Introduction

Supercapacitors (SCs) are widely applied in portable electronics as well as smart-grid electrochemical energy storage, because of their higher energy densities compared to electrostatic capacitors and higher power densities compared to batteries1,2. According to the charge storage mechanism, there are two kinds of typical SCs: electric double layer capacitors (EDLCs) with charge absorbed on the surface of an electrode and pseudocapacitors (PCs) with reversible Faradic redox reactions on the surface and near-surface. Generally transition-metal oxides and conductive polymers are representative pseudocapacitive materials3. Their capacitance is generally 10 times larger than that of a carbon-based EDLC4. Among various transition-metal oxides, ruthenium oxides (RuOx) has been intensively investigated because of its attractive pseudocapacitive behaviour and stable long-term performance. However, the disadvantages of high cost and environmental harmfulness hamper its development in commercial applications. In this regard, several transition-metal oxides, such as manganese oxides5,6,7, vanadium oxides8,9,10,11, titanium oxides12,13,14 and their composites15,16,17,18 have been explored as the alternatives, due to their designed structures, excellent electrochemical performances and relatively reasonable price.

In practical applications, a high electrochemical performance and durability of a SC greatly depends on the combination of multiple factors, including the measurement techniques, electrode materials, electrolyte as well as the approach to assemble materials on the current collectors19. Normally in a typical fabrication process, a slurry of the mixture of binders, active materials and conductive additives with desired density is spread onto a proper current collector, followed by a drying and roll-pressing process to achieve a uniform coating layer1. However, the binders and other additives used in the fabrication process could induce several problems such as an increased resistivity and the addition of dead weight of an electrode20,21. Therefore, it is desirable to develop the binder-free and additive-free electrodes to overcome the mentioned problems.

To fabricate the binder-free electrodes, several techniques such as atomic layer deposition (ALD)22, vacuum filtration23,24, chemical vapor deposition (CVD)25,26,27 and electrospinning4,28,29 have been developed. Nevertheless, they always involve a disadvantage of expensive equipment, which hinders the large-scale production for practical applications30. In comparison, electrostatic spray deposition (ESD) is a versatile technique that has a cheap and simple set-up and can be worked under ambient conditions. It has been demonstrated successfully to fabricate various metal-oxide films with well-defined nano-structures31,32,33,34.

Herein, we fabricate binder-free VO2/TiO2 nanosponges by an ESD route and explore their electrochemical performance in SCs. The porous VO2/TiO2 film delivers a highest electrochemical capacitance of 86.2 mF cm−2 (~548 F g−1) between −1 and −0.3 V (vs. SCE) at a scan rate of 10 mV s−1. The three-dimensional (3D) pores grown on the porous nickel foam greatly enlarge the working area of film electrodes, help to facilitate the infiltration of electrolyte, thus largely enhance the electrochemical capacity35,36. Moreover, the rational design involving both high theoretical specific capacitance of vanadium oxides and high structural stability of titanium oxides improves the long-term durability during charge/discharge processes.

Results and Discussion

For a detailed study, the films with an increasing molar ratio between V to Ti were prepared and marked as VT1, VT2 and VT3. To examine the surface electronic state and chemical composition, the products with different vanadium contents deposited on the glass slides followed by a certain heat treatment were characterized by X-ray photoelectron spectroscopy (XPS) (Fig. 1 and Fig. S1). A typical XPS spectrum for the VT2 film (Fig. 1a) indicates the existence of V, Ti, O and C elements. The high-resolution XPS spectrum for V2p3/2 can be deconvoluted into two peaks (Fig. 1c), where the peaks at 516.4 and 517.4 eV correspond to the reported binding energy for V4+ and V5+10,37. The existence of V5+ may result from the surface oxidation in air, which has been mentioned by other group38. The peaks at 464.28 and 458.60 eV are assigned to Ti2p1/2 and Ti2p3/2, which are characteristics of Ti(IV) in TiO2 (Fig. 1b)39,40. The O 1s in the XPS spectrum (Fig. 1d) can be fitted by three peaks centered at 530.0, 530.2 and 530.4 eV. The peak at 530.0 eV is ascribed to O 1s in the Ti-O linkage of TiO2, whereas the peak at 530.0 and 530.4 eV belong to the V-O linkage of VO2 and V2O5. In addition, the corresponding X-ray diffraction (XRD) patterns (Fig. S2) and energy-dispersive X-ray (EDX) (Fig. S3) analysis further confirm the phase and the composition of the films. Although VO2 and TiO2 can form a solid solution, the diffraction peaks in the XRD patterns of VT1, VT2 and VT3 films and powders indicate a mixture of monoclinic VO2 and anatase TiO2. This may due to the different formation steps of the two components. TiO2 is formed from the hydrolysis of TTIP during the early stage of deposition, while VO2 is formed after the certain heat treatment. However, Ti in TiO2 shares the same valence state and similar ion diameter to V in VO2, it is still very possible that a slight amount of Ti can be doped into VO241. The atomic ratio between V to Ti from the EDX results present a good agreement with the composition in the precursor solution. This suggests that the composition of the products can be easily controlled and tailored as desired.

Figure 1
figure 1

XPS spectra of the VT2 film.

(a) Survey spectrum. High-resolution XPS spectra of (b) V 2p, (c) Ti 2p and (d) O 1s.

Full size image

The 3D porous morphologies of the as-formed films were revealed by field-emission scanning electron microscopy (FESEM) test. Figure 2a,c,e present the low-magnification FESEM images for the VT1, VT2 and VT3 films on the flat aluminum foil, while the insets present the films on the porous nickel foam. A highly uniformly distributed film on a flat aluminum at a large scale of 100 μm and even on a porous nickel foam (Fig. 2a,c,e) demonstrates a great advantage of the ESD technique – substrate-insensitive. The high-magnification SEM images (Fig. 2b,d,f) show that each film presents a 3D porous network structure with the diameter of the inner connected holes up to few micrometers. The skeleton of the porous sponge is comprised of tangled spherical particles and bunches and their sizes depend greatly on the experimental conditions, including the flow rate, physical properties of the precursor solution and the applied voltage42. Generally, spray droplets become smaller with the increase of the applied voltage and decrease of the flow rate. This is also confirmed by the smallest particles in the VT3 film. The formation of such a highly porous structure can be ascribed to the two-step volatilization process caused by the solvent system43. The particles in VT3 show an average diameter less than 100 nm and the sticks are approximately 100 nm in length and 30 nm in diameter. Interestingly, the surface morphology remains almost unchanged during the heat treatment, indicating a highly stable feature of this 3D sponge. For comparison, the annealed powder from the dried precursor was also prepared. Results show that the shape of the resulting powder is irregular (Fig. S4). Scanning transmission electron microscopy (STEM) and X-ray elemental mappings characterization provide more insights into the detailed stucture. Figure 3a displays a representative TEM image for VT2 film. The diameter of the particles is around 20 to 40 nm and some big holes between the particles could be beneficial to the infiltration of electrolytes. The STEM results (Fig. 3b,c,d) reveal that the TiO2 and VO2 nanoparticles are homogeneously distributed in nano-scale, indicating a stable combination of the two main components.

Figure 2
figure 2

SEM images of the obtained thin films on the aluminum foil.

(a,b) VT1, (c,d) VT2 and (e,f) VT3. The insets show the films on the nickel foam.

Full size image
Figure 3
figure 3

TEM morphologies of the VT2 film.

(a) TEM image of VT2 film and elemental mapping images of (b) O, (c) V and (d) Ti.

Full size image

To evaluate the advantages of the 3D porous structure, the electrochemical performance of VT2 was measured in a three-electrode cell. An aqueous LiCl solution (8M) was selected as the electrolyte because of the negligible solubility and better cycling performance of vanadium oxides in it22,44. The thin films on the porous nickel foam and followed by an annealing treatment were used as the electrodes directly without adding any binders and other additives. The nickel foam performs as the current collector, ensuring a good pathway for electrons without introducing much extra capacitance (Fig. 4a). All the Cyclic voltammetry (CV) curves display a quasi-rectangular shape without obvious redox peaks, which may be caused by the low crystallinity of the particles and the selected potential window9. We also explored the electrode made of irregular powders with the traditional roll-pressing fabrication process. It is observed that the area enclosed by CV curves of the VT2 film triples that of powders at a scan rate of 100 mV s−1, implying a larger areal capacitance. Remarkably, the gravimetric capacitance of the electrodes made of the VT2 films calculated from the formula of Cm = Q/(2SρΔE) is 6 times higher than that of the powders. This observation can be assigned to the unique 3D structure obtained by the ESD method. The interconnected pores provide higher specific area and more active sites for ion absorption and intercalation45. Figure 4b displays the CV curves of the VT2 film at various scan rates ranging from 10 to 1000 mV s−1. The maximum cathodic/anodic current density increases linearly with the scan rate, indicating a diffusion-controlled process. This can be attributed to the mass transport restrictions of the electrolyte, limiting the reaction with electrolyte species and active materials. According to the calculation using Cs = Q/(2SΔE) and Cm = Q/(2SρΔE), the thin film of VT2 exhibits a specific capacitance of 26.8 mF cm−2 (~174 F g−1) at the scan rate of 100 mV s−1, which agrees well with the previous results17,18. Figure 4c exhibits the galvanoastatic charge-discharge (GCD) curves at various current densities from 3 to 20 A g−1. The typical curves with a triangular shape show an obvious distortion, which is indicative of pseudocapacitive behaviour influenced by redox reactions along with electric double layer behaviour. A high specific capacitance of 715.8 F g−1 is obtained at a current density of 5 A g−1. It is observed that the iR drop increases with the increase of discharge current densities, but is still relatively small.

Figure 4
figure 4

Electrochemical performance of the VT2 film.

CV curves for (a) the VT2 film, the VT2 powder and the nickel foam at a scan rate of 100 mV s–1 and (b) the VT2 film at the scan rates of 10, 20, 100, 200, 500 and 1000 mV s−1. (c) Galvanoastatic charge-discharge curves of the VT2 film at current densities of 3, 4, 5, 8, 10, 15 and 20 A g−1. (d) Cycling performance at a current density of 10 A g−1 over 1000 cycles. The inset shows the specific capacitance as a function of scan rate ranging from 10 to 1000 mV s−1.

Full size image

Furthermore, long-term cycling behaviour was investigated using the GCD measurement at a charge/discharge density of 10 A g−1 (Fig. 4d). After charging/discharging for 1000 cycles, the capacitance still remained 84.3% of the initial value. Notably, the Faradaic capacitance brings a large volumetric change in the vanadium oxide, thus leading to a severe decay of the electrochemical performance during the expansion/contraction process. However, the typical zero-strain material, anatase TiO2, can server as the stabilizer that restrains the volumetric change and keeps the framework stable during ion intercalation/deintercalation, therefore improving the cyclability. It is also reported that in the mixed system of titanium oxides and vanadium oxides, TiO2 prohibits a reorganization of the microstructure of the material, which improve the cycling stability46. Therefore, the largely improved electrochemical performance can be atributed to the 3D sponge-like structure, which increases the surface area and helps to improve the infiltration of the electrolyte as well as the accessibility of ions.

To further understand the influence of the VO2/TiO2 composition on the electrochemical performance, the electrodes with different molar ratios of VO2 to TiO2 were tested. Figure 5a exhibits the CV curves of the electrodes with different VO2/TiO2 ratios at a scan rate of 100 mV s−1. All the CV curves show symmetrical cathodic/anodic current peaks, implying a good reversibility of the electrodes made of 3D porous VO2/TiO2. Evidently, increasing the vanadium content can greatly enhance the current of redox reactions. The capacitances calculated from Cm = Q/(2SρΔE) exhibit a linear increase to the molar percentage of VO2 (Fig. 5c), demonstrating a tailored capacitance through a designed compositions. Figure 5b summarized the rate capability calculated from CV measurements at scan rates between 10 to 1000 mV s−1. The thin film of VT3 exhibits the highest specific capacitance of 86.2 mF cm−2 (~548 F g−1) at the scan rate of 10 mV s−1. The specific capacitance can still be kept to be 13.0 mF cm−2 (~82 F g−1), when the scan rate reaches a high value of 1000 mV s−1. This is a satisfactory rate capability of pure metal oxide without any other additives. It can be ascribed to the decreased crystallite size as well as the increased surface area and the vanadium content in the VT3 film. The above results demonstrate that the additive-free thin-film electrodes of VO2/TiO2 exhibit an excellent supercapacitive performance. The interconnected pores provide an increased voids and defects, thus may lead to more active sites for redox reactions. Meanwhile, the monoclinic VO2 offers a high electrochemical capacity, while anatase TiO2 serving as a stabilizer both buffers the volumetric change and guarantees the long-term cyclability.

Figure 5
figure 5

Electrochemical performances of VT1, VT2 and VT3 films.

(a) CV curves at a scan rate of 100 mV s−1, (b) the specific capacitance as a function of scan rate and (c) the specific capacitance as a function of molar ratio between VO2 to TiO2 for the VT1, VT2 and VT3 films.

Full size image

In summary, the binder-free VO2/TiO2 nanosponge electrodes have been successfully fabricated by an ESD technique. The electrodes present an easily tailored and controlled composition as well as the 3D sponge-like structure. Our results demonstrate that the additive-free VO2/TiO2 electrodes show a greatly enhanced electrochemical capacity of 86.2 mF cm−2 (~548 F g−1) at a scan rate of 10 mV s−1 and a satisfactory long-term stability of 84.3% retention at a current density of 10 A g−1. This superior supercapacitive performance can be ascribed to the unique highly porous structure that improves electrolyte infiltration and the zero-strain anatase TiO2. Furthermore, the facile and low-cost ESD route can be extended to fabricate other functional nanostructured thin films for energy-storage devices.

Experimental Section

Materials Synthesis

All the chemical reagents were commercially purchased and used directly without any purification. 0.0585 mg of NH4VO3 and 0.1576 mg of oxalic acid (C2H2O4) were dissolved into 10 ml of ethanol at 70 °C for 2 h, which is marked as solution A. 0.0716 mg of titanium (IV) isopropoxide (TTIP) was added into a mixture of 2.5 ml of ethanol and 2.5 ml of acetic acid, which is marked as solution B. A proper amount of solution A, i.e., 2.5, 5 and 10 ml, was added into solution B (5 ml) while vigorous stirring, which is marked as VT1, VT2 and VT3. Then 40 ml 1, 2-propylene was added into the solution.

The as-prepared precursor was transferred to a computer-controlled syringe pump, where the metallic needle was connected to a high-voltage power supply. Aluminum foils, nickel foams and glass slides were chosen as the substrates for scanning electron microscopy (SEM), electrochemical measurements and X-ray diffraction (XRD) analysis, respectively. During the ESD procedure, the substrate was kept at 260 °C and the distance between the substrate and the needle was 3 cm. The feeding rate and the applied voltage for VT1, VT2 and VT3 are shown in Table S1. To remove the excessive carbonaceous substances, the resulting thin films were thermally treated at 350 °C in air for 15 min before thermal treatment at 500 °C (heating rate : 1 °C min−1) under 5% NH3/Ar atmosphere for 90 min. For comparison, the precursor was dried at 260 °C and treated with the same heat treatment.

Materials characterization

X-ray diffraction (XRD) analysis was performed by X’Pertb PRO (PANalytical B.V., Holland) Diffractometer with high-intensity Cu Kα1 irradiation (λ = 1.5406 Å). The morphology was characterized using a field-emission scanning electron microscopy (FESEM, FEI, Sirion 200) coupled with an energy-dispersive X-ray (EDX, Oxford Instrument) spectrometer and transmission electron microscopy (TEM, JEOL 2100F). X-ray photoelectron spectroscopy (XPS) was measured on a VG MultiLab 2000 system with a monochromatic Al Kα X-ray source (ThermoVG Scientific).

Electrochemical measurements

All the electrochemical performances were carried out on a CHI 660D electrochemical workstation. The SC tests were performed in a three-electrode cell with 8 M LiCl aqueous solution as the electrolyte and a Pt foil and a saturated calomel electrode (SCE) as the counter and reference electrodes, respectively. The substrates deposited with active materials were cut into pieces with average working area of ~1.2 cm2 and used as working electrodes directly. For comparison, the dried powder was mixed with polytetrafluoroethene (PTFE) and super P in a weight ratio of 80:10:10. The mixture was pressed onto the nickel foam for measurements. Cyclic voltammetry (CV) was performed at scan rates of 10 to 1000 mV s−1 over the potential range from −1.0 to −0.3 V (vs. SCE). Galvanostatic charge-discharge (GCD) measurements were carried out at current densities from 3 to 20 A g−1 and potential window between −0.9 and −0.3 V (vs. SCE). The calculation of the integrated-average volumetric and gravimetric capacitance of each electrode was according to: Cs = Q/(2SΔE) and Cm = Q/(2SρΔE), where Cs (F cm−2) and Cm (F g−1) is specific capacitance, Q (C) is the integrated cathodic charge, S (cm2) is the working area of electrodes, ρ (g cm–2) is the mass of active material per unit area and ΔE (V) is the width of potential window. Considering for the accuracy and reproducibility, all electrodes were tested in order: CV (scan rates increasing from 10 to 1000 mV s−1) was followed by GCD (current densities increasing from 3 to 20 A g−1).

Additional Information

How to cite this article: Hu, C. et al. VO2/TiO2 Nanosponges as Binder-Free Electrodes for High-Performance Supercapacitors. Sci. Rep. 5, 16012; doi: 10.1038/srep16012 (2015).