Rechargeable metal-air batteries have attracted significant attention as energy storage devices owing to their higher theoretical energy density than commercialized Li-ion batteries, especially non-aqueous Li-O2 and Na-O2 batteries [1-6]. Na-O2 batteries are of particular interest as Na exhibits similar physicochemical properties to Li. Considering the limited Li resources that are inadequate to satisfy the increasing demand for batteries based on Li chemistry, many researchers call for using Na to replace Li owing to its abundance and inexpensive nature [7-9]. However, before the practical application of Na-O2 batteries becomes a reality many significant technical challenges need to be overcome, such as poor rate capacities, low round trip efficiencies (caused by a high overpotential of both the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER)), unstable electrolytes and polymer binders (which undergo side reactions), and short cycle lives [10, 11]. And, similar to Li-O2 batteries, the properties of the cathode material (such as morphology, specific surface area, structure, activity, and conductivity) play an important role in the performance of Na-O2 batteries [12-15]. In response, many scientists have tried to overcome these above mentioned limitations by finding a suitable cathode catalyst to accelerate the ORR and OER kinetics of rechargeable Na-O2 batteries. They tried to achieve this by adjusting the porous structure of the cathode catalyst to improve cycle stability through providing enough paths for oxygen and ion migration and enough sites for deposition of the discharge product. Additionally, researchers have tried tailoring the cathode structure to enhance its structural stability [16]. For example, carbon supported catalysts, such as mesoporous carbon, carbon fiber, and N-doped graphene nanosheets have been widely used as cathodes for Na-O2 batteries owing to their enhanced electronic conductivity and catalytic activity [17-19]. In general, cathodes are prepared by coating a homogenous slurry containing conductive carbon powder, polymer binder (e.g. PVDF and PTFE), and catalyst onto the current collector. However, the carbon itself and the commonly used polymer binder are reported to be unstable, forming decomposition products during the discharge/charge process [20, 21]. To solve these problems, the development of a binder free or non-carbon cathode is of great importance. Prioritizing the mechanical strength of the cathode, we chose carbon textiles as the base owing to their superior flexibility and excellent electrochemical stability.
Herein, we report a facile and efficient hydrothermal synthesis method to fabricate a flexible and binder free cathode with Co3O4 nanowire arrays vertically grown onto carbon textiles (this cathode is referred to as COCT throughout the rest of this manuscript). When employed as cathode in a Na-O2 battery, the COCT cathode endows the Na-O2 battery with lower overpotential, enhanced specific capacity of 4687 mAh/g (based on weight of Co3O4), and enhanced cycling performance (62 cycles).
Co(NO3)2·6H2O and NH4F were purchased from Sinopharm Chemical Reagent Co. Ltd.,Shanghai,China. Urea (CO(NH2)2), tretraethylene glycol dimethyl ether (TEGDME), and sodium triflate (NaCF3SO3) were purchased from Aladdin Reagent. Carbon textiles (CT) were purchased from Torray.
The COCT cathode was synthesized using a hydrothermal method. The CTs were ultrasonically cleaned several times with acetone, absolute ethanol, and distilled water and then dried at 60 °C in a vacuum oven for 12 h. The precursor solutions were obtained by dissolving a desired amount of mixed salt (the molar ration of Co(NO3)2·6H2O:NH4F:CO(NH2)2 = 1:1:2) into 40 mL distilled water at room temperature. After stirring for 1 h, the above solutions were transferred to a Teflon lined stainless steel autoclave (50 mL) and the cleaned CTs were immersed in the precursor solutions. Then the Teflon lined stainless steel autoclaves were heated at 120 °C for 5 h in an oven. After cooling down, the CTs were removed from the autoclaves and rinsed several times with deionized water and ethanol, then dried in a vacuum oven at 120 °C for 12 h. Finally, the samples were annealed in air at 400 °C for 2 h. The synthesis steps for COCT cathodes are schematically illustrated in Scheme 1.
Powder X-ray diffraction (XRD) measurements were performed with a Bruker D8 Focus Powder X-ray diffractometer using Cu Kα radiation (40 kV, 40 mA). Scanning electron microscopy (SEM) was performed using a HITACHI S-4800 field mission scanning electron microscope. Transmission electron microscopy (TEM) and high resolution transmission electron microscopy (HRTEM) were undertaken on a FEI Tacnai G2 electron microscope operated at 200 kV. X-ray photoelectron spectroscopy (XPS) analysis was carried on a VG Scientific ESCALAB MKⅡ X-ray photoelectron spectrometer.
Cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) measurements were performed on a BioLogic VMP3 electrochemical workstation. Na-O2 battery measurements were cycled on a LAND CT2001AS2 multi-channel battery testing system.
The electrochemical performance of COCT as a cathode in a Na-O2 battery was tested using a coin2025 type cell. The cathodes were dried in a vacuum oven at 80 °C for 24 h. All batteries were assembled in a glove box under Ar atmosphere, using a sodium metal foil anode, glass fiber separator, oxygen cathode, and electrolyte containing 0.5 mol/L NaCF3SO3 in TEGDME. Galvanostatic discharge-charge tests were conducted within a voltage window of 1.8-4.2 V (vs. Na/Na+) at ambient temperature after a 2-5 h rest period. EIS measurements of the cells were carried out using an AC impedance analyzer within a frequency range of 106 to 10−2 Hz. The CV curves were measured form 1.8 to 4.0 V at a voltage sweep rate of 0.5 mV/s.
We investigated the morphology and structure of the COCT cathode using SEM and TEM. Fig. 1(a) shows an SEM image of the pristine CT, revealing that the CTs are woven by carbon fibers with diameters of about 10 μm. Fig. 1(b) and (c) show SEM images at different magnifications of the carbon textiles after Co3O4 NWs were grown on them; it could be observed that the Co3O4 NWs vertically grew uniformly on the CT frame without the help of any additional polymeric binder. This contributes to the formation of a low resistance path for electron transport. The average diameter of the nanowires was about 40 nm (Fig. 1(d)). The HRTEM image collected from the surface of the Co3O4 NWs exhibits well resolved lattice fringes with d spacings of 0.24 and 0.47 nm, corresponding to the (311) and (111) planes of Co3O4, respectively (Fig. 1(e)). The selected area electron diffraction (SAED) pattern shown in the inset of Fig. 1e demonstrates the polycrystalline nature of the NWs. Moreover, the XRD pattern (Fig. 1(f)) shows that the diffraction peaks can be assigned to the crystalline Co3O4 phase (JCPDS 42-1467). Further, no other diffraction peaks were observed, which confirms that there were no impurities.
To examine the elemental distribution and composition of the COCT surface, we used energy dispersive spectrometer (EDS) mapping (Fig. 2(a)-(e)). O and Co are uniformly distributed around a single CT fiber, and the diameter is increased from 10 μm (pure CT fiber) to about 13 μm for the resulting composite COCT. The SEM and EDS images demonstrate the core-shell configuration of COCT, so this structure can effectively prevent the carbon textiles from decomposing. A more detailed elemental composition and the oxidation state of the Co3O4 NWs are measured by XPS. From the C 1s spectra (Fig. 2(f)), we can see that C is not oxidized (the standard binding energy of C 1s is 285.0 eV). By using a Gaussian fitting method the best deconvolution of the Co 2p profile was achieved, which showed two pairs of spin-orbit doublets indicating the existence of Co2+ and Co3+, which is consistent with previously published results on Co3O4 (Fig. 2(g)) [22, 23].
Compared with the traditional air cathode, this flexible electrode material exhibits many advantageous properties that are vital to the transportation of oxygen, electrons, and ions, and thus improves the electrochemical performance of Na-O2 batteries. More importantly,Co3O4 NWs are vertically grown onto carbon textiles without the help of any polymeric binder, thereby effectively reducing the resistance of the cathode and avoiding side reactions originating from the decomposition of the non-conductive polymeric binder. Finally, the flexibility of the COCT cathode paves the way for the practical application of flexible Na-O2 battery devices.
The electrochemical performance of COCT cathodes compared with pure CT cathodes and with Co3O4 pasted on CT (CO@CT) cathodes was examined in Na-O2 batteries. We choose TEGDME (because of its relatively high stability toward O2−) with NaCF3SO3 as the electrolyte [24, 25]. The discharge-charge curves of the Na-O2 batteries with the three different cathodes are shown in Fig. 3(a). The overpotential of the Na-O2 batteries with the COCT cathode is lower than for the other two cathodes. We found that the discharge specific capacity and charge-discharge voltage was improved by the Co3O4 NWs. Specifically, the discharge capacity of the COCT cathode was 4687.2 mAh/g, which is three times that of the CO@CT cathode (1533.4 mAh/g). We attribute this to more void space being available for the deposition of discharge products and to the binder free character of the COCT cathode (polymer binder usually decomposes). The capacity of the Na-O2 battery with the pure CT cathode is 1113.7 mAh/g, which is almost one-fourth that of the cell with the COCT cathode. The discharge voltage and charge voltage of the COCT devices indicate that the Co3O4 NWs have superior ORR and OER catalytic activity. Furthermore, to exclude possible electrochemical contributions from intercalation reactions with CT and/or Co3O4 NWs, we investigated the CV curves of Na-O2 batteries (Fig. 3(b)). Compared with the CT cathode, the COCT cathode exhibits a higher ORR onset potential and ORR/OER peak current, which indicates the Co3O4 NWs have excellent catalytic activity. The enhanced ORR/OER kinetics could lead to improvements in the energy output, recharging characteristics, and round trip efficiency of the Na-O2 batteries.
Inspired by the superior catalytic activity of Co3O4 NWs, we further examined the energy efficiency of Na-O2 batteries. We found that the COCT cathode had a higher discharge capacity (Fig. 3(c)) and capacity retention rate (Fig. 3(d)) at both 100 mA/g and 1000 mA/g current density compared with the pure CT cathode, which we attribute to a synergistic effect of increased catalytic activity and increased porosity of the Co3O4 NWs. To clarify the above point, we investigated the SEM images of these two kinds of cathodes (Fig. 4). The discharge products on the CT cathode were shaped like sheets (Fig. 4(c)), while the discharge products grown on the Co3O4 NWs were uniformly distributed on the COCT cathode (Fig. 4(d)). The porous Co3O4 NWs provide enough space for the deposition of the discharge products, thereby resulting in a high discharge specific capacity. In addition, the Co3O4 NWs offer more oxygen and electrolyte paths in the electrode, which is crucial to improve the rate capability, as reported in other types of metal-air batteries.
Additionally, another enhancement of Na-O2 batteries with COCT cathodes is their increased cycling stability. As shown in Fig. 5(a) and (b), the batteries with COCT cathodes can cycle for 62 cycles at a current density of 100 mA/g with the limited capacity of 500 mAh/g, compared with the pure CT cathode with only 16 cycles. This enhanced cycling stability may be attributed to the unique properties of the COCT cathode because its ordered, porous Co3O4 electrocatalyst NWs may accelerate the formation and decomposition of the discharge products thereby improving the rechargeability of the cathode. Evidence for this is found in the SEM images (Fig. 4) of cathode: the discharge products on the CT cathode were shaped like sheets (Fig. 4(c)), while they formed a film on the Co3O4 NWs (Fig. 4(d)). After recharge, the discharge products completely decompose on the COCT cathode (Fig. 4(f)), indicating the recovery of COCT cathode, which ensures the rechargeability of the batteries. Conversely, the discharge products incompletely decomposed on the pure CT cathode. We investigated the different electrochemical characteristics affecting charge efficiencies of the cathodes using EIS of both CT and COCT cathode Na-O2 batteries at different discharged/charged states. As shown in Fig. 5(c) and (d), we found that the impedances of pure CT and COCT Na-O2 batteries in the pristine state are almost the same. After discharge, the impedances of both batteries increased significantly, which was caused by the poor electronic conductivity of the discharge products generated on the cathode. Interestingly, the impedances of Na-O2 batteries with COCT cathodes can almost be restored to the pristine state (Fig. 5(d)) after being recharged, which indicates that the discharge products can be almost fully decomposed during the charging process; this conclusion is in agreement with the findings from the SEM images of the COCT cathode (Fig. 4(f)). Conversely, the impedance of Na-O2 batteries with pure CT cathodes increases monotonously during the charging processes (Fig. 5(b)) owing to the incomplete decomposition of the discharge products (Fig. 4(e))—this once again highlights the unique properties of COCT cathodes.
Based on the thermodynamic standard potential of different sodium oxides and considering the real discharge voltage and the similar standard potentials, we deduce that the most likely discharge products are Na2O2 or NaO2 or a mixture of Na2O2 and NaO2:
To understand the discharge products of our Na-O2 batteries, we investigated the cathode surface using XRD. As shown in Fig. 6, both of the XRD patterns reveal that the composition of discharge products is a mixture of NaO2 and Na2O2. The peaks associated with these discharge products appeared after discharging and vanished after recharging, which is consistent with the SEM images and EIS data, demonstrating the reversibility of our Na-O2 batteries. Specifically, the XRD peaks associated with the COCT cathode were weaker than those of the CT cathode, indicating the poor crystallinity of the discharge products, which causes a high specific capacity and cycling stability of the batteries. After recharging, the XRD peaks of the discharge products on the COCT cathode vanished, indicating that the products completely decomposed, which is in agreement with the EIS results (Fig. 5(f)) and SEM images (Fig. 4(d)).
In summary, we demonstrated a facile and efficient method of fabricating a flexiblebinder free COCT cathode. When directly employed as the O2 cathode,Na-O2 batteries exhibit high specific capacity and enhanced cycling stability, which was attributed to the enhanced catalytic activity, porous nature, and flexible binder free structure of the COCT cathode. Although flexible power sources are crucial for the realization of next-generation flexible electronics, their application in such devices is hindered by their low theoretical energy density. Therefore, flexible Na-O2 batteries with COCT cathodes show great potential. However, to develop Na-O2 batteries for practical device applications, enormous challenges and technological issues must still be overcome, such as protecting the sodium and eliminating side reactions.
2016,
Vol. 37 Issue (7): 1172-1179 
PDF (1234 KB)
2016,
Vol. 37 Issue (7): 1172-1179 PDF (1234 KB)