In situ soft-chemistry synthesis of β-Na_0.33V₂O₅ nanorods as high-performance cathode for lithium-ion batteries

Abstract

β-Na_0.33V₂O₅ nanorods were prepared via a facile soft-chemistry strategy using Na⁺ intercalated (NH₄)_0.5V₂O₅ nanosheets as precursor. Based on X-ray diffraction, Fourier transform infrared spectra and scanning electron microscope analysis, the formation mechanism of β-Na_0.33V₂O₅ nanorods is proposed, which involves cation co-intercalation and crystal structure slip as well as a phase transformation process induced by cation release. When used as cathode for lithium-ion batteries, β-Na_0.33V₂O₅ nanorods calcined at 600 °C exhibited good stable cycling behaviour with high capacity retention of 81.3% after 50 cycles. Reversible discharge capacities of 237.8, 199.8, 183.5, 151.7 mA h g⁻¹ and 110.5 mA h g⁻¹ can be delivered at 30, 60, 150, 300 and 600 mA g⁻¹, respectively. It is expected that the Na_0.33V₂O₅ nanorods could be employed as a promising cathode material in rechargeable lithium-ion batteries.

---

Introduction

Rechargeable lithium-ion batteries (LIBs), which possess various advantages, like high energy density, long cycle life, environmentally benignity and good safety,^1,2 have been widely used as the electrical-storage for portable electronics, hybrid electric vehicles (HEVs) and clean energy sources.^3–5 Currently, commercial cathode materials, such as LiFePO₄,^6–9 deliver relatively low reversible discharge capacity of only ∼170 mA h g⁻¹, which is still insufficient to fulfil the growing demand for large-scale energy-storage application with high power capability and large energy densities. Therefore, the task of exploring alternative cathode materials with higher specific capacity and energy density is a top priority in the lithium-ion battery market.^10,11

Among the promising cathode materials, vanadium-based oxides^12–14 have gained much attention mainly because of their low cost, abundant sources and layered structure combined with a wide range of valence states for the element V (e.g. +3, +4, and +5), which allows high theoretical specific capacity of 442 mA h g⁻¹ for V₂O₅. However, the rate capability and long-term cyclic performance of V₂O₅ are greatly hampered owing to the intrinsic low lithium-ion diffusion coefficient, poor electrical conductivity and the irreversible phase transformation upon deep charge and discharge.^15,16 Introducing additional alkaline or alkali-metal ions into the V₂O₅ interlayer has been utilized to stabilize the structure framework, facilitate Li diffusion and enhance electrochemical properties.^17–23 The resulting materials possess interesting crystal structures such as double-layered or 3D tunnel geometries.

Notably, quasi-one-dimensional (1D) vanadium bronze β-Na_0.33V₂O₅ with a stable and rigid three-dimensional (3D) tunnelled crystal structure can be obtained by doping Na into the V₂O₅ host lattice.^24–27 β-Na_0.33V₂O₅ contains three types of vanadium sites (V1, V2 and V3) and three different Li intercalation sites (M1, M2 and M3) within the tunnel, endowing the compound with excellent structural reversibility upon high lithium uptake. As reported,²⁸ the β-Na_0.33V₂O₅ host lattice shows no phase transition within a fairly wide Li content range of 0.0 ≤ x ≤ 1.66 and delivers an initial specific capacity of up to 230 mA h g⁻¹. Recently, Liang et al.²⁹ synthesized mesoporous β-Na_0.33V₂O₅ via a typical hydrothermal reaction, which delivered a high initial capacity of 339 mA h g⁻¹ at 20 mA g⁻¹ yet poor cycling stability with a capacity of 168 mA h g⁻¹ after 35 cycles. Tan et al.³⁰ prepared hierarchical β-Na_0.33V₂O₅ microspheres by a solvothermal method, which exhibited an initial specific capacity of 157 mA h g⁻¹ but that decreased to 111 mA g⁻¹ after 35 cycles at a current density of 1000 mA g⁻¹. In addition to the conventional hydrothermal/solvothermal routes based on a dissolution–crystallization mechanism, the soft-chemical method^31,32 is a popular and suitable approach in the design of functional inorganic nanomaterials with specific crystal structure as well as controllable chemical composition, typically involving morphology and phase transformation from templated precursor with insertion of certain ions or molecules to target product with desirable crystal structure under given conditions, and includes hydrothermal/solvothermal or heat treatment. For example, Kong et al.³³ adopted an in situ soft-chemical way to prepare 1D bead-like AgVO₃ nanoarchitectures from a layered K₂V₆O₁₆·2.7H₂O plate-like particle precursor at room-temperature; the material they produced exhibited a higher discharge capacity of 127 mA h g⁻¹ at current density of 100 mA g⁻¹. However, the synthesis of 3D tunnelled metal vanadium oxide bronze, e.g. β-Na_0.33V₂O₅, through an in situ soft-chemistry process for use as cathode for LIBs has not been reported yet.

Herein, a simple soft-chemistry route, involving a hydrothermal process and subsequent phase transformation, was employed for the first time, we believe, to prepare β-Na_0.33V₂O₅ single crystalline nanorods. The formation mechanism of β-Na_0.33V₂O₅ nanorods is proposed based on morphology and structure evolution with annealing temperature. In addition, the electrochemical properties of β-Na_0.33V₂O₅ were evaluated by galvanostatic charge–discharge test. The as-prepared β-Na_0.33V₂O₅ annealed at 600 °C exhibited enhanced electrochemical performance with high specific capacity, good cycling stability and rate capability, which can be attributed to its excellent structural reversibility upon repetitive electrochemical cycling.

Experimental section

β-Na_0.33V₂O₅ nanorods were synthesized via a facile soft-chemistry method using (NH₄)_0.5V₂O₅ nanosheets as precursor template, followed by calcination. First, 0.635 g NH₄VO₃ was dissolved in 25 mL deionized water under magnetic stirring at 80 °C; then stoichiometric amounts of oxalic acid and NaCl were added. The obtained clear solution was transferred into a 50 mL Teflon-lined stainless steel container and heated in an electrical oven at 180 °C for 12 h. The resulting precipitate was collected by centrifugation, washed with pure ethanol and dried at 60 °C overnight. Finally, the obtained precursor was annealed from 400 °C to 650 °C under Ar atmosphere with a ramping rate of 3 °C min⁻¹. The calcination samples are designated NVO400, NVO450, NVO500, NVO550, NVO600 and NVO650, respectively. For comparative study, (NH₄)_0.5V₂O₅ nanosheets were also prepared as described above but without the addition of NaCl.

The change in mass of the precursor during heat treatment in Ar atmosphere was studied by thermogravimetric analysis (TG, Netzsch STA449C, Germany) ranging from 25 °C to 650 °C at a ramping rate of 10 °C min⁻¹. The morphology and microstructure of the as-prepared products were characterized by field-emission scanning electron microscopy (FE-SEM, FEI Quanta 200F) and transmission electron microscopy (TEM, JEM 2100, 200 kV). The crystallinity and phase structure of the final products were recorded by power X-ray diffraction (XRD, Rigaku D/Max2500, Cu Kα radiation, λ = 1.5406 Å). The chemical bond structure was analysed by Fourier transform infrared spectrometry (FT-IR, Thermo Scientific Nicolet iS10). The element valency analysis of the products was done by X-ray photo-electron spectrometry (XPS, Thermo Fisher Scientific, Al Kα radiation).

To fabricate the working electrode, active materials, Super P and polyvinylidene fluoride (PVDF) in a weight ratio of 80 : 10 : 10 were mixed and coated onto aluminium foil, which was dried at 120 °C and punched into round discs with a diameter of 14 mm. Lithium metal foil served as the counter/reference electrode, porous polypropylene membrane was used as the separator and 1 M LiPF₆ in ethylene carbonate/dimethyl carbonate/ethyl methyl carbonate solvent (EC/DMC/EMC) (1 : 1 : 1 v/v/v) acted as the electrolyte. The 2025 type coin-cells assembly was operated in an Ar-filled glovebox (MBraun, Germany) with H₂O and O₂ contents less than 5 ppm. The cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) measurements were conducted on a CHI 660E electrochemical station at a scan rate of 0.1 mV s⁻¹ and over the frequency range of 100 kHz to 0.01 Hz by applying an AC signal with amplitude of 5 mV. The galvanostatic discharge/charge experiment was performed on the Land CT 2001A Battery Test system (Jinnuo Wuhan Corp., PR China). All the electrochemical tests were carried out in the potential window 1.5–4.0 V (vs. Li/Li⁺).

Results and discussion

XRD patterns of the obtained precursor and products of calcination at 400 °C and 450 °C are given in Fig. S1.† As can be seen, the precursor mainly displays the monoclinic (NH₄)_0.5V₂O₅ phase (space group P2₁/m, JCPDS no. 31-0075). The intense peak at 9.1° corresponds to the (001) plane of the (NH₄)_0.5V₂O₅ phase due to the external cation co-intercalation into the lamellar V₂O₅-based structure during the hydrothermal process. The open [VO₅] layered frameworks can easily accommodate small guest cations like Na⁺, H₃O⁺ and NH₄⁺ with modest volume changes.³⁴ Thus, the layered framework structure of the precursor is well-preserved when alkali metal ions are incorporated. Herein, the obtained layered precursor is defined as the Na⁺ intercalated (NH₄)_0.5V₂O₅ precursor. After the thermal treatment of the layered precursor at 400 °C, the phase structure of the product was almost unchanged with slight (001) plane shifts towards lower angles, which are likely to be induced by the release of crystallization water and NH₄⁺ species from crystal layers. For NVO450, the strong peak intensity of the (001) plane decreased, along with the appearance of a new peak around 12.2°, which corresponds to the (002) reflection of β-Na_0.33V₂O₅, suggesting the phase transformation from layered Na⁺ intercalated (NH₄)_0.5V₂O₅ precursor to β-Na_0.33V₂O₅ product starts around 450 °C.

In order to better understand the weight loss process during heat treatment, TG of the precursor was carried out (as shown in Fig. S2†). In the temperature range from 200 to 420 °C, an obvious mass loss of around 5.6% was observed on the TG curve, which can be mainly ascribed to the removal of physically absorbed/chemically bonded water, as well as the release of NH₄⁺ inorganic cations from the hydrothermal products. Thereafter, the TG curve tends to be flat above 450 °C, suggesting that the guest cation escape occurred below 450 °C and phase transformation takes place above 450 °C with no mass loss.

To make a qualitative observation on the morphology transformation with annealing temperature, SEM images of the precursor with and without thermal treatment are shown in Fig. 1. It can be seen that both the (NH₄)_0.5V₂O₅ and Na⁺ intercalated (NH₄)_0.5V₂O₅ precursor (Fig. 1a and b) consisted of agglomerated flake-like nanosheets with an average length of about 2 μm and width around 1 μm, suggesting the incorporation of Na⁺ into (NH₄)_0.5V₂O₅ did not influence the morphology of layered precursor mainly because of the similar ionic radius of Na⁺ and NH₄⁺. Upon annealing at low temperatures of 400 and 450 °C, the sample was still composed of well-dispersed nanoflakes as shown in Fig. 1c and d, indicating the layered structure was still retained when the calcination temperature was below 450 °C. On further increasing the calcination temperature to 500 °C, the separated nanosheets merged together and generated rod-like structure as shown in Fig. 1e. At the higher temperature of 600 °C, the nanosheet morphology disappeared completely and distinct nanorods (Fig. 1g) formed with widths of ∼200–500 nm and length around 1 μm, while the particle size of the sample annealed at 650 °C grew significantly to around 1 μm in width and 2 μm in length with poor homogeneity (Fig. 1h). As seen, when the calcination temperature was above 500 °C, the nanosheet precursor was transformed into nanorods accompanied by phase transformation.

Fig. 1 SEM images of the precursor and products calcined at different temperatures. (a) (NH₄)_0.5V₂O₅; (b) the Na⁺ intercalated (NH₄)_0.5V₂O₅ precursor; (c) NVO400; (d) NVO450; (e) NVO500; (f) NVO550; (g) NVO600; (h) NVO650.

XRD patterns of samples annealed at temperatures above 500 °C are presented in Fig. 2. All the XRD diffraction patterns can be readily assigned to the monoclinic β-Na_0.33V₂O₅ bronze phase (space group: A2/m, JCPDS no. 86-0120). The (NH₄)_0.5V₂O₅ phase disappeared with the generation of the new phase, suggesting the precursor was thoroughly transformed into the β-Na_0.33V₂O₅ phase when calcined above 500 °C. With increasing temperature, the intensity of the diffraction peaks was clearly enhanced, demonstrating the improved crystallinity of the samples treated at higher annealing temperatures.

Fig. 2 XRD patterns of the β-Na_0.33V₂O₅ samples calcined at temperatures of 500–650 °C.

Structural information regarding the obtained precursor and the calcined products was further provided by FT-IR spectra, as shown in Fig. 3. For the precursor, there are only two absorption bands, located at 996.1 and 540.0 cm⁻¹, which are ascribed to the stretching vibration of V=O bonds and edge-sharing V–O bonds, respectively.³⁵ Upon calcination at 400 and 450 °C, the new band at 877.0 cm⁻¹ can be associated with the asymmetric stretching vibrations of the V–O–V unit, which are sensitive to the NH₄⁺ ion extraction with the content variation of V⁴⁺ and V⁵⁺ in the sodium vanadium compound and the change of V–O bond length as a consequence. When calcination is at higher temperature, the band around 996.1 cm⁻¹ splits into three peaks located at 939.2, 961.4 and 996.1 cm⁻¹, which suggest the formation of three disordered V–O crystal groups, and probably originate from V=O vibrations for distorted octahedral and distorted square pyramids, respectively.

Fig. 3 FT-IR curves of the precursor and the products of calcination at different temperatures.

TEM and selected-area electron diffraction (SAED) analysis were carried out to investigate the detailed structural characteristics of β-Na_0.33V₂O₅ nanorods obtained at 600 °C. As shown in Fig. 4a, the nanorods are ∼200–500 nm in width and ∼1–2 μm in length. The clear lattice fringes with interplanar spacing of ∼0.32 nm correspond well to the distances between (111) lattice planes of the monoclinic β-Na_0.33V₂O₅ phase, and the clear diffraction spots in the SAED pattern (inset in Fig. 4b) further reveal the nanorods are single-crystalline.

Fig. 4 (a) TEM image of β-Na_0.33V₂O₅ nanorods calcined at 600 °C. (b) High resolution TEM image of a nanorod taken from the area indicated by a rectangle in (a). Inset in (b) is the corresponding SAED pattern.

To confirm the chemical compositions and element valence state in the NVO600 sample, XPS was performed, as shown in Fig. 5. The binding energy obtained in the XPS analysis was calibrated with the C 1s peak at 284.6 eV as reference binding energy. A series of peaks for C 1s, V 2p and O 1s located at around 284.7, 516.7 and 530.1 eV are clearly observed on the full range XPS spectrum in Fig. 5a. The Na 1s spectrum at 1071.2 eV further confirms the existence of Na atoms in 3D tunnelled β-Na_0.33V₂O₅ lattice as shown in Fig. 5b. The high resolution XPS spectrum of V 2p displayed two peaks at 515.8 and 522.9 eV (Fig. 5d), corresponding to V 2p_3/2 and V 2p_1/2 of V⁵⁺, respectively. Notably, the V 2p_3/2 core peak can be deconvoluted into two components, V 2p_3/2 (IV) and V 2p_3/2 (V), denoting the mixed valences of V⁵⁺ and V⁴⁺ in ternary vanadium oxide bronzes.

Fig. 5 XPS spectra of the NVO600 sample: (a) survey; (b) Na 1s; (c) O 1s; (d) V 2p.

On the basis of the above discussion, a phase transformation mechanism from layered precursor to 3D tunnelled β-Na_0.33V₂O₅ bronze is proposed as shown in Fig. 6. After hydrothermal treatment, Na⁺ and NH₄⁺ along with water molecules are embedded between V₂O₅ polyhedral layers, where some NH₄⁺ sites are occupied by Na⁺ ions owing to their similar ionic radius, giving rise to a layered crystal structure in (NH₄)_0.5V₂O₅. Herein, the Na⁺ and NH₄⁺ are conceived to be distributed between V–O polyhedral layers along the b axis as shown schematically in Fig. 6a. Upon calcination at temperatures below 450 °C, the inter-layer water and NH₃ molecules are removed, with only Na⁺ remaining between VO₅ double layers, and the layered structure is preserved with subtle altering of V–O bond length, as shown in Fig. 6b. The subsequent high annealing temperature initiates the crystallographic slip and reconstruction process, which involves VO₅ layers slipping along the a axis and condensation along the c axis. As a result, double layers of VO₆ octahedra are reconstructed, along with the breaking of V–O bonds and generation of VO₅ square pyramids. The 3D tunnel-like structure finally obtained is presented in Fig. 6c.

Fig. 6 Schematic mechanism for formation of 3D tunnelled β-Na_0.33V₂O₅ nanorods from layered precursor. Projection of the crystal structure of (a) Na⁺ intercalated (NH₄)_0.5V₂O₅ precursor; (b) Na⁺ intercalated precursor after the remove of H₂O and NH₃; (c) 3D tunnelled β-Na_0.33V₂O₅ on a–c plane.

Fig. 7 compares the charge–discharge curves of β-Na_0.33V₂O₅ nanorods annealed at different temperatures at 60 mA g⁻¹ in the potential window of 1.5–4.0 V (vs. Li/Li⁺). It can be seen that the discharge voltage platform of NVO500 around 2.4–3.0 V fades quickly as cycling proceeds, which is mainly due to its low degree of crystallinity. For NVO600 nanorods, four well-defined voltage plateaus (3.3, 2.8, 2.5 and 2.0 V) can be observed clearly, which are strongly connected with progressive Li accommodation in the available interstitial sites (M3, M2, and M1) in the β-Na_0.33V₂O₅ 3D tunnelled structure. The correlation of the working potential with successive Li⁺ filling of tunnel sites M in the β-Na_0.33V₂O₅ structure has been made previously by Pereira-Ramos et al.³⁶ by thermodynamic and kinetic investigations, and further verified by R. Baddour-Hadjean et al.²⁸ using XRD and Raman probes on a lithiated β-Na_0.33V₂O₅ electrode over the wide composition range 0 < x ≤ 1.66. The M3 sites are occupied by Li⁺ preferentially when the voltage reaches 3.3 V with 0 < x ≤ 0.33. The half-occupancy of M2 sites by the Li ions occurs at the second voltage of around 2.8 V with 0.33 < x ≤ 0.66. And the complete Li⁺ filling of the remaining M1, M2, and M3 sites corresponds to the large voltage plateau at 2.5 V with 0.66 < x ≤ 1.67. In addition, it can be seen that all these plateaus are well-maintained after 50 charge–discharge cycles, demonstrating a very stable host lattice and desirable structural reversibility. The good electrochemical performance of the NVO600 electrode can be mainly attributed to its high degree of crystallinity and well-defined 1D nanosized morphology, which provides a fast transfer channel for electrons and shorter diffusion path for Li insertion/extraction. However, increasing the calcination temperature to 650 °C leads to higher crystallinity but larger size of nanorods; the specific discharge capacity of NVO650 is lower than that of NVO600, which is consistent with previous reports that larger particle size will prolong the Li diffusion distance. As seen, the balance between crystallinity and particle size plays a vital role in improving the electrochemical performance of the β-Na_0.33V₂O₅ material.

Fig. 7 Charge–discharge curves of β-Na_0.33V₂O₅ nanorods at 60 mA g⁻¹. (a) NVO500; (b) NVO550; (c) NVO600; (d) NVO650.

In order to investigate the reaction mechanism of β-Na_0.33V₂O₅ with Li⁺, the as-prepared NVO electrode and those cycled twice and 50 times under charge state were characterized using SEM, EDX and XPS, and the results are shown in Fig. S3.† The β-Na_0.33V₂O₅ electrodes were disassembled in an Ar-filled glovebox and washed with DMC for XPS and SEM-EDX measurements. The presence of Na is evident even after 50 cycles. Although EDX is not good for quantitative composition analysis, based on XPS analysis it is calculated that the ratio of Na/V does not change upon charge–discharge testing. It has been shown that Na ions are immobile in β-Na_0.33V₂O₅ during the charge–discharge process. In this work, the NVO600 electrode was discharged first. Based on the ex situ XPS and EDX results, the reversible Li⁺ insertion/extraction reaction in β-Na_0.33V₂O₅ can be described as follows:

Na_0.33V₂O₅ + xLi⁺ + xe⁻ ↔ Li_xNa_0.33V₂O₅

Upon initial discharge, Li⁺ is intercalated in the β-Na_0.33V₂O₅ host and all the vacant tunnel sites are filled by Li⁺ and Na⁺. In the next charge process, the Li⁺ is extracted from Li_xNa_0.33V₂O₅, while the larger alkaline cation Na⁺ is trapped in the β structure during the delithiation step, which might be governed by thermodynamic and kinetic diffusion.³⁶

The cycling performance of NVO500, NVO550, NVO600 and NVO650 electrodes at current density of 60 mA g⁻¹ is compared in Fig. 8a. The NVO500, NVO550, NVO600 and NVO650 electrodes deliver initial discharge capacities of 303.0, 241.1, 223.9, and 202.7 mA h g⁻¹, respectively. As seen from Fig. 8a, the NVO500 cathode exhibits higher specific discharge capacity for the initial 10 cycles than the other samples treated at higher temperatures. However, the capacity of the NVO500 electrode degrades seriously and only a discharge capacity of 152.3 mA h g⁻¹ remains after 50 cycles. The capacity decay occurring on NVO500 can be ascribed to its relatively low crystallinity, which is beneficial for delivering high initial capacity but has adverse effects on the cyclability performance. In order to get superior electrochemical performance, higher annealing temperature is applied and the cycling stability of NVO550, NVO600 and NVO650 is much improved when compared with the NVO500 cathode. After 50 cycles, the specific discharge capacity of NVO550, NVO600 and NVO650 decreases to 110.2, 182.1 and 163.0 mA h g⁻¹, and the fading ratios are ca. 1.09%, 0.37% and 0.39% per cycle, respectively. The relatively lower capacity of NVO650 under higher annealing temperature is likely to be related to the larger-sized particles, which will increase the Li⁺ diffusion distance and be unfavourable to achieving higher discharge capacity. In our work, the NVO600 electrode possesses the best electrochemical performance, with relatively high capacity and good cycling stability, which is mainly due to the optimal balance between size and crystallinity. A comparison of the electrochemical properties of our β-Na_0.33V₂O₅ electrode with those in other works is presented in Table S1.† Despite the relatively lower specific capacity, the cycling stability of the NVO600 electrode is better than the results reported by several researchers.^29,30

Fig. 8 (a) Cycling performance of β-Na_0.33V₂O₅ at a current density of 60 mA g⁻¹. (b) Rate capability of β-Na_0.33V₂O₅ electrodes cycling at different current densities. (c) CV curves of NVO600 nanorods at a scan rate of 0.1 mV s⁻¹. (d) Nyquist plots of NVO600 electrodes at different cycles (1^st, 10^th, 20^th, 50^th). Solid lines are fitting results.

The rate capability of the NVO600 electrode at various current densities from 30 to 600 mA g⁻¹ is displayed in Fig. 8b. A specific discharge capacity of 237.8 mA h g⁻¹ can be obtained for the NVO600 electrode at a current density of 30 mA g⁻¹, and discharge capacities of 199.8, 183.5, and 151.7 mA h g⁻¹ are retained at the higher current densities of 60, 150 and 300 mA g⁻¹, respectively. Even at 600 mA g⁻¹, the NVO600 sample can still deliver a satisfactory discharge capacity of 110.5 mA h g⁻¹. More importantly, a high specific discharge capacity of 208.0 mA h g⁻¹ can be recovered when the current density is returned to 30 mA g⁻¹. As shown in Fig. 8b, almost no capacity fading is observed under each current density for 10 cycles. The superior rate performance of the pure-phase β-Na_0.33V₂O₅ nanorods without any carbon materials incorporated can be attributed to their high crystallinity and remarkable structure stability.

Typical cyclic voltammograms (CVs) of the NVO600 electrode at a scan rate of 0.1 mV s⁻¹ between 1.5 and 4.0 V vs. Li/Li⁺ are shown in Fig. 8c. Four strong cathodic peaks located at 3.2, 2.9, 2.4 and 1.8 V vs. Li/Li⁺ are clearly observed in the first cathodic process, which involves the specific lithium intercalation steps into the β-Na_0.33V₂O₅ electrode.^37,38 In the β-Na_0.33V₂O₅ 3D tunnelled structure, the [VO₆] octahedra double chains are linked by edge-sharing [VO₅] pyramids along the b axis, which contain three distinct types of tunnel sites available for Li⁺ occupation per unit cell, namely, interstitial equivalent sites (M1), eight-coordinated sites (M2) and tetrahedral sites (M3). The pre-inserted 0.33Na⁺ occupy half of the M1 positions. During the cathodic scan, the current peak around 3.2 V is ascribed to Li⁺ accommodation in M3 sites. The second cathodic peak around 2.9 V corresponds to Li⁺ half occupancy of the M2 sites. The cathodic peak located at 2.4 V is attributed to continuous Li⁺ filling of the remaining M1, M2, and M3 vacant sites. For V₂O₅, an irreversible phase is formed when the discharge voltage is below 2.0 V with Li⁺ successive insertion, as seen from the charge–discharge curves of the V₂O₅ electrode provided in Fig. S4.† For β-Na_0.33V₂O₅, the cathodic peak around 1.8 V in the CV curves is highly overlapped, further implying the good reversibility of the β-sodium vanadium bronze electrode. Correspondingly, four distinct anodic peaks located at 2.3, 2.9, 3.0 and 3.3 V vs. Li/Li⁺ for the first anodic process are observed, which are attributed to the deintercalation of the lithium ions from the β-Na_0.33V₂O₅ crystal lattice. In the following cycles, there is a minor difference in the intensity and position of redox peaks on the CV curves, suggesting highly reversible lithium insertion and extraction processes for NVO600 electrode. In addition, the redox peaks in the CV curves are in good accordance with the discharge–charge profiles displayed in Fig. 7c.

In order to investigate the transport kinetics of the obtained β-Na_0.33V₂O₅ nanorods, the electrochemical impedance spectra of cycled NVO600 electrodes after different cycles are measured, as shown in Fig. 8d. The impedance spectra mainly consist of a depressed semicircle in the middle- to high-frequency region and a straight slope line in the low-frequency region. The semicircle is associated with the electrochemical charge-transfer process. The straight line is related to the diffusion of Li⁺ in the electrode bulk material. The corresponding equivalent circuit model in Fig. S5† was employed to acquire the fitting data. R_e is the ohmic resistance of the involved electrolyte, current collector and separator. R_ct is the charge-transfer resistance. CPE_dl and Z_W are the double-layer capacitance and the Warburg impedance, respectively. The simulation parameters of electrochemical impedance spectra are presented in Table S2.† As seen, the charge-transfer resistance values for the electrode after the 1^st and 10^th discharge cycles are 135.8 and 223.7 Ω, respectively. When the electrode has undergone 50 cycles, the charge transfer resistance is slightly increased, to 263.4 Ω. The small charge-transfer resistance variation can be attributed to the excellent structure stability of the β-Na_0.33V₂O₅ bronze cathode during the cycling processes.

Conclusions

In summary, highly crystalline β-Na_0.33V₂O₅ nanorods are fabricated by a facile soft-chemistry route. The annealing temperature has a great effect on the morphology, phase crystallinity and electrochemical properties of the β-Na_0.33V₂O₅ product. The β-Na_0.33V₂O₅ nanorods calcined at 600 °C exhibit the highest reversibility capacity, and the best cycling stability and rate performance, delivering the desirable discharge capacity of 223.9 mA h g⁻¹ recorded at 60 mA g⁻¹, with high capacity retention of 81.3% after 50 cycles. Moreover, the β-Na_0.33V₂O₅ can maintain a discharge capacity of 110.5 mA h g⁻¹ even at the high current density of 600 mA g⁻¹. Based on its excellent rate performance, β-Na_0.33V₂O₅ is a promising cathode candidate for high-power lithium-ion batteries. The soft-chemistry method reported here might be generalizable to the synthesis of other metal vanadium bronzes.

Acknowledgements

This work was supported by the Program for New Century Excellent Talents in University, Ministry of Education, China (NCET-11-0810). C. Y. X. acknowledges the Fundamental Research Funds for the Central Universities (HIT.BRETIII.201203) and the China Postdoctoral Science Foundation (200902381).

References

1. M. Armand and J. M. Tarascon, Nature, 2008, 451, 652–657.
2. V. Etacheri, R. Marom, R. Elazari, G. Salitra and D. Aurbach, Energy Environ. Sci., 2011, 4, 3243–3262.
3. B. Scrosati and J. Garche, J. Power Sources, 2010, 195, 2419–2430.
4. B. Scrosati, J. Hassoun and Y. K. Sun, Energy Environ. Sci., 2011, 4, 3287–3295.
5. M. Stanley Whittingham, Chem. Rev., 2004, 104, 4271–4301.
6. W. J. Zhang, J. Power Sources, 2011, 196, 2962–2970.
7. T. F. Liu, L. Zhao, J. S. Zhu, B. Wang, C. F. Guo and D. L. Wang, J. Mater. Chem. A, 2014, 2, 2822–2829.
8. C. W. Sun, S. Rajasekhara, J. B. Goodenough and F. Zhou, J. Am. Chem. Soc., 2011, 133, 2132–2135.
9. H. M. Xie, R. S. Wang, J. R. Ying, L. Y. Zhang, A. F. Jalbout, H. Y. Yu, G. L. Yang, X. M. Pan and Z. M. Su, Adv. Mater., 2006, 18, 2609–2626.
10. B. Xu, D. N. Qian, Z. Y. Wang and Y. S. Meng, Mater. Sci. Eng., R, 2012, 73, 51–65.
11. K. X. Wang, X. H. Li and J. S. Chen, Adv. Mater., 2015, 27, 527–545.
12. Y. Q. Qiao, X. L. Wang, J. P. Zhou, J. Zhang, C. D. Gu and J. P. Tu, J. Power Sources, 2012, 198, 287–293.
13. P. Mei, X. L. Wu, H. M. Xie, L. Q. Sun, Y. P. Zeng, J. P. Zhang, L. H. Tai, X. Guo, L. N. Cong, S. C. Ma, C. Yao and R. S. Wang, RSC Adv., 2014, 4, 25494–25501.
14. A. Q. Pan, H. B. Wu, L. Zhang and X. W. Lou, Energy Environ. Sci., 2013, 6, 1476–1479.
15. A. Q. Pan, J. G. Zhang, Z. M. Nie, G. Z. Cao, B. W. Arey, G. S. Li, S. Q. Liang and J. Liu, J. Mater. Chem., 2010, 20, 9193–9199.
16. M. Winter, J. O. Besenhard, M. E. Spahr and P. Novák, Adv. Mater., 1998, 10, 725–763.
17. K. West, B. Zachau-Christiansen and T. Jacobsen, Solid State Ionics, 1998, 28–30, 1128–1131.
18. P. P. Wang, C. Y. Xu, W. D. Li, L. Wang and L. Zhen, Electrochim. Acta, 2015, 169, 440–446.
19. Y. G. Tang, D. Sun, H. Y. Wang, X. B. Huang, H. Zhang, S. Q. Liu and Y. N. Liu, RSC Adv., 2014, 4, 8328–8334.
20. W. D. Li, C. Y. Xu, X. L. Pan, Y. D. Huang and L. Zhen, J. Mater. Chem. A, 2013, 1, 5361–5370.
21. H. Y. Wang, K. L. Huang, Y. Ren, X. B. Huang, S. Q. Liu and W. J. Wang, J. Power Sources, 2011, 196, 9786–9791.
22. S. Bacha, A. Boudaoud, N. Emery, R. Baddour-Hadjean and J. P. Pereira-Ramos, Electrochim. Acta, 2014, 119, 38–42.
23. Y. L. Zhao, C. H. Han, J. W. Yang, J. Su, X. M. Xu, S. Li, L. Xu, R. P. Fang, H. Jiang, X. D. Zou, B. Song, L. Q. Mai and Q. J. Zhang, Nano Lett., 2015, 15(3), 2180–2185.
24. E. Khoo, J. M. Wang, J. Ma and P. S. Lee, J. Mater. Chem., 2010, 20, 8368–8374.
25. J. K. Kim, B. Senthilkumar, S. H. Sahgong, J. H. Kim, M. F. Chi and Y. Kim, ACS Appl. Mater. Interfaces, 2015, 7, 7025–7032.
26. H. M. Liu, Y. G. Wang, L. Li, K. X. Wang, E. Hosono and H. S. Zhou, J. Mater. Chem., 2009, 19, 7885–7891.
27. Y. K. Lu, J. Wu, J. Liu, M. Lei, S. S. Tang, P. J. Lu, L. Y. Yang, H. R. Yang and Q. Yang, ACS Appl. Mater. Interfaces, 2015, 7, 17433–17440.
28. R. Baddour-Hadjean, S. Bach, N. Emery and J. P. Pereira-Ramos, J. Mater. Chem., 2011, 21, 11296–11305.
29. S. Q. Liang, J. Zhou, G. Z. Fang, C. Zhang, J. Wu, Y. Tang and A. Q. Pan, Electrochim. Acta, 2014, 130, 119–126.
30. Q. G. Tan, Q. Y. Zhu, A. Q. Pan, A. P. Wang, Y. Tang, X. P. Tan, S. Q. Liang and G. Z. Cao, CrystEngComm, 2015, 17, 4774–4780.
31. C. Antonyraj, D. Srivastava, G. Mane, S. Sankaranarayanan, A. Vinu and K. Srinivasan, J. Mater. Chem. A, 2014, 2, 6301–6304.
32. M. Niederberger, N. Pinna, J. Polleux and M. Antonietti, Angew. Chem., Int. Ed., 2004, 43, 2270–2273.
33. X. G. Kong, Z. L. Guo, C. B. Zeng, J. F. Huang, L. Y. Cao, L. Li, L. X. Yin, P. H. Wen, Q. Feng and Z. W. Xu, J. Mater. Chem. A, 2015, 3, 18127–18135.
34. D. Sun, G. H. Jin, H. Y. Wang, P. Liu, Y. Ren, Y. F. Jiang, Y. G. Tang and X. B. Huang, J. Mater. Chem. A, 2014, 2, 12999–13005.
35. G. T. Chandrappa, P. Chithaiah, S. Ashoka and J. Livage, Inorg. Chem., 2011, 50, 7421–7428.
36. S. Bach, J. P. Pereira-Ramos, N. Baffier and R. Messina, J. Electrochem. Soc., 1990, 137, 1042–1048.
37. W. D. Li, C. Y. Xu, Y. Du, H. T. Fang, Y. J. Feng and L. Zhen, J. Electrochem. Soc., 2014, 161(1), A75–A83.
38. I. Seo, G. C. Hwang, J. K. Kim and Y. Kim, Electrochim. Acta, 2016, 193, 160–165.

---

Footnote

† Electronic supplementary information (ESI) available: XRD patterns of (NH₄)_0.5V₂O₅, Na-intercalated precursor and calcination products at 400 and 450 °C. TG curve of Na-intercalated precursor. SEM images, EDX and XPS spectra of NVO600 electrode in different states: (a and b) before cycling; (c and d) 2^nd charge state; (e and f) 50^th charge state. (g) Comparison of Na 1s spectra. Charge–discharge curves of the V₂O₅ electrode. Comparison of electrochemical performance for β-Na_0.33V₂O₅ cathode. EIS fitted results for NVO600 electrode after different cycles. See DOI: 10.1039/c6ra23484d

---