Self-standing Bi₂O₃ nanoparticles/carbon nanofiber hybrid films as a binder-free anode for flexible sodium-ion batteries

Abstract

A flexible binder-free film composed of bismuth oxide nanoparticles embedded in carbon nanofibers (Bi₂O₃/C) was prepared by a feasible electrospinning method and directly used as a sodium ion battery (SIB) anode. As a binder-free and flexible anode for SIBs, Bi₂O₃/C delivers a high reversible capacity of 430 mA h g⁻¹ after 200 cycles at a current density of 100 mA g⁻¹ and an exceptional rate capability of 230 mA h g⁻¹ at 3200 mA g⁻¹. It has a stable capacity of 252 mA h g⁻¹ after 50 cycles at 400 mA g⁻¹ in a Na-ion full cell device. The high capacity, good cyclability and rate capability are attributed to synergistic effects of the uniform distribution of ultra-small Bi₂O₃ nanoparticles (≈10 nm) in the carbon nanofibers and the conducting framework of 3-D interconnected carbon nanofibers, which can effectively alleviate the volume expansion during sodiation/desodiation processes and maintain the high electrical conductivity throughout the electrode. This self-standing flexible Bi₂O₃/C nanocomposite electrode may hold great promise for high-performance SIBs.

---

Introduction

Recently, sodium ion batteries (SIBs) have attracted great interest owing to their low cost and the natural abundance of Na sources compared to lithium ion batteries (LIBs).^1–6 Although the fundamental principle of the electrochemical reaction of sodium is similar to that of lithium, the Na-ion (0.106 nm and 23 g mol⁻¹) has a larger ionic radius and molar mass than the Li-ion (0.076 nm and 7 g mol⁻¹) which makes it difficult to explore suitable and stable sodium storage materials. In particular, appropriate anode materials with high capacity and excellent cycling performance are still limited.⁷ As a successful strategy, metal oxides (MO), metal sulphides (MS) and metal selenides (MSe) with conversion reactions have been developed as new anode materials for SIBs, owing to their high gravimetric specific capacity (M = Sn, Sb, Bi).^8–10 Among them, bismuth oxide (Bi₂O₃) is an important semiconductor with a band gap of 2.8 eV and only a few research studies about Bi₂O₃ as an anode material for SIBs have been reported. Bi₂O₃ is an attractive candidate due to its stable chemical properties, environmental friendliness, high theoretical capacity (669.8 mA h g⁻¹) and volumetric capacity (5961.2 mA h cm⁻³).^11–16 However, electrochemical conversion reactions are usually accompanied by large volume variation, continuous solid electrolyte interface (SEI) reestablishment and structure collapse for the electrode materials, leading to a rapid deterioration of the capacity for SIBs.¹⁷

Several approaches have been used to address the drawbacks and great efforts have been made to fabricate nanoscale Bi₂O₃ to moderate the absolute strain generated upon sodiation/desodiation processes.¹⁸ It is noteworthy that the effect of particle size is very significant for both basic research and industrial manufacture. Although nanomaterials can definitely improve the Na-storage ability, the electrochemistry performance is still far away from meeting large-scale applications because of aggregation of nanoparticles with high surface energy and the rate performance is also unsatisfactory due to the poor conductivity of Bi₂O₃. Dispersing nanoparticles in a carbon matrix could significantly restrain the particle agglomeration, accommodate the volume expansion, and enhance the electric conductivity.^19–24 For example, Sun's group successfully fabricated Bi₂O₃/graphene composites (Bi₂O₃ size ≈ 100 nm, content 16.6 wt%) using aerosol spray pyrolysis, which displayed a discharge capacity of 440 mA h g⁻¹ after 100 cycles at 200 mA g⁻¹ for LIBs. However, a Bi₂O₃@carbon composite which simultaneously controls the Bi₂O₃ particle size to less than 100 nm and allows high Bi₂O₃ loading mass has not been reported yet. Considerable improvements in the design and optimization of the Bi₂O₃@carbon nanostructure are still required.

1-D carbon nanofibers can effectively enhance the ionic diffusion length and electronic transport along their geometry, leading to fast diffusion kinetics.^25–30 Electrospinning is a feasible technique to prepare various 1-D carbon-containing composites.^31–34 This method can also produce a flexible and free-standing film, which can be directly used as a working electrode without auxiliary additives (such as carbon black and binders). As is well-known, organic polymeric binders would result in poor electronic conductivity and irreversible capacity fading. Synthesis of flexible and binder-free anode materials for rechargeable SIBs can not only benefit the electrochemical performance, but also further improve the volumetric energy and power density of batteries.

Herein, we report on the preparation of Bi₂O₃ by utilization of Bi₂O₃ nanoparticles uniformly encapsulated in carbon nanofibers (denoted as Bi₂O₃/C) using a feasible electrospinning method. The size of the Bi₂O₃ nanoparticles is ≈10 nm and the content is optimized to 48 wt%. This customized Bi₂O₃/C is directly used as a flexible anode for SIBs, exhibiting high reversible capacity (430 mA h g⁻¹ at a current density of 100 mA g⁻¹ after 200 cycles) and high rate capability (230 mA h g⁻¹ even at 3.2 A g⁻¹). It has a stable capacity of 252 mA h g⁻¹ after 50 cycles at 400 mA g⁻¹ for a Na-ion full cell, showing great potential for practical applications. The improvement in electrochemical performance is due to the following merits. Firstly, the 1-D carbon matrix can prevent aggregation of Bi₂O₃ grains during prolonged sodiation/desodiation processes. Secondly, the CNFs can effectively enhance the electronic conductivity. More importantly, the high loading mass of Bi₂O₃ and ultra-small Bi₂O₃ nanoparticles can significantly improve the utilization of the electrode materials and reduce the irreversible capacity losses.

Experimental

Materials synthesis

In the typical synthesis, polyacrylonitrile (PAN, average M_w = 150 000, Aldrich) was used as the carbon source, N,N-dimethylformamide (DMF, anhydrous, 99.8%, Aldrich) as the solvent, and BiCl₃/Bi(NO₃)₃ (99%, Alfa Aesar) as the source of Bi₂O₃. To prepare the precursor solution, 1 g PAN and 0.75 g BiCl₃/Bi(NO₃)₃ were dissolved in 8 mL of DMF for 6 h under magnetic stirring. The mixture was transferred into a 5 mL plastic syringe equipped with a 21-gauge blunt tip needle and electrospun on a spinning system at 21 kV. Aluminium foil as the collector was located 15 cm away from the needle to collect the film. The as-collected nanofibers and comparison composite were all first heat-treated in a tube furnace at 5 °C min⁻¹ up to 280 °C in air for 2 h to stabilize the fiber structure by reaction of PAN. Subsequently, the Bi/C nanofibers were obtained by carbothermal reaction at 700 °C for 6 h in Ar (90 vol%)/H₂ (10 vol%). Finally, the films were treated at 300 °C for 3 h in air to get Bi₂O₃/C (BiCl₃ for embedded Bi₂O₃/C and Bi(NO₃)₃ for bare Bi₂O₃/C, respectively). In addition, Bi₂O₃ macroparticles were also synthesized using the same procedures except that the oxidation temperature was increased from 300 to 500 °C for 3 h in order to eliminate the carbon.

Material characterization

Morphological characterization of the Bi₂O₃/C nanofibers before and after sodium insertion were performed by scanning electron microscopy (SEM, NOVA 450, FEI) and by transmission electron microscopy (TEM, G2 FEI). The crystalline structures of the as-prepared nanofibers were characterized by X-ray diffraction (XRD, Shimadzu XRD-6000). The Raman spectra of the products were collected on a Thermo scientific FT-Raman spectrometer (FRA 106/s) using an Nd-line laser source with an excitation wavelength of 532 nm. The composition analysis of Bi₂O₃/C was performed with TG measurements (Diamond TGA/DSC 6300) with a heating rate of 10 °C min⁻¹ in air and N₂. The analysis of the valence states of the Bi₂O₃/C was performed with an X-ray electron spectrometer (XPS, AXIS-ULTRA DLD-600W).

Fabrication of hierarchical CuBi₂O₄ microsphere electrodes

The hierarchical CuBi₂O₄ microsphere anodes were prepared by mixing 70 wt% CuBi₂O₄ microspheres, 20 wt% super P, and 10 wt% vinylidene fluoride (PVDF) to form a slurry, which was then coated onto copper foil and dried at 90 °C overnight under vacuum. The area of the electrode is about 1.53 cm² while the loading of the whole material is about 2.5 mg cm⁻².

Fabrication of Li-ion batteries

Li-ion batteries were fabricated using lithium as the anode and 1 mol L⁻¹ LiPF₆ in a mixture of ethylene carbonate/diethyl carbonate (EC/DEC, 1 : 1 by volume) as the electrolyte. All the cells were assembled in a glove box with a water/oxygen content lower than 0.1 ppm and tested at room temperature.

Electrochemical measurements

The Bi₂O₃/C films were cut into circular samples (d = 12 mm) to be directly used as working electrodes without adding any polymeric binder and conductive additives. The average weight of the active material was about 0.53 mg cm⁻² and the thickness of the working electrodes is about 0.30 mm. The cells were assembled with sodium as the counter electrode, 1 M NaPF₆ in a mixture of ethylene carbonate/dimethyl carbonate (EC/DMC, 1 : 1 by volume) with 5% fluoroethylene carbonate (FEC) as the electrolyte, and Celgard®3501 (Celgard, LLC Corp., USA) as the separator. All the cells were assembled in a glove box with a water/oxygen content lower than 0.1 ppm and tested at room temperature. The galvanostatic charge–discharge test was conducted on a LAND cycler (CT 2001, Wuhan Kingnuo Electronic Co., China). Cyclic voltammetric and electrochemical impedance spectroscopy measurements were carried out with coin cells using a CHI 760D electrochemical workstation (ChenHua Instruments Co., China).

Results and discussion

The preparation details are shown in Scheme 1. In this process, the bismuth ions facilitate the formation of carbon and the carbon frame confines the growth of Bi₂O₃ grains. Thus Bi₂O₃ nanoparticles uniformly embedded in the carbon matrix are successfully synthesized. Correspondingly, the morphology of the composites and the content of Bi₂O₃ particles can be controlled by changing the precursor solution composition and reaction conditions.

Scheme 1 Schematic illustration of the formation process for Bi₂O₃/C nanofibers.

Fig. 1a shows the X-ray diffraction (XRD) patterns of the as-prepared samples. All diffraction peaks of the Bi/C sample correspond to an R3̄m space group of a rhombohedral crystal system Bi (JCPDS File No. 85-1331). After calcination in air, the entire diffraction peaks can be well indexed to the tetragonal Bi₂O₃ (JCPDS File No. 74-1374). No peaks of impurities are observed, demonstrating that Bi₂O₃/C with high purity has been successfully synthesized using the facile electrospinning method. The Raman spectra are applied to certify the existence of carbon and Bi₂O₃. As shown in Fig. 1b, two characteristic peaks located at 1344 and 1589 cm⁻¹ can be attributed to the typical disorder-induced feature (D bands) and the E_2g patterns of graphite carbon (G bands), respectively.^35,36 Moreover, the peaks around 199, 302, and 432 cm⁻¹ in a low band are Bi₂O₃ characteristic signals.³⁷ Meanwhile, XRD patterns and Raman spectra of the pure carbon nanofibers (CNFs) are displayed (Fig. S1, ESI†). The X-ray photoelectron spectroscopy (XPS) spectra further clarify the structural change from simple substance bismuth to bismuth trioxide (Fig. 1c). The binding energy of 4f_7/2 from 157 eV shifts to 159 eV and 4f_5/2 from 162 eV shifts to 164 eV of bismuth elements, respectively, due to the elimination of twelve electrons (4Bi + 3O₂ − 12e⁻ → 2Bi₂O₃). The process of oxidation leads to the formation of a more stable electron configuration (4f¹⁴5d¹⁰6s²), which requires a higher electronic excitation energy. The XPS spectra of carbon and oxygen are shown in Fig. S2 (ESI†). All results effectively indicate that the carbonization and oxidation processes are complete and the final product is verified.

Fig. 1 (a) XRD patterns of the as-prepared samples. (b) Raman spectra of the as-prepared samples. (c) XPS spectra of bismuth (4f).

Fig. 2a and b reveal the morphologies of electrospun Bi/C and Bi₂O₃/C as investigated by field-emission scanning electron microscopy (FESEM). The fibers of the two samples are smooth and continuous with lengths of up to hundreds of micrometers. The average diameters of Bi/C and Bi₂O₃/C are ≈120 and 160 nm, respectively, and the fibers well interlink into a conductive 3-D network. The interlinked structure of Bi₂O₃/C nanofibers could effectively form binder free and flexible films, which can facilitate the contact between active materials and the electrolyte, leading to rapid charge and discharge processes. Simultaneously, the Bi₂O₃ nanoparticles stick on the carbon nanofibers, denoted as Bi₂O₃/C (bare) (Fig. S3, ESI†). The Bi₂O₃ nanoparticles are clearly visible and unordered sticking on the CNFs. Fig. 2d shows different resolution transmission electron microscopy (TEM) images of the Bi₂O₃/C sample. As expected, the Bi₂O₃ nanoparticles are uniformly embedded in CNFs without aggregation and the diagrams of particle size distribution in Fig. S4 (ESI†) manifest that the Bi₂O₃ grains in Bi₂O₃/C are around 10 nm. It is noteworthy that these ultra-small Bi₂O₃ nanoparticles provide much more surface area than bulk Bi₂O₃ and increase the electrochemical activity for SIBs.

Fig. 2 SEM images of (a) Bi/C and (b) Bi₂O₃/C. (c) SEM elemental mapping of Bi₂O₃/C. (d) TEM image. (e and f) The SAED pattern and lattice plane spacing. (g) TEM (HAADF C, O, Bi) elemental mapping images of Bi₂O₃/C.

The selected area electron diffraction (SAED) pattern (Fig. 2e) is well indexed to the pure Bi₂O₃ phase, corresponding to the diffraction peaks of the (201), (221) and (400) planes shown in the XRD pattern. Meanwhile, the HRTEM image (Fig. 2f) shows a group of parallel fringes with a d-spacing of 0.16 nm clearly, corresponding to the crystal (221) plane of rhombohedral Bi₂O₃. These results are consistent with the XRD patterns (Fig. 1a). Fig. 2c and g show the SEM and TEM elemental mapping images of Bi₂O₃/C, revealing that the Bi₂O₃ nanoparticles are homogeneously located in the carbon matrix. The Bi₂O₃ grain content in the CNFs is calculated from the TGA–DSC curves (Fig. S5, ESI†), based on the weight loss of carbon combustion and the mass ratio is ≈48 wt%. It is noteworthy that the content of Bi₂O₃ nanoparticles is determined by the BiCl₃ concentration in the precursor solution (BiCl₃, PAN and DMF). A series of content gradients of Bi₂O₃ in the composites are shown in Fig. S6 (ESI†). The lower the concentration of BiCl₃, the lower the particle content in the carbon nanofiber. The Bi₂O₃ nanoparticles could be seen clearly with uniform distribution at the low concentration (Fig. S6a, ESI†) and their content is only 20 wt% in the carbon nanofibers (Table S1, ESI†). However, when the Bi₂O₃ nanoparticle content is more than 48 wt%, these nanoparticles aggregate together and even overflow from the carbon fibers. The surrounding carbon matrix exerts a continuous structure which plays a key role as a strong buffer region to accommodate the mechanical stress during ion insertion/deinsertion, especially for ions with a larger ionic radius such as sodium.

The binder free and flexible films (Bi₂O₃/C) were directly used as anodes for SIBs and the conductivity of the Bi₂O₃/C is revealed by the electrochemical impedance spectrum (EIS) and cyclic voltammograms (CVs). To demonstrate the hypothesis that better electrochemical performance could be achieved by the Bi₂O₃ nanoparticles embedded in CNFs, Bi₂O₃/C (Bare) and Bi₂O₃ (Macroparticle) (Fig. S7, ESI†) were used as references. The electrode made from Bi₂O₃/C shows higher electronic conductivity and lower charge-transfer resistance compared with those of Bi₂O₃/C (Bare) and Bi₂O₃ (Macroparticle) as revealed by the Nyquist plots (Fig. 3a). The simplified-contact-Randles equivalent circuit for the simulation of the Nyquist plots and more simulation information are shown in Depiction S1 (ESI†). It is noteworthy that the Na-ion diffusion coefficient (D) directly influences the cycling performance of the electrode material. The calculated formulas of the D value are presented in Depiction S2 (ESI†) and are directly determined by the Warburg factor (σ).³⁸ The D value has negative correlation with the σ value and the values of σ are clearly shown in Fig. 3b. The σ values at open circuit are calculated to be 17.63, 50.87 and 578.65 for the Bi₂O₃/C, Bi₂O₃/C (Bare) and Bi₂O₃ (Macroparticle) electrodes, respectively, which are in accordance with the slopes of the real part of the complex impedance versus the square root of frequency, ω^−1/2. For further emphasis, the CVs of the as-prepared Bi₂O₃/C sample at different scan rates from 0.1 to 1 mV s⁻¹ are shown in Fig. 3c. When the scan rate is 0.1 mV s⁻¹, two obvious reduction peaks at 0.33 and 0.57 V and an oxidation peak at 0.74 V are clearly observed, which recur intensively in the subsequent scans. With increase of the scan rate, the peak current increases without being proportional to the square root of the scan rate, which illustrates that the charge and discharge processes are composed of non-Faradaic and Faradaic behaviour.^39,40 The relationship between the peak current (i) and scan rate (v) could be illustrated by the equations as follows:

where a and b are adjustable parameters. The b value determines the style of Na-ion insertion/extraction. When the value of b is 0.5, the electrochemical reaction is dominated by ionic diffusion (non-Faradaic). When the value of b is 1, the process is mainly decided by pseudo-capacitive (Faradaic) control. Fig. 3d depicts the linear relationship between log(i) and log(v) plots at every peak potential. The b values of the four peaks are 0.82, 0.81, and 0.88, which means that the electrochemical reactions of the Bi₂O₃/C electrode are dominantly controlled by pseudo-capacitive behaviour, leading to fast Na-ion insertion/extraction (high rate capability).

Fig. 3 Electrochemical characterizations of the as-prepared samples. (a) Nyquist plots of the impedance spectra. (b) The real part of the complex impedance versus ω^−1/2 at open circuit voltage. (c) Cyclic voltammograms of Bi₂O₃/C at varied sweep rates from 0.1 to 1 mV s⁻¹. (d) Relationship between the peak current density and scan rate in the logarithmic algorithm.

Due to the enhanced electronic conductivity and Na-ion diffusion coefficient, the cycling and galvanostatic curves of the Bi₂O₃/C electrode are further demonstrated (Fig. 4a and b). The reversible capacity of the Bi₂O₃/C composite is perfectly maintained compared with the other two samples and the electrode was galvanostatic charged and discharged at 100 mA g⁻¹ for 200 cycles. After a few stabilization cycles, the reversible capacity can reach to 430 mA h g⁻¹, and a high capacity retention of 63% can still be achieved over 200 cycles. For comparison, the Bi₂O₃/C (Bare) electrode exhibits a reversible capacity of 205 mA h g⁻¹ after 50 cycles, while Bi₂O₃ (Macroparticle) displays a fast capacity fading (30 mA h g⁻¹ after 20 cycles). The higher capacity retention of Bi₂O₃/C indicates that the uniform dispersion of Bi₂O₃ nanoparticles in the carbon matrix can generate a balanced stress upon prolonged cycling and maintain the integrity of the electrode. As shown in Fig. 4b, the voltage profiles showed sloping lines during both charge and discharge, which were consistent with the broad current peaks observed during CV scans (Fig. 3c). The sodiation capacity for the first cycle was 979 mA h g⁻¹ with a Coulombic efficiency of around 51.1%. The large initial irreversible capacity is mainly due to the formation of an SEI film and electrochemical decomposition of the solvent on the electrode. It is noteworthy that some pyrolysis carbon (52%) is present in the CNFs, which would decrease the initial Coulombic efficiency and capacity of the Bi₂O₃/C electrode.⁴¹ Since CNFs can only provide a sodiation capacity of 70 mA h g⁻¹ at the same current density after 200 cycles (Fig. S8b, ESI†), the capacity contributed by the Bi₂O₃ nanoparticles was around 430 mA h g⁻¹ after 200 cycles, which provided a high capacity retention (69%). The Coulombic efficiency of the second cycle exceeded 98% and remained stable for the rest of the cycles. In addition, the typical charge and discharge voltage profiles of Bi₂O₃/C (Bare) and Bi₂O₃ (Macroparticle) are shown in Fig. S9 (ESI†).

Fig. 4 (a) Specific capacity retention at a current density of 100 mA g⁻¹ and the corresponding Columbic efficiency. (b) Charge–discharge curves of the Bi₂O₃/C electrode. (c) Rate capability at current densities ranging from 100–3200 mA g⁻¹. (d) Comparison of the cycling capability of Bi₂O₃/C with other Bi and Bi₂O₃ electrodes reported recently.

To get further evidence of the improved electrochemical performance of Bi₂O₃/C for SIBs, the rate capabilities of both Bi₂O₃/C (Bare) and Bi₂O₃ (Macroparticle) electrodes were investigated (Fig. 4c). The Bi₂O₃/C electrode delivers a reversible capacity as high as 657, 546, 466, 379, 334, and 265 mA h g⁻¹ when cycled at 60, 100, 200, 400, 800 and 1600 mA g⁻¹, respectively. Even at a high current density of 3200 mA g⁻¹, it still remains nearly at 230 mA h g⁻¹. When the current density was set back to 100 mA g⁻¹, the reversible capacity can be recovered up to 352 mA h g⁻¹, showing excellent rate reversibility. The good rate performance can be attributed to the following facts: firstly, the 1-D carbon nanofibers cross-linked together to form a 3-D network, which could effectively facilitate the ionic/electron diffusion kinetics. Furthermore, the ultra-small Bi₂O₃ nanoparticles are readily accessible to the electrolyte and can easily be fully utilized for sodium-storage. It is noteworthy that the binder-free electrodes can improve the electronic conductivity and energy density. Lastly, the electrochemical reactions of the Bi₂O₃/C electrode are mainly controlled by the pseudo-capacitive behaviour, leading to a fast Na-ion insertion/extraction, and efficiently rendering the high rate capability. By contrast, the rate capabilities of Bi₂O₃/C (Bare) and Bi₂O₃/C (Macroparticle) are presented and both of them show very poor capacity even at the current density of 100 mA g⁻¹ (170 and 80 mA h g⁻¹, respectively). As can be concluded, the smallest particle size and high conductivity of the electrode material can significantly facilitate the electron transfer and Na-ion diffusion kinetics, which endow Bi₂O₃/C with excellent rate capability.

Fig. 4d illustrates the superiority of bismuth metal oxidation to prepare Bi₂O₃ nanoparticles over other Bi₂O₃ electrodes reported in the recent literature. Bismuth nanoparticles wrapped by graphene have been synthesized and deliver a capacity of 240 mA h g⁻¹ at 50 mA h g⁻¹ after 80 cycles (ref. 9); bismuth nanoparticles embedded in carbon spheres were prepared using an aerosol spray pyrolysis technique and deliver a capacity of 100 mA h g⁻¹ at 123.5 mA g⁻¹ after 100 cycles (ref. 17). Both of them showed poor cycling performance and provided low capacity due to the inherent specific capacity (384 mA h g⁻¹). Furthermore, the capacities of the Bi₂O₃ anodes are 410 and 413 mA h g⁻¹ for ref. 15 and 16, respectively, which are far above the highest capacity (240 mA h g⁻¹) of the bismuth metal. It is noteworthy that the Bi₂O₃/C electrode delivers a high reversible capacity of 430 mA h g⁻¹, which is more superior over other Bi₂O₃ electrodes. These results can emphasize that the significance of Bi₂O₃ by bismuth metal oxidation and Bi₂O₃ ultra-small nanoparticles embedded in carbon nanofibers can effectively improve the cycling performance for SIBs.

We have measured the electrochemical performance of full cells which are assembled with a coupling NASICON-type Na₃V₂(PO₄)₃ cathode and Bi₂O₃/C anode. The NASICON-type Na₃V₂(PO₄)₃ has been considered as a promising cathode material (theoretical capacity of 117.6 mA h g⁻¹) for SIBs owing to its fast Na-ion diffusion, wide voltage plateau, and high thermal stability. So, it is used as a cathode material to couple with the Bi₂O₃/C anode.^42,43Fig. 5a illustrates the composition of the Bi₂O₃/C–Na₃V₂(PO₄)₃ full cell. The Bi₂O₃/C anode has been demonstrated to achieve a six-electron reaction, and the Na₃V₂(PO₄)₃ cathode possesses a theoretical 2Na extraction capability. The reaction of the full cell is shown in eqn (3):

Fig. 5b illustrates the actual-effect depiction of the Bi₂O₃/C–Na₃V₂(PO₄)₃ full cell. The interlinked CNFs ensure that the full cell possesses excellent flexibility and the charged full cell can light a high colour temperature red-LED (Fig. 5c) after repeated inflexion. Fig. 5d exhibits the cycling performance of the full cell with excessive Na₃V₂(PO₄)₃ between 0.01 and 3.0 V. The full cell delivers a discharge capacity of 252 mA h g⁻¹ based on the mass of Bi₂O₃/C at a current density of 400 mA g⁻¹ after 50 cycles, demonstrating its promising potential for portable equipment and wearable products.

Fig. 5 (a and b) Schematic illustration of the SIBs with the Na₃V₂(PO₄)₃vs. Bi₂O₃/C couple. (c) The practical application with an R-LED. (d) Cycling performance of the Na₃V₂(PO₄)₃vs. Bi₂O₃/C full cell with excessive Na₃V₂(PO₄)₃ at a current density of 400 mA g⁻¹.

Conclusions

In summary, a flexible binder-free film composed of bismuth oxide nanoparticles embedded in carbon nanofibers (Bi₂O₃/C) was prepared by a feasible electrospinning method and used as a sodium ion battery (SIB) anode. The CNFs form a conductive 3-D network and thus a flexible binder-free film is constructed. When used as an anode material for SIBs, the Bi₂O₃/C electrode exhibits an excellent cycling stability (430 mA h g⁻¹ after 200 cycles), rate capability (230 mA h g⁻¹ at a current density of 3.2 A g⁻¹) and sustained capacity in a full cell (252 mA h g⁻¹ after 50 cycles at 400 mA g⁻¹) due to its ultra-small Bi₂O₃ nanoparticles embedded in CNFs, no auxiliary additives, enhanced conductivity and high Na-ion diffusion coefficient. The superior electrochemical performance further demonstrates that Bi₂O₃ nanoparticles have great potential for practical applications, especially in flexible energy storage devices.

Acknowledgements

This work was supported by the National Basic Research Program of China (Grant No. 2015CB755602 and 2013CB922104), NSFC (51673077 and 21474034). We also thank the Analytical and Testing Center of Huazhong University of Science and Technology and the Center of Micro-Fabrication and Characterization (CMFC) of WNLO for the use of their facilities.

Notes and references

1. X. Y. Lai, J. E. Halpert and D. Wang, Energy Environ. Sci., 2012, 5, 5604–5618.
2. J. Wang, H. Tang, H. Ren, R. Yu, J. Qi, D. Mao, H. Zhao and D. Wang, Adv. Sci., 2014, 1, 1400011.
3. J. Wang, N. Yang, H. Tang, Z. Dong, Q. Jin, M. Yang, D. Kisailus, H. Zhao, Z. Tang and D. Wang, Angew. Chem., Int. Ed., 2013, 52, 6417–6420.
4. H. Ren, J. Sun, R. Yu, M. Yang, L. Gu, P. Liu, H. Zhao, D. Kisailus and D. Wang, Chem. Sci., 2016, 7, 793–798.
5. J. Qi, X. Lai, J. Wang, H. Tang, H. Ren, Y. Yang, Q. Jin, L. Zhang, R. Yu, G. Ma, Z. Su, H. Zhao and D. Wang, Chem. Soc. Rev., 2015, 44, 6749–6773.
6. H. Tang, C. M. Hessel, J. Wang, N. Yang, R. Yu, H. Zhao and D. Wang, Chem. Soc. Rev., 2014, 43, 4281–4299.
7. N. Mahmood, T. Y. Tang and Y. L. Hou, Adv. Energy Mater., 2016, 6, 1600374.
8. Y. Xu, Y. Zhu, Y. Liu and C. Wang, Adv. Energy Mater., 2013, 3, 128–133.
9. D. Su, S. Dou and G. Wang, Nano Energy, 2015, 12, 88–95.
10. Y. Zhu, X. Han, Y. Xu, Y. Liu, S. Zheng, K. Xu, L. Hu and C. Wang, ACS Nano, 2013, 7, 6378–6386.
11. Y. Li, M. A. Trujillo, E. Fu, B. Patterson, L. Fei, Y. Xu, S. Deng, S. Smirnov and H. Luo, J. Mater. Chem. A, 2013, 1, 12123.
12. X. Cui, X. Yu, L. Hou, A. Gagnoud, Y. Fautrelle, R. Moreau, Z. Ren, X. Lu and X. Li, J. Alloys Compd., 2016, 684, 21–28.
13. H. Che, J. Chen, K. Huang, W. Hu, H. Hu, X. Liu, G. Che, C. Liu and W. Shi, J. Alloys Compd., 2016, 688, 882–890.
14. Q. Wang, X. Shi, E. Liu, J. C. Crittenden, X. Ma, Y. Zhang and Y. Cong, J. Hazard. Mater., 2016, 317, 8–16.
15. C. Nithya, ChemPlusChem, 2015, 80, 1000–1006.
16. M. K. Kim, S. H. Yu, A. Jin, J. Kim, I. H. Ko, K. S. Lee, J. Mun and Y. E. Sung, Chem. Commun., 2016, 52, 11775–11778.
17. F. Yang, F. Yu, Z. Zhang, K. Zhang, Y. Lai and J. Li, Chem. – Eur. J., 2016, 22, 2333–2338.
18. T. N. Ramesh and P. V. Kamath, Electrochim. Acta, 2008, 53, 4721–4726.
19. Y. Li, H. Zhang, Y. Wang, P. Liu, H. Yang, X. Yao, D. Wang, Z. Tang and H. Zhao, Energy Environ. Sci., 2014, 7, 3720–3726.
20. M. Chen, J. Wang, H. Tang, Y. Yang, B. Wang, H. Zhao and D. Wang, Inorg. Chem. Front., 2016, 3, 1065–1070.
21. F. Wang, J. Wang, H. Ren, H. Tang, R. Yu and D. Wang, Inorg. Chem. Front., 2016, 3, 365–369.
22. W. Fang, N. Zhang, L. Fan and K. Sun, J. Power Sources, 2016, 333, 30–36.
23. J. Y. Wang, H. J. Tang, H. Wang, R. B. Yu and D. Wang, Mater. Chem. Front., 2017, 1, 414–430.
24. J. Zhang, H. Ren, J. Wang, J. Qi, R. Yu, D. Wang and Y. Liu, J. Mater. Chem. A, 2016, 4, 17673–17677.
25. H. Li, H. Ma, M. Yang, B. Wang, H. Shao, L. Wang, R. Yu and D. Wang, Mater. Res. Bull., 2017, 87, 224–229.
26. X. X. Yu, H. Yin, H. X. Li, W. Zhang, H. Zhao, C. Li and M. Q. Zhu, Nano Energy, 2017, 34, 155–163.
27. H. Ren, R. Yu, J. Wang, Q. Jin, M. Yang, D. Mao, D. Kisailus, H. Zhao and D. Wang, Nano Lett., 2014, 14, 6679–6684.
28. H. Yin, Q. Li, M. Cao, W. Zhang, H. Zhao, C. Li, K. Huo and M. Zhu, Nano Res., 2017 DOI:10.1007/s12274-016-1408-z.
29. L. Wang, B. Gao, C. Peng, X. Pensg, J. Fu, P. K. Chu and K. Huo, Nanoscale, 2015, 7, 13840–13847.
30. J. Wang, H. Tang, L. Zhang, H. Ren, R. Yu, Q. Jin, J. Qi, D. Mao, M. Yang, Y. Wang, P. Liu, Y. Zhang, Y. Wen, L. Gu, G. Ma, Z. Su, Z. Tang, H. Zhao and D. Wang, Nat. Energy, 2016, 1, 16050.
31. L. Zeng, W. Zeng, Y. Jiang, X. Wei, W. Li, C. Yang, Y. Zhu and Y. Yu, Adv. Energy Mater., 2015, 5, 1401377.
32. X. Pu, L. Li, H. Song, C. Du, Z. Zhao, C. Jiang, G. Cao, W. Hu and Z. L. Wang, Adv. Mater., 2015, 27, 2472–2478.
33. E. C. Self, E. C. McRen, R. Wycisk and P. N. Pintauro, Electrochim. Acta, 2016, 214, 139–146.
34. C. Kim, K. S. Yang, M. Kojima, K. Yoshida, Y. J. Kim, Y. A. Kim and M. Endo, Adv. Funct. Mater., 2006, 16, 2393–2397.
35. Y. Kang, Y. Yang, L. C. Yin, X. Kang, G. Liu and H. M. Cheng, Adv. Mater., 2015, 27, 4572–4577.
36. M. J. Xie, J. Yang, J. Y. Liang, X. F. Guo and W. P. Ding, Carbon, 2014, 77, 215–225.
37. M. M. Martins, D. S. Silva, L. R. P. Kassab, S. J. L. Ribeiro and C. B. d. Araújo, J. Braz. Chem. Soc., 2015, 26, 2520–2524.
38. Y. Zhao, L. Peng, B. Liu and G. Yu, Nano Lett., 2014, 14, 2849–2853.
39. H. Yin, M. L. Cao, X. X. Yu, C. Li, Y. Shen and M. Q. Zhu, RSC Adv., 2017, 7, 13250–13256.
40. X. Xiang, K. Zhang and J. Chen, Adv. Mater., 2015, 27, 5343–5364.
41. H. Yin, X. X. Yu, Q. W. Li, M. L. Cao, W. Zhang, H. Zhao and M. Q. Zhu, J. Alloys Compd., 2017, 706, 97–102.
42. F. Shen, W. Luo, J. Dai, Y. Yao, M. Zhu, E. Hitz, Y. Tang, Y. Chen, V. L. Sprenkle, X. Li and L. Hu, Adv. Energy Mater., 2016, 6, 1600377.
43. Y. Xu, M. Zhou and Y. Lei, Adv. Energy Mater., 2016, 6, 1502514.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7qm00128b

---