) suggest that the flower core was densely filled with leaves whereas the flower edge had a porous structure due to the gaps between the flower leaves. The lattice fringe spacing of 0.23 nm marked in Fig. 3f corresponds to the (111) plane of monoclinic CuO. The selected area electron diffraction (SAED) pattern (Fig. 3g) can be indexed to a pure monoclinic phase. The appearance of periodic diffraction spots indicates that these nanostructures were self-assembled into highly oriented aggregates and diffracted as a single crystal.
![]() | ||
Fig. 3 (a–c) SEM images; (d and e) TEM images of CuO flowers taken at different magnifications; (f) HRTEM image of the fringe part of flower, and (g) the corresponding SAED result. |
CuO corals and spheres were prepared using a different surfactant or without surfactant, and Fig. 4 shows their morphologies. The corals had an average length of 2.7 μm, approximate widths of 200–450 nm, and thicknesses of 300–550 nm (Fig. 4a and c). The simpler geometry and the elongated shape may imply that the absence of surfactant led to faster growth along the length of nanostructures, relative to the growth into the other geometries. Meanwhile, spherical particles of 0.8–1.5 μm in diameter with a relatively clean surface were created when the PVA surfactant was added into the reaction solution (Fig. 4b and d). When PVA was used as a template, the aqueous solution of Cu2+ and PVA led to the formation of nucleation seeds to act as an initial nucleus for the growth of particles. When the particle reached a critical dimension, PVA was adsorbed onto the small particles through –OH bonds to serve as a template for the formation of spheres. The rough surface of the solid CuO spheres indicates that they consisted of many primary nanoparticles.
The largely different morphologies due to the use of different types of surfactants need further explanations. It is well known that the surfactant in a solution aggregates into micelles when its concentration exceeds a critical micelle concentration (CMC). Depending on the concentration and surrounding medium of the surfactant, the micelles display different shapes (Fig. 5). During the reaction in the presence of SDS, Cu(OH)4− could be dissociated to Cu2+ and OH−, and SDS coordinates with Cu2+ to form Cu(SDS)2. In the subsequent reaction, Cu(SDS)2 acts as surfactant. The concentration of 0.07 M SDS in water is considered above the CMC, so that micelles are organized in the system where the micelles play the role of “seeds”35 to form CuO clusters on them which further grow to become 3D hierarchical flowers.
To determine the porosity of CuO nanostructures, N2 adsorption–desorption isothermal analyses were performed. According to the IUPAC classification of hysteresis loops,36 Fig. 6a displays the type IV isotherms with type H3 hysteresis loops. All three nanostructures did not exhibit limited adsorption at relative pressures from 0 to 1, proving the presence of typical hierarchical porosities.37,38 An increase in slope at about 0.4, especially for CuO flowers, corresponds to capillary condensation, typical of mesoporous materials, while a further increase in adsorbed volume at higher relative pressures indicates inter-particle porosity.39 The specific surface areas of the nanostructures calculated by the Brunauer–Emmett–Teller (BET) model are given in Table 2, showing increasingly larger surface areas and pore volumes in the ascending order of spheres, corals and flowers. The pore size distribution shown in Fig. 6b indicates that both the flowers and corals had both small (<10 nm) and large mesopores (>10 nm). On the contrary, the spheres showed only small mesopores (∼4.75 nm). In view of the dependence of pore size distributions on nanostructures, it can be summarized that the structural evolution is an important factor influencing the interconnected pores inside the particles and the void spaces between them.40
![]() | ||
Fig. 6 (a) Nitrogen adsorption and desorption isotherms measured at 77 K for CuO nanostructures; and (b) corresponding Barrett–Joyner–Halenda (BJH) pore size distribution curves. |
Electrodes | Surface area/m2 g−1 | Pore volume/cm3 g−1 | Rs/Ω | Rct/Ω |
---|---|---|---|---|
CuO flowers | 119.6 | 0.413 | 0.94 | 0.54 |
CuO corals | 73.5 | 0.351 | 0.99 | 1.17 |
CuO spheres | 36.1 | 0.233 | 1.31 | 5.94 |
The performance of these CuO nanostructures as supercapacitor electrodes was characterized by cyclic voltammetry (CV) measurements in KOH aqueous solution at room temperature (Fig. 7). All the CV curves showed a pair of redox peaks, very much different from that of an EDLC, which is normally in the form of an ideal rectangle. The well-defined redox peaks in the potential range of 0–0.6 V arose mainly from the reversible Faradaic redox reactions of Cu+ and Cu2+ species associated with the OH− ions.41 In addition, as the scan rate increased from 5 to 100 mV s−1, the current density increased while the CV curve shapes changed little. Especially for CuO flower electrode, the anodic peak potential at about 0.56 V shifted towards the anodic direction whereas the cathodic peak potential at about 0.22 V shifted towards the cathodic direction. Compared with the CuO coral and sphere counterparts, the CuO flower electrode maintained distinct redox peaks at high scan rates (Fig. 7, S3a and b†). This observation might be due to the bigger size pores in flowers than the other nanostructures, which promoted ion diffusion. The electrode made from CuO flowers exhibited the highest current density (Fig. 7) making them most promising for pseudocapacitor electrodes among the three nanostructures. The smaller active specific surface areas and pore volumes of the CuO coral and sphere electrodes presented inferior cyclic voltammetry performance.
![]() | ||
Fig. 7 Cyclic voltammetry (CV) curves of CuO nanostructures in 1 M KOH aqueous solution over a potential range from 0 to 0.6 V at a scan rate of (a) 5, (b) 20, (c) 50 and (d) 100 mV s−1. |
Fig. 8a shows a Nyquist plot of CuO-based electrode materials measured in 1 M KOH in the frequency range from 0.01 Hz to 100 kHz. The intercept at the real part of high frequency represents the combination (Rs) of ionic resistance of electrolyte, intrinsic resistance of substrate and contact resistance at the active material/current collector interface.42 The charge transfer resistance (Rct) is caused by the Faradic reactions and the double-layer capacitance on the grain surface.43 The electrical equivalent circuit used for fitting the impedance spectra is shown in the inset of Fig. 8a and Table 2 presents these parameters fitted by Zview software. Both the Rs and Rct values of the CuO flower electrode were lower than those of the corals and spheres. At low frequencies, the flower electrode had a linear line with a steeper slope and shorter length than the other nanostructures, indicating a lower Warburg impedance (Zw) encountered during the ion transportation in the aqueous electrolyte. Moreover, ion diffusion coefficient of all the morphologies is calculated according to the following equation:44
D = R2T2/2A2n4F4C2σ2 | (6) |
Zw = Rs + Rct + σω−1/2 | (7) |
Fig. 8b shows a summary of the CV curves of the flower electrode taken from Fig. 7 with a typical pseudocapacitive behaviour and distinct redox peaks even at a high potential scan rate of 100 mV s−1. Similar CV curves for the other electrode materials are given in Fig. S3a and b.† Fig. S4† investigates the relationship between the SCs of the flower electrode or capacity retention and scan rate, exposing excellent specific capacitance of 615, 595, 563, 549, 525, and 500 F g−1 at scan rates of 5, 10, 20, 30, 50 and 100 mV s−1, respectively. This shows that about 81.3% of the capacitance is still retained when the scan rate increased from 5 to 100 mV s−1. In pseudocapacitive materials, we can conclude from the scan rate (v) dependence of voltammetric current (I) that the capacitance originates from surface redox reactions or from bulk diffusion.25 That is to say, the I ∝ v for surface redox reactions and I ∝ for semi-infinite bulk diffusion.45 In the case of CuO flower, the relationship between peak current and scan rate is plotted in Fig. S5a,† the peak current varied linearly with scan rate. As far as we know, ion movement is limited only to the surfaces of the electrode material at higher voltammetric scan rates (>30 mV s−1). Therefore, the high SCs of flower electrode resulted from surface redox reaction, and remained practically constant with scan rate. Meanwhile, we also surveyed the other two electrodes (Fig. S5b and c†). From both the linear relationship between peak current and scan rate as well as the symmetry of anodic/cathodic peak, CuO flower electrode exhibits prominent electrochemical reversibility.
The SCs of the electrode materials were measured in the voltage range of 0–0.6 V (vs. Ag/AgCl). Typical galvanostatic discharge profiles of the three electrodes measured at different current densities are shown in Fig. 8c, S3c and d.† The potential drop in the discharge curves is generally caused by internal resistance and incomplete Faradic reactions in the electrode.39 The SCs of the electrodes calculated according to eqn (1) are plotted as a function of current density, as shown in Fig. 8d. The flower electrode delivered the highest SCs among the three at all current densities studied: e.g. 520 and 405 F g−1 at 1 and 60 A g−1, respectively. These values are at least 30% higher than those of the coral electrode or 50% higher than those of the sphere electrode at the same discharge current densities.
High rate capability and long cyclic life are among the vital indicators for practical application of a pseudocapacitor. The CuO flower electrode was subjected to continuous charging and discharging for 5000 cycles at increasingly higher current densities, as shown in Fig. 9a. The SC progressively increased in the first 500 cycles at 1 A g−1 before reaching a maximum of 529 F g−1, as a result of the gradual activation of the electrode material. It is worth noting that even after 3000 cycles, the electrode retained a remarkable SC of 405 F g−1 at a high current density of 60 A g−1 while it showed 495 F g−1 after the current density was returned to 1 A g−1. During all these charge/discharge cycles at different rates, the Coulombic efficiency was steadily kept higher than 99%, confirming exceptional cyclic stability. Several synergies arising from the unique morphological features and the hierarchical pore structure of the CuO flowers gave rise to such remarkable electrochemical performance. As the basis of the flower-shaped structure, the well-separated primary nanoleaves were exposed to the electrolyte solution, to the benefit of fast electron/ion transfer. The hierarchical porous architecture greatly increased the surface area, offering sufficient diffusion channels, facilitating the electrolyte to diffuse more easily into the inner region of the electrode (Fig. 9b). In addition, the sufficient void spaces between the neighbouring nanoleaves, owing to their role of ion buffering reservoirs, ensured sufficient redox reactions to take place at high current densities.
![]() | ||
Fig. 9 (a) Cycling performance and Coulombic efficiency of the CuO flower electrode at progressively higher current densities; (b) schematic of the hierarchical porous structure. |
Footnote |
† Electronic supplementary information (ESI) available: TG-DTA curves, size distribution of CuO flowers, and electrochemical characterization of the CuO flowers, corals and spheres. See DOI: 10.1039/c4ra16924g |
This journal is © The Royal Society of Chemistry 2015 |