Hydrothermally synthesized nickel cobalt oxide for bifunctional electrochemical supercapacitor and nonenzymatic glucose biosensor

Herein, various nickel cobalt oxide nanostructures with different Ni concentrations are prepared via a hydrothermal route followed by annealing process for an electrochemical supercapacitor as well as a nonenzymatic glucose biosensor. The electrode synthesized on carbon cloth using Ni0.9Co2.1O4 nanosheet-like morphology showed a maximum 516.51 F g−1 specific capacitance at 10 mV s−1 scan rate and the cyclic stability of 87.7% over 2000 GCD cycles. The electrode prepared with Ni0.3Co2.7O4 on CC offered a linear response from 0 to 0.3 mM glucose concentration and exhibited a maximum of 759.5 µA mM−1 cm−2 glucose sensitivity.


Introduction
Nowadays, human beings have suffered from problems such as deficiency of fossil fuels, global warming, and pollution. Hence, it is essential to produce energy from renewable sources like solar, tide, and wind. Also, low-cost, environmentally friendly, advanced energy storage and conversion devices must store energy harvested from renewable resources [1][2][3][4]. There are different energy storage devices such as conventional capacitors, biofuel, lithium-ion batteries (LIBs), and supercapacitors [5][6][7][8][9][10]. Supercapacitors, also called electrochemical supercapacitors, are the most desirable candidates for energy storage. Their unique electrochemical properties like high power density, excellent coulombic efficiency, and high cyclic stability make them compatible with other energy storage devices. Still, supercapacitors suffered from low energy density as compared to LIBs. So, to improve the energy density of supercapacitors, efforts have been made to develop novel electrodes with suitable synthesis strategies [9,[11][12][13][14].
The different conducting polymers [15], transition metal oxides, and transition metal hydroxides such as Co 3 O 4 [12] , NiO [16], Co(OH) 2 [17], Ni(OH) 2 [18,19], RuO 2 [20], MnO 2 [9,21] have been widely studied as an electrode for pseudocapacitors. The transition metal oxides deliver higher specific capacitance as compared to conducting polymers due to multi-electron redox reactions. In addition, ternary metal oxides, because of their better electrical conductivity and rich redox-active sites, shows better supercapacitor performance than single metal oxide [22]. Spinel nickel cobalt oxide (NiCo 2 O 4 ) has been widely used as high-performance capacitive electrode material in recent years. It shows better electrical conductivity, relatively lower electron transport activation energy, and better electrochemical activity than cobalt oxide (Co 3 O 4 )and nickel oxide (NiO) [22,23]. This high electrical conductivity and improved specific capacitance of NiCo 2 O 4 are because of contributions from both the Co and Ni ions with diverse valence states.
Researchers have made many efforts to enhance electrochemical performance by developing nickel cobalt oxide with different morphologies. Waghmode et al. [24] synthesized NiCo 2 O 4 nanoflowers by chemical bath deposition method and observed capacitance was 610 F g −1 at 1 mA cm −2 . Ma et al. [25], with the solution precursor thermal spray method, developed NiCo 2 O 4 hollow microspheres, which showed specific capacitance 902 F g −1 at 1 A g −1 . Qi et al. [26] synthesized NiCo 2 O 4 hollow microspheres by a hydrothermal method which showed a specific capacitance of 720 F g −1 at 2 A g −1 . Saravanakumar et al. [27] synthesized NiCo 2 O 4 nanoparticles via hydrothermal route and observed specific capacitance was 294 F g −1 at 1 A g −1 . Uke et al. [28] synthesized NiCo 2 O 4 nanomorphs by hydrothermal method and observed specific capacitance was 479 F g −1 with better retention.
In the medical industries, quality control in food and clinical diagnosis of diabetes; rapid, accurate, and sensitive detection of glucose is essential [29,30]. Recently, sensitive and selective enzyme-based sensors have been studied for glucose sensing. But they cannot be reused for continuous glucose detection. They suffer from low stability, which limits their use in actual applications [31,32]. The other method, i.e., nonenzymatic glucose biosensing, exhibited high stability, sensitivity, and reproducibility. Qin et al. [33]  In the present investigation, the different nickel cobalt oxide nanostructures by varying concentrations of Ni in cobalt oxide are prepared by the hydrothermal route and studied for both supercapacitor and glucose biosensing applications. The electrodes are synthesized using a flexible carbon cloth (CC) current collector with a simple doctor blade method. The synthesized Ni 0.9 Co 2.1 O 4 nanosheets exhibited 516.51 F g −1 specific capacitance at 10 mV s −1 with a capacity retention of 87.7% over 2000 cycles. At the same time, Ni 0.3 Co 2.7 O 4 nanosheets showed a maximum 759.5 µA mM −1 cm −2 glucose sensitivity with a linear response from 0.0 to 0.3 mM glucose concentration.

Preparation of different Nickel cobalt oxide powders
The nickel cobalt oxide powders with different Ni percentages were synthesized by the hydrothermal route and then calcination. In a synthesis of Co 3

Preparation of electrode
In an ultrasonication bath, a piece of CC (2 cm × 1 cm) was cleaned with ethanol, acetone, and DDW each for 10 min respectively, and dried overnight at 60 °C in the oven. The actual electrode was prepared by mixing prepared powders (active material), PVDF as a binder, and a carbon black ratio of 80:10:10. Then, a sufficient amount of NMP solvent was mixed to obtain a uniform slurry. Then, the prepared slurry was coated on CC (1 cm × 1 cm) and subjected to 80 °C in a hot air oven for 12 h. These prepared electrodes were used for electrochemical supercapacitor and nonenzymatic glucose biosensor study. The experimental procedure can be presented in Scheme 1. XRD pattern of prepared powders recorded using a Bruker D2 phaser tabletop model under Cu Ka radiation. FTIR was taken by Lambada-7600, Australia. The morphology was analyzed using FE-SEM from the Mira-3, Tescan, Brno-Czech Republic. Biologic SP-300 electrochemical workstation used for electrochemical supercapacitor and glucose biosensor study. Figure

FTIR study
The different nickel cobalt oxide powders were further characterized by the FTIR study. Figure 1B presents the FTIR spectra of all synthesized nickel cobalt oxide powders. The FTIR spectra display two sharp peaks at about 661 cm −1 and 565 cm −1 which originated from Co-O bonds (stretching vibrations), confirming the formation of Co 3 O 4 [38,39]. As the percentage of Ni increases, these two peaks shift towards a higher wavenumber. The corresponding change in the peak is mentioned in supplementary information Table S1. The relatively weak peaks at about 1632 and 3432 cm −1 are associated with the O-H bending and stretching mode of vibration of water [14,40,41].  Fig. 2A shows nanosheets like nature. As the different percentages of Ni are added into Co 3 O 4 , surface modifications are observed. In Fig. 2B, the nanosheet-like morphologies are observed, which are merged. As the percentage of Ni increases, the nanosheets are incorporated a lot in Fig. 2C. The flakes are started covered with nanoparticles in Fig. 2D, and finally, the number of nanoparticles increased as shown in Fig. 2E corresponding to the Ni 1.2 Co 1.8 O 4 . Figure 2F shows the EDS spectra of Ni 1.2 Co 1.8 O 4, which confirms the presence of Ni, Co, and O. Figure 2G-I presents the elemental distribution of Ni, Co, and O, respectively.

EIS study
The electrochemical behavior at the electrode-electrolyte interface was studied by using the EIS study. Figure 3

Electrochemical supercapacitor study
The three-electrode system with SCE as a reference electrode, platinum wire counter electrode, and prepared CC electrodes as a working electrode is used to study   Figure 4A presents the cyclic voltammetry (CV) curves of all nickel cobalt oxide electrodes studied at a 10 mV s −1 scan rate . The maximum capacitance values estimated from CV curves are 189. 31 Fig. S1(A-E). From the graphs in Fig. S1, it is clear that with an increase in scan rate, the area under the curve increases as the current drawn increases with the voltage in the CV graph. But increase in scan rate limits the interaction time between electrolyte and electrode. This limited-time interaction between the electrode and electrolyte also limits the utilization of the active surface area of the electrode. Hence specific capacitance decreases with the increase in scan rate [43].
The galvanostatic charge-discharge curves are displayed in Fig. 4B Fig. S2(A-E). The cyclic stability is measured for the Ni 0.9 Co 2.1 O 4 electrode as it exhibited a higher capacitance value than other nickel cobalt oxide nanostructured electrodes. The observed stability for the Ni 0.9 Co 2.1 O 4 electrode is 87.7% over the 2000 chargingdischarging cycles as depicted in Fig. 4C. The inset image of Fig. 4C shows the first and last charging and discharging cycles. The higher specific capacitance and stability of Ni 0.9 Co 2.1 O 4 electrode may be due to micro sheets covering nanospheres like morphology. The micro sheets and nanosphere morphology provide maximum surface area so that the interaction between electrode and electrolyte becomes easier, enhancing the specific capacitance value. The calculated specific capacitance from CV vs. scan rate is mentioned in Fig. 4D. In contrast, the variation of specific capacitance vs. current density is mentioned in Fig. 4E. Figure 4F presents the Ragone plot. The maximum energy density for the Ni 0.9 Co 2.1 O 4 electrode is 15.81 Wh kg −1 (at 0.75 kW kg −1 power density).
The total capacitance of an electrode includes surfacecontrolled capacitive process (ion adsorption/desorption) and diffusion-controlled charge storage (fast faradaic redox reaction of redox species) [44]. Hence, to study detailed charge storage behavior of different prepared samples, we performed the kinetics based on CV measurement. The relationship between current (i) and scan rate (ν) is as follows [45], where i is current at specific voltage, ν is the scan rate, a and b are constants. The values of b for surface-controlled capacitive process and diffusion-controlled process are 1 and 0.5, respectively. Figure   where k 1 and k 2 are equation parameters and their values can be found by the slope and intercept of graph (iν 1/2 vs ν 1/2 ), respectively (Fig. S3(A)). The capacitive and diffusion-controlled contributions in the Ni 0.9 Co 2.1 O 4 electrode are mentioned in Fig. 5B. The surface-controlled capacitive contributions for Ni 0.9 Co 2.1 O 4 electrode are 47.05%, 55.94%, 66.75%, 71.75%, and 73.91% at 10, 20, 50, 80, and 100 mV s −1 , respectively. Other electrodes capacitive and diffusion-controlled contributions are mentioned in Fig.  S3(B-F). In case of all electrodes, it is observed that the capacitive behavior increases with scan rate. The comparison of the supercapacitor performance of present work and previously reported work is mentioned in Table 1.

Nonenzymatic glucose biosensing study
Here, the same three-electrode system is used for the glucose-sensing study except for 0.1 M NaOH electrolyte. Figure 6A- During the oxidation, Ni 2+ and Co 2+ ions present in the NiOOH and CoOOH compounds oxidize to Ni 3+ and Co 3+ with the release of two electrons (Eq. 1-3). After adding glucose, the glucose molecules dissociate and are converted into gluconolactone with the release of two electrons (Eq. 4). During reduction, the oxidized species Ni 3+ and Co 3+ reduce back to the Ni 2+ and Co 2+ by accepting electrons and returning to their original states. Figure 8A-E presents the CV of different nickel cobalt oxide electrodes in the presence of 1.0 M glucose at various scan rates (10-100 mV s −1 ). It can be seen that with the scan rate, the anodic and cathodic peaks also increase. Figure 9A-E presents  Fig. S4. The estimated glucose-sensing parameters are given in Table 2. As shown in Fig. 10A-E, the selectivity of samples was examined by the addition of 0.2 mM glucose and other interfering species [0.0125 mM of uric acid (UA), Ascorbic acid (AA), Fructose (FR), Sucrose (SA), and Lactic acid (LA)]. All the electrodes showed a negligible effect of interfering species as compared to glucose. The response times for all prepared electrodes are mentioned in Fig. S5 (Supplementary data). The comparative glucose-sensing study of current work and previously reported work is mentioned in Table 3.

Conclusion
In summary, the nickel cobalt oxide by varying Ni concentrations has been prepared by the hydrothermal method. The nanosheet-like Ni 0.9 Co 2.1 O 4 electrode exhibited 516.51 F g −1 specific capacitance at 10 mV s −1 and cyclic stability of 87.7% over 2000 galvanostatic charge-discharge cycles. Also, it showed an energy density of 15.81 Wh kg −1 at 0.75 kW kg −1 power density. The Ni 0.3 Co 2.7 O 4 nanosheets on CC showed linear response from 0 to 0.3 mM glucose concentration and exhibited the highest glucose sensitivity of 759.5 µA mM −1 cm −2 . The overall study is a platform for the development of a flexible, high-stability supercapacitor and reliable nonenzymatic glucose sensor.