Transition metal oxides, as well as multivalent metal oxides like Co3O4, have attracted huge attention for energy-storing materials. Combining Co3O4 with carbon materials is of interest to improve the limited electrical conductivity and electrochemical performance of Co3O4 composites. However, most synthesis methods of those Co3O4 and carbon composites commonly need plenty of chemicals with long calcination and release many harmful byproducts. Thus, our green strategy based on laser sintering to synthesize a carbon-Co3O4 composite from a single source – cobalt acetate – is proposed. With a low-power laser, our method is applicable to making flexible devices. A demonstration of flexible supercapacitor cells formed on a silver-nanowires-coated polyimide substrate is presented. Furthermore, the effects of laser pulse duration and current collector conductance on the device performance are investigated. As tested in Na2SO4 and KOH electrolytes, these cells can deliver specific capacitances of 13.2 F g−1 (at 0.08 A g−1) and 48.2 F g−1 (at 1 A g−1), respectively, with a remarkable cycling behavior of 90% after 3000 cycles. Notably, the usability of our composite with the neutral Na2SO4 electrolyte is worth exploring for eco-friendly energy storage. Altogether, our study shows opportunities to create green composites of carbon and multivalent metal oxides quickly and cost-effectively.
For making flexible SC electrodes, silver nanowires (Ag NWs, dispersed in IPA, NovaWire) were spray-coated on a polyimide substrate (PI; 1 × 2 cm2) as the charge collector. After spin-coating of the cobalt salt solution at 500 rpm for 10 s, the samples were dried at 100 °C for 3 min. Here, we used 20 μL of the solution for 1 cm2 of sample area. It is worth noting that the short 3-min annealing period is critical to keep a small amount of water in the coated layer, which helps reduce the burning of carbon groups in the air, leading to the formation of carbon black, as illustrated in Equation (2)–(6). If the film is too wet (with plenty of water), the formation of cobalt oxides (CoO and Co2O3/Co3O4) will be affected while the formation of cobalt hydroxides (Co(OH)2 and CoOOH) will be more significant, as proven in Supporting Fig. S1. Next, the samples were loaded into a portable laser engraver (NEJE; 1 W laser power, 405-nm wavelength, and 5 V power supply) for sintering. The scanning mode of the laser module (5–20 ms per pulse, the area of laser beam having a spot diameter of 70 μm) was adjusted to expose the laser light on the cobalt salt layer (1 × 1 cm2). During scanning, the decomposition of the salt solution slightly emitted a vinegar odor, which is the acid acetic vapor, which is not harmful to human health or to the environment) [24,25]. The samples could be dipped in deionized water to remove residual salts, but this is an optional step because residual Na cations in the electrodes (tiny amount) can participate to the electrochemical process). Finally, the fabricated electrodes were assembled with a cellulose membrane containing the aqueous Na2SO4-based electrolyte to construct a flexible SC cell.On the other hand, there are some studies using laser for the carbonization of organic compounds [30–32]. Therefore, the salt decomposition during laser scanning could produce cobalt atoms. To find the existence of carbon black in the CoOx layer, X-ray diffraction (XRD) and energy dispersive X-Ray spectroscopic (SEM-EDS) measurements were performed. Fig. 2b displays the XRD patterns of cobalt salt and CoOx layers coated on a SiO2 substrate. Here, the cobalt salt sample (after the concentration step) exhibited a few small characteristic peaks of tiny particles of Co2O3 or Co3O4 (JCPDS card 76–1802 or 42–1467) and Co(OH)2 (JCPDS card 30–443). After laser-sintering (15-ms pulse), the XRD pattern of the C-CoOx layer contains more intense peaks due to the formation of larger polycrystals of not only Co3O4 (majorly) but also CoO (minorly; JCPDS card 65–2902). Notably, the peaks at 2θ = 25, 31, 37, 38, and 45° are matched with the crystal the planes (200), (220), and (311) of CoO; and the ones at 2θ = 32, 38, 51, and 58 correspond to the planes (100), (002), (102), and (110) of Co(OH)2 [33–36]. Comparing the intensity of their characteristic peaks in the XRD data (Supporting Fig. S2d), we found that the contents of cobalt oxides in the sintered sample dominated others (e.g., Co(OH)2 and Na2O). Besides, the intensity increase of Co3O4 peaks and the appearance of new CoO peaks in the sintered samples are consistent with the growth and merging of CoOx NPs, as fore mentioned. On the other hand, we also recognized the XRD pattern of the CoOx layer containing characteristic peaks of carbon black [37]. The SEM-EDS data in Fig. 2b also revealed the existence of carbon elements at a high atomic concentration in the sintered sample. As we observed, the appearance of carbon black can originate from the incomplete burning of organic compounds by laser pulses (e.g., CH3COOH → 2C + 2H2O). Despite the short pulse, the high intensity of the laser light may provide sufficient heat for the carbonization as well as the formation of carbon-carbon aromatic bonds (>250 °C) [38–42]. Furthermore, a short annealing period at 100 °C after the spin-coating step is critical to keep a small amount of water in the layer, which helps reduce the conversion of carbon groups into CO2 and raises the amount of carbon black. However, if the water amount is plenty, the formation of cobalt hydroxide will be dominated, as illustrated in Equation (2) and (3). Regardless, the formation of carbon black along with cobalt oxide in our samples was confirmed, and that is advantageous since both are good candidates for energy-storing applications [43].Technically, with an increase in laser pulse duration, the samples certainly received a higher amount of heat, which is beneficial for the decomposition of the cobalt salt, the formation of the carbon mixed cobalt oxide, and the crystallization of the CoOx layer. However, the total period for a laser scan would be prolonged (3–10 min for 15–30-ms pulses, scanning on an area of 1 × 1 cm2). In addition, the flexible PI substrate could be deformed and carbonized [30] due to receiving too much heat, as shown in Supporting Fig. S3a. Please kindly note that we found no effect of long laser pulses on the outer appearance of solid substrates like SiO2 and FTO. Therefore, for making flexible energy devices, controlling the laser pulse duration is critical to not only achieve the high electrochemical performance of the CoOx layer but also avoid the deformation of flexible substrates.Particularly regarding the robustness of substrates covered by the cobalt salt layer, with the short pulse duration of 7 ms, no clear PI deformation was observed. Meanwhile, with the laser pulse duration of 15 ms or higher, the thermal deformation of the PI substrate happened more significantly, but it is worth noting that the deformed electrodes (due to laser pulses ≤30 ms) still worked well, as displayed in Supporting Figs. S3b–e. If the aesthetic of flexible devices is of concern, the optimum duration of laser pulses considered for the PI substrate is about 7 ms. For other transparent flexible substrates like polyethylene naphthalate (PEN) and polyethylene terephthalate (PET), it was surprising that they could endure a longer laser pulse duration of ∼20 ms before the thermal deformation as well as a burning area was observed Supporting Fig. S3a). Regardless, for making flexible low-cost SCs, we still prefer using the PI substrate because PEN and PET substrates are quite expensive and have a low flex lifespan [44].In this section, we investigated the electrochemical performance and stability of our flexible SC cells. Fig. 4a shows CV characteristics of the flexible SC cell in 1 M Na2SO4 obtained at different scan rates (5–100 mV s−1). Its CV curves appeared a quasi-rectangular shape. Moreover, with the increase in the scan rate, the area of the CV curve also increases in a linear fashion. Those indicated the excellent charge transportation and the charge storage capacity at the surface of the C-CoOx electrodes. Fig. 4b shows the galvanic charge-discharge (GCD) behavior of the flexible SC cell in a potential window of 0–0.8 V with different current loadings from 0.05 to 1 mA. Because of the low internal resistance, the IR drop was small in the discharging cycles, implying that the current-voltage response was quite rapid. The specific capacitance of our working electrodes was estimated with an active area (laser-scanned) of 1 cm2 and the material loading of 0.6 mg. The specific capacitance of our flexible device, Cs, was ca. 5.8–13.2 F g−1 obtained at the current loading of 1–0.05 mA. However, these are not the best electrochemical performance of our samples. As presented in Supporting Fig. S5, our SC cell tested in a basic KOH electrolyte exhibited significant enhancement in the current density (ca. 20-fold higher) and the specific capacitance (Cs = 48.2 F g−1 at 0.6 mA or 1 A g−1). Besides, the stability of the SC cell was evaluated by continuously operating the cell through 3000 cycles of GCD (with a current loading of 1 mA or 1.6 A g−1), as shown in Fig. 4c and d. In addition, Fig. 4e and f display the first and last cycles of the GCD tests. It can be seen from these figures that the flexible cell exhibited excellent capacitance retention ca. 90% after 3000 cycles of GCD. In general, these obtained results are yet comparable to those of reported stiff SC composites based on cobalt composites, as shown in Table 2. Regardless, comparing to other flexible SC composites, the electrochemical performance of our carbon-mixed-CoOx composite is promising for further investment and practical use.We gratefully acknowledge the financial support of National Research Foundation (2018H1D3A1A02074733 & 2018R1D1A1B07050008) from the Ministry of Science and ICT and the Ministry of Education, Republic of Korea.We gratefully acknowledge the financial support of National Research Foundation ( 2018H1D3A1A02074733 & 2018R1D1A1B07050008 ) from the Ministry of Science and ICT and the Ministry of Education, Republic of Korea .
