One-step synthesis of TiO₂ nanorod arrays on Ti foil for supercapacitor application

Over the past few years, energy storage devices have received tremendous interest due to the increasing demand for sustainable and renewable energy in modern electronic industry [1, 2]. Supercapacitors are considered as one of the most promising energy storage devices because of their high power density and long cycle life [3, 4]. Based on the charge mechanism, supercapacitors can be classified into two categories: electrochemical double layer capacitors (EDLCs) using carbon-based materials, and pseudocapacitors using redox-active materials such as transition metal oxides and conducting polymers. Compared with EDLCs, pseudocapacitors have higher specific capacitance because the fast and reversible surface Faradaic reactions dominate the charge storage; while EDLCs simply utilize the electrical charge at the electrode–electrolyte interface [5]. Therefore, most of recent supercapacitor studies have been focused on the development of redox-active materials, especially metal oxides, with high specific capacitance and good stability for pseudocapacitors [6–13].

Ruthenium oxide (RuO₂) is widely recognized as the most promising electrode material for supercapacitors due to its high specific capacitance [6, 7]. However, the high cost and toxicity of RuO₂ limits its practical applications. Therefore, other inexpensive and environmental friendly metal oxides (e.g. MnO₂ [8, 9], TiO₂ [10, 11], NiO₂ [12] and Co₂O₃ [13]) have been investigated as alternatives. Among these candidates, TiO₂ shows many advantages such as low cost, natural abundance and environmentally friendly, making it a very promising electrode material for supercapacitors. However, it suffers from poor electrical conductivity and low specific surface area, leading to a relatively low specific capacitance [14]. To overcome such limitations, nanostructures have been demonstrated to be unique in facilitating the fast ion diffusion and electron transfer, thus dramatically boosting the electrochemical performance. Among different nano-architectures, three dimensional (3D) nanostructures, such as nanorod, nanowire and nanotube arrays, exhibit additional advantages such as providing direct pathways for ion/electron transport and large surface area for charge storage enhancement [15]. Different methods have been reported to synthesize 3D TiO₂ nanostructures including template sol–gel [16], anodic oxidation [17] and chemical vapor deposition methods [18], although instrumental complexity poses high investment costs as well as limits of scalability. In this respect, low cost solution-based methods are of immense interest for inexpensive mass production. Hydrothermal growth is one of simplest and most cost-effective solution-based synthesis methods, which is capable of achieving not only high-purity nanostructures with high-crystallinity, but also can be easily scaled up for large area nanostructures fabrication for various device applications [19, 20].

In this paper, we demonstrate a facile one-step hydrothermal method to grow TiO₂ nanorod arrays directly on Ti metal foil for supercapacitor application. Owing to the unique 3D configuration of TiO₂ nanorod arrays, a decent electrochemical property with high specific capacitance and excellent cycling stability was achieved. In addition, the flexible supercapacitor electrode using Ti foils with TiO₂ nanorod arrays was also demonstrated.

2.1. Synthesis of TiO₂ nanorod arrays

2.2. Material characterizations

2.3. Electrochemical measurements

Figures 1(a) and (b) show typical FESEM images of the as-grown TiO₂ nanorod arrays on Ti foil after the hydrothermal reaction. As seen in figure 1(a), TiO₂ nanorods with clear tetragonal shape were grown from the entire surface of the Ti foil. The inset, enlarged from one nanowire tip, shows pyramid shape with the tetragonal edge arranging from 50 to 200 nm. A cross-sectional view of the TiO₂ nanorod arrays is shown in figure 1(b), revealing the nanorods grew out randomly from the Ti foil with length about 1 μm long. The crystal structure of the nanorods was determined by XRD as shown in figure 1(c) and all diffraction peaks are in good agreement with the tetragonal rutile phase (JCPDS No. 88-1175) except the peaks from the Ti foil, as denoted by asterisks. XPS (figure 1(d)) was further employed to confirm the chemical composition and oxide state of the nanorods through photoelectron peaks of Ti and O. The binding energies of Ti 2p3/2 and Ti 2p1/2 peaks observed at 458.8 eV and 463.4 eV, respectively, with a spin–orbit splitting of 4.6 eV (shown in the supporting information, figure S1), match with the values for bare TiO₂ [21] very well, suggesting that the oxidation state of the Ti in the nanorods is 4.0.

Zoom In Zoom Out Reset image size

The morphology and structures of the TiO₂ nanorods were further characterized by using TEM. Figure 2(a) is the low-magnification TEM image of a single TiO₂ nanorod. The corresponding selected-area electron diffraction patterns (SADPs) taken from [110] (figure 2(b)) and [111] (figure 2(c)) zone axis reveal the single-crystalline feature of the nanorod, which can be indexed to the pure rutile TiO₂ phase. From the SADPs, the growth direction of the TiO₂ nanorod can be easily determined along the [001] direction, which agrees with the tetragonal crystal shape and pyramid tip observed under FESEM. The EDS spectrum shown in figure 2(d) further confirms the successful synthesis of TiO₂ nanorods.

Zoom In Zoom Out Reset image size

To better understand the growth mechanism of TiO₂ nanorods, a cross-sectional TEM specimen of the TiO₂ nanorod arrays was prepared by focused ion beam (FIB). As shown in figure 3(a), a buffer layer between TiO₂ nanorod and Ti substrate can be observed. Note the nanorod is much shorter than average length because it has been broken during the FIB process. The detail EDS spectra from each area can be found in figure S2. The corresponding EDS mapping images of Ti and O distribution are displayed in figures 3(b) and (c), respectively, clearly indicating the formed buffer layer between TiO₂ nanorods and Ti substrate is composed of Ti and O, whose structure will be further characterized below.

Zoom In Zoom Out Reset image size

Figure 4(a) is a low magnification bright-field TEM image, showing a rough Ti/TiO₂ interface. SADPs, performed on different regions of areas b, c and d of the interface, are shown in figures 4(b)–(d), respectively. The diffraction pattern from the substrate (figure 4(b)) can be indexed as pure Ti hexagonal close-packed structure. Besides the diffraction spots from single crystal TiO₂ nanorod, denoted in the small rings in figure 4(c), there is also the presence of diffraction rings in the diffraction pattern from the interface. Such diffraction rings can be indexed as rutile TiO₂ (simulated with the PCED2.0 program) [22], which further confirms the formation of polycrystalline TiO₂ in the buffer layer. Figure 4(d) shows the SADP taken from the area with several small TiO₂ nanorods, in which the diffraction rings can be indexed to the rutile TiO₂ phase, matching EDS mapping results above. Furthermore, no special growth orientation between the nanorod and the substrate was observed. Based on the above experimental results, the possible growth mechanism of the TiO₂ nanorod is detailed as follows.

Zoom In Zoom Out Reset image size

Similar to the reported mechanism of TiO₂ nanorods grown on carbon fibers [23], the Ti foil reacts with H⁺ ions in HCl solution at high temperature and pressure, leading to the gradually dissolving of Ti foil and releasing of the Ti³⁺ ions. Because Ti³⁺ ions are not stable in aqueous solution, they are further hydrolyzed and oxidized to TiO₂. Apparently, due to the big polycrystalline grains of Ti substrate (figure 4(a)) and different crystal structures between the TiO₂ (tetragonal) nanorods and the Ti (hexagonal) substrate, it seems unlikely for TiO₂ nanorods to directly grow out from the Ti surface. Instead, TiO₂ polycrystalline buffer layer was observed to form on the surface of the Ti substrate (figure 3). It is well known that the growth on the existing seeds is more favorable compared to the formation of new nuclei because the heterogeneous nucleation in solution has a lower activation energy barrier than homogeneous nucleation [24]. Hence, the TiO₂ buffer layer can provide initial nucleation sites by reducing the nucleation energy. Since the buffer layer was composed of small randomly oriented TiO₂ crystallites, each nanorod tended to grow randomly, which is consistent with the cross-sectional SEM observation shown in figure 1(b). However, the growth rate of rutile TiO₂ nanorod follows the sequence of (110) < (100) < (101) < (001) [25]. Therefore, TiO₂ nanorods growing along the [001] direction was eventually dominated. The schematic diagram of the formation process is shown in figure S3.

The electrochemical properties of TiO₂ nanorod arrays were investigated by CV and galvanostatic charge/discharge measurements in a conventional three-electrode configuration. Figure 5(a) shows the CV curves of TiO₂ nanorod arrays on Ti foil at different scan rates. All the CV curves exhibit almost rectangular shapes without obvious redox peaks, indicating ideal supercapacitance behavior and fast charging-discharging process based on good charge propagation. The calculated areal capacitances as a function of scan rate are shown in figure 5(b). The capacitance values of the TiO₂ nanorod arrays are in the range of 83.7–856.2 μF cm⁻², as the applied scan rate is varied from 1000 mV s⁻¹ to 1 mV s⁻¹, which is substantially higher than the values recently reported for TiO₂ nanoparticles (33 μF cm⁻² at scan rate of 100 mV s⁻¹) [14] and TiO₂ nanorod arrays on FTO glass (85 μF cm⁻² at scan rate of 100 mV s⁻¹) [26]. The enhanced electrochemical performance of TiO₂ nanorod arrays can be attributed to two major aspects. First, the 3D configuration of single crystalline TiO₂ nanorod arrays provided direct pathways for ion and electron transport as well as large surface area for charge storage. Second, direct growth of TiO₂ nanorod arrays on Ti foil provided good mechanical adhesion and electrical conduction between the active electrode material and the current collector. In addition, heat-treatment improves significantly the electrochemical performance of the TiO₂ nanorod arrays (figure S4) [27]. The electrochemical performance of TiO₂ nanorod arrays were further studied by galvanostatic charge/discharge method. Figure 5(c) illustrates the charge/discharge curves at different current densities of 10, 20, 50 and 100 μA cm⁻². The charge/discharge curves of the TiO₂ electrode show good symmetry and a small internal resistance drop (0.07 V), indicating the TiO₂ nanorod arrays have ideal electrochemical capacitance behavior and good redox reaction reversibility [8, 28].

Zoom In Zoom Out Reset image size

The long-term cycling stability is one of the most important characteristics for supercapacitor. The stability of TiO₂ nanorod arrays was tested by galvanostatic charge/discharge at 100 μA cm⁻² for 1000 cycles, and the results are shown in figure 5(d). The decrease of capacitance after 1000 cycles is less than 1.6%, revealing the TiO₂ nanorod arrays have a superior long-term cyclic performance. Meanwhile, as shown in the insert, the typical charge–discharge curves within ten cycles exhibit almost identical sharp triangular curve shapes, indicating a stable redox reaction, which further confirms the outstanding long-term stability.

EIS analysis was also measured to further understand the fundamental electrochemical behavior of TiO₂ nanorod arrays. The corresponding Nyquist plots are shown in figure 6(a). The impedance plot, as inserted, is composed of a semicircle arc in the high frequency region followed by a linear response in the low frequency. The high frequency region corresponds to the combined resistances of ionic resistance of electrolyte, intrinsic resistance of substrate, and contact resistance at the active material/current collector interface [29]. The linear part in the low frequency region is related to the frequency dependence diffusive resistance of the ion diffusion at electrolyte/electrode interface. It is noticeable that the Nyquist plot shows a large slope at lower frequencies, indicating the high capacitive behavior of the TiO₂ nanorod arrays [30]. Figure 6(b) represents the Mott–Schottky plot observed in the range of positive dc potentials. The correlation between 1/C² and the applied potential exhibits a positive linear dependence, indicating the n-type semiconductor nature. The carrier density calculated from the slope of this Mott–Schottky plot is 2.1 × 10²¹ cm⁻³, which is accordant with the expected values for TiO₂ [31].

Zoom In Zoom Out Reset image size

To evaluate the feasibility of this TiO₂ nanorod arrays directly grown on Ti foil for flexible supercapacitor application, the CV test under different bending conditions were performed. The foil with as-synthesized TiO₂ nanorod arrays shows highly flexibility upon repeated bending (figure 7(a)) and, more importantly, the CV curves collected did not significantly change under different bending angles (figures 7(b) and S5), demonstrating that the TiO₂ nanorod arrays on Ti foil have high mechanical flexibility, which can be potentially applied for flexible supercapacitor fabrication.

Zoom In Zoom Out Reset image size

In summary, a facile one-step hydrothermal method is developed to directly grow TiO₂ nanorod arrays on Ti foil. The TiO₂ nanorods are single rutile crystalline with [001] growth involving dissolving of the Ti substrate and growing of the nanorods. The electrochemical performance of the nanorod arrays, used as supercapacitor electrodes, was systematically investigated. The results demonstrate that TiO₂ nanorod arrays exhibit an ideal capacitive behavior, such as high specific capacitance and excellent cyclic stability. The TiO₂ nanorod arrays was further used as a flexible electrode and showed good stability under different bending conditions, which suggests that TiO₂ nanorod arrays grown directly on Ti foil can be employed for flexible energy storage applications as well.

This research was financially supported by Louisiana Board of Regents under contract No. LEQSF(2011-13)-RD-B-08. Ms Kayla Tarr thanks the support from LA Board of Regents LaSIGMA and NSF(2010-15)-RII-UNO grants.