Carbon nanofiber-supported B₂O₃–SnO_x glasses as anode materials for high-performance lithium-ion batteries

Abstract

A series of carbon nanofiber-supported B₂O₃–SnO_x glasses (0B₂O₃–SnO_x/CNFs, 0.5B₂O₃–SnO_x/CNFs, 1.5B₂O₃–SnO_x/CNFs and 3B₂O₃–SnO_x/CNFs) have been prepared by an electrospinning technique, and a subsequent stabilization and carbonization treatment. The amorphous B₂O₃–SnO_x glass nanoparticles homogeneously dispersed in the carbon nanofibers (CNFs) matrix have been utilized as anode materials for lithium-ion batteries. The typical 1.5B₂O₃–SnO_x/CNF nano hybrid anode material demonstrates the highest reversible capacity as well as superior cycling stability and rate performance by virtue of the special glass network structure of non-bridging oxygen (NBO: B–O⋯Sn) in the B₂O₃–SnO_x glasses, which can not only buffer the large volume change of SnO_x, but also prohibit the aggregation of Sn nanoparticles during the charge–discharge cycles.

---

1. Introduction

Lithium-ion batteries (LIBs) have been considered as one of the most important power sources for microelectronic devices, electric vehicles (EVs) and hybrid electric vehicles (HEVs) owing to their outstanding properties of high energy density, long cycle life, low cost and environmental benignity.^1–4 Currently, as the commercial anode material for LIBs, graphite has a relatively low theoretical capacity of 372 mA h g⁻¹, which cannot meet the requirements for high-performance LIBs.^5–10 Tin and tin oxides have attracted considerable attentions as potential substitutes for graphite owing to their high theoretical reversible capacities (Sn: 990 mA h g⁻¹, SnO: 880 mA h g⁻¹ and SnO₂: 780 mA h g⁻¹).^6–13 Although Sn-based materials have very high capacities, their practical use have been seriously hampered by the large volume change (up to 300%) associated with reversible lithium alloying and de-alloying process, which often result in mechanical degradation and poor cycling performance.^13–16 Therefore, in order to solve the problem, several strategies have been developed.

A common strategy is dispersing nano-sized Sn-based particles into a carbon matrix for releasing the internal strain and shortening the diffusion length of lithium ions to improve cycling performance.^13,17–19 Various forms of carbon supporters for Sn-based nanoparticles have been reported, such as carbon nanotubes (CNTs),²⁰ graphene,^21,22 and carbon nanofibers (CNFs).^13,23 Currently, a simple and versatile technique of electrospinning has been used to prepare SnO_x-containing CNFs nanohybrids (SnO_x/CNFs), which can be used as self-standing and binder-free anodes for LIBs. In this case, the SnO_x particles are uniformly distributed in the CNFs matrix, which can effectively buffer the volume change and hinder the aggregation of Sn particles.^13,18,24,25 Nevertheless, owing to the insufficient electronic conductivity of the SnO_x/CNFs electrodes due to the amorphous CNFs matrix, it is highly desired to further improve the reversible capacity and rate capability of the SnO_x/CNFs anode materials.

Recent researches on the modification of Sn-based oxides by introducing various heteroatoms such as Sb,²⁶ Cu,²⁷ Ni,²⁸ Ti,²⁹ and P³⁰ suggest that the addition of either metallic or nonmetallic elements should be an effective route to enhance the reversible capacity and rate capability of Sn-based oxides electrodes. Recently, it is reported that B₂O₃ play a promotional role in enhancing the cycling and rating performance of SnO₂ anodes, because boron atoms can act as an electron-acceptor during charge/discharge process, thus making more lithium species are reversibly inserted into the host.^31,32 Besides, the non-bridging oxygen network (NBO, B–O⋯Sn) existed in the SnO_x–B₂O₃ glasses can provide additional free spaces for compensating the volume changes of SnO_x.³³ Therefore, it should be attractable and significant to develop a novel carbon nanofibers-supported B₂O₃–SnO_x glasses (B₂O₃–SnO_x/CNFs) nanohybrids, and to investigate the role of B₂O₃ in enhancing the SnO_x/CNFs anode materials for LIBs.

In this work, we successfully synthesized several B₂O₃–SnO_x/CNFs nanohybrids with different amounts of B₂O₃ through electrospinning the precursor solution of polyacrylonitrile (PAN) and tin tetrachloride with addition of varied amounts of boric acid followed by thermal treatments. The addition of boron into SnO_x/CNFs forms a B₂O₃–SnO_x glasses network in CNFs matrix, which can effectively hinder the aggregation of Sn nanoparticles during cycling, and thus producing a typical B₂O₃–SnO_x/CNFs electrode with greatly enhanced reversible capacity and rate capability compared with SnO_x/CNFs electrode.

2. Experimental

2.1. Preparation of B₂O₃–SnO_x/CNFs

2.0 g polyacrylonitrile (PAN, M_w = 100 000 g mol⁻¹, UK Courtaulds Co.) was dissolved in 15 ml N,N-dimethylformamide (DMF, Beijing Chemical Co.) at 60 °C for 8 h (solution A). Tin tetrachloride (SnCl₄, Beijing Chemical Co.) and ethylene glycol (EG, Beijing Chemical Co.) were mixed with a volume ratio of 1 : 3 under magnetic stirring to acquire a homogenous solution (solution B). Then, 1 ml solution B together with varied amounts of boron acid (H₃BO₃, Beijing Chemical Co.) (0, 0.01, 0.03, and 0.06 g) were added into solution A under magnetic stirring at room temperature for 24 h to form homogenous electrospinning solutions, respectively. All the precursor solutions were electrospun at a voltage of 14 kV using a high voltage supply (DWP303-1AC, China), with a flow rate of 0.3 ml h⁻¹ and a collection distance of 14 cm. The as-prepared electrospun nanofiber webs were then stabilized at 270 °C for 3 h under air atmosphere, and subsequently carbonized in N₂ atmosphere at 700 °C for 1 h to prepare the B₂O₃–SnO_x/CNFs samples, which were named as 0B₂O₃–SnO_x/CNFs, 0.5B₂O₃–SnO_x/CNFs, 1.5B₂O₃–SnO_x/CNFs and 3B₂O₃–SnO_x/CNFs, respectively, where the numbers (0, 0.5, 1.5, and 3) represent the approximate weight percent of boric acid to PAN in the precursor solutions.

2.2. Characterizations

The morphologies of samples were observed by using a field emission scanning electron microscope (FE-SEM, Supra55, Carl Zeiss). Transmission electron microscope (TEM, Tecnai G2 20, FEI Company) and a high resolution transmission electron microscope (HR-TEM, JEM-3010, JEOL) were used to detect the structure of single nanofiber. Thermogravimetric analysis (TGA) was carried out on a TA-Q50 with a heating rate of 10 °C min⁻¹ from 25 to 900 °C in air. X-ray diffraction (XRD) was carried out on a Brucker diffractometer (WAXD, D8 Advance, Brucker, Cu Kα, λ = 0.154 nm) from 5 to 90°. Raman spectroscopy was recorded on a Renishaw with 514 nm diode laser excitation. X-ray photoelectron spectroscopy (XPS) was performed on an ESCALAB 250 with Al K R radiation (1486.6 eV).

2.3. Electrochemical measurements

Electrochemical performance of the samples was evaluated on CR2025 coin cells under ambient temperature following our previous study.²⁴ Four samples were cut into small discs with typical diameter of 12 mm, inserted into two nickel meshes with typical diameter of 14 mm, and pressed at a pressure of 20 MPa to be used directly as a working electrode. The cells were assembled in an Ar-filled glove box (OMNI-LAB) using lithium as a counter and reference electrode, Celgard 2300 membrane as a separator, and 1 M LiPF₆ in ethylene carbonate (EC)/dimethyl carbonate (DMC) (1 : 1 v/v) as an electrolyte. Galvanostatic charge/discharge cycles were evaluated on a LAND CT 2001A battery tester in the potential range of 0.005–3.0 V at a constant current density of 200 mA g⁻¹ or varied current densities from 100 to 2000 mA g⁻¹. The capacity is calculated based on B₂O₃–SnO_x/CNFs. Cyclic voltammetry (CV) were measured in the voltage range from 0.005 to 3 V using an Autolab PGSTAT 302 N (Metrohm) workstation at a scanning rate of 0.1 mV s⁻¹ or different scan rates of 0.1, 0.5, 1, 3, and 5 mV s⁻¹, respectively. Electrochemical impedance spectroscopic (EIS) studies were performed in a frequency range of 100 kHz to 0.01 Hz under the potentiostatic signal amplitude of 10 mV.

3. Results and discussion

Fig. 1a presents the field emission scanning electron microscopy (FE-SEM) image of the 0B₂O₃–SnO_x/CNFs, which exhibits a long continuous fibrous morphology with a diameter in the range of 200 nm to 300 nm. In Fig. 1b, the morphology of the 1.5B₂O₃–SnO_x/CNFs is similar to that of 0B₂O₃–SnO_x/CNFs. It was noted that the surface of both 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs nanofibers were smooth, and no nanoparticles were observed at the surface even under higher magnification, as shown in the inset images in Fig. 1a and b, suggesting that the B₂O₃–SnO_x nanoparticles may be dispersed inside the CNFs. The transmission electron microscopy (TEM) images of 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs (Fig. 1c and d) show the uniform dispersion of B₂O₃–SnO_x nanoparticles. Moreover, as observed in Fig. 1e and f, no distinct crystalline structures were observed, indicating that the B₂O₃–SnO_x nanoparticles are amorphous and homogeneously dispersed as ultrafine nanoparticles. Such homogenous dispersion of B₂O₃–SnO_x nanoparticles in CNFs is believed to be beneficial to the electrochemical performances of the B₂O₃–SnO_x/CNFs electrodes.

Fig. 1 The FE-SEM images of (a) 0B₂O₃–SnO_x/CNFs and (b) 1.5B₂O₃–SnO_x/CNFs, the insets are enlarged images; the TEM images of (c) 0B₂O₃–SnO_x/CNFs and (d) 1.5B₂O₃–SnO_x/CNFs; the high resolution TEM images of (e) 0B₂O₃–SnO_x/CNFs and (f) 1.5B₂O₃–SnO_x/CNFs.

Thermogravimetric analysis (TGA) was carried out to determine the amounts of B₂O₃–SnO_x nanoparticles inside the CNFs as shown in the Fig. 1S (seen in the ESI†). The weight loss in the first stage from room temperature to 200 °C is resulted from desorption of water vapor and evacuation of gaseous contents of the samples,³⁴ and the sharp weight loss in the range of 200–550 °C is ascribed to the removal of carbon component. The four samples show similar compositions, which are composed of about 20 wt% of the B₂O₃–SnO_x.

Fig. 2a shows the X-ray diffraction (XRD) patterns of the B₂O₃–SnO_x/CNFs samples. All the curves exhibit a broad peak of amorphous graphitic carbon (JCPDS 13-0148) near 25°, while no other diffraction peaks can be observed, indicating the amorphous nature of the B₂O₃–SnO_x particles. Moreover, no obvious difference can be seen from the Fig. 2a among the four B₂O₃–SnO_x/CNFs samples, suggesting that boron addition has little influence on the crystal structure of SnO_x/CNFs. Fig. 2b shows the Raman spectra of all B₂O₃–SnO_x/CNFs samples. All the Raman spectra display two prominent peaks of the G band (1580 cm⁻¹) and the D band (1350 cm⁻¹) which are corresponding to the sp²-conjugated carbon structure and the disordered carbon structures.^30,35 Generally, the intensity ratio of D band to G band (I_D/I_G) is used to estimate the disorder degree of carbon materials. The I_D/I_G values for all the B₂O₃–SnO_x/CNFs samples were calculated to be ca. 1.2, with an indicator of low graphitization degree of the CNFs. However, it is worth noting that the I_D/I_G value falls from 1.19 for 0B₂O₃–SnO_x/CNFs to 1.16 for 3B₂O₃–SnO_x/CNFs, which shows the increase in the graphitization of CNFs with boron addition. This change of I_D/I_G value with boron addition is helpful for the efficient electron conduction when used as the anode materials for LIBs.³⁶

Fig. 2 (a) XRD patterns and (b) Raman spectra of (1) 0B₂O₃–SnO_x/CNFs, (2) 0.5B₂O₃–SnO_x/CNFs, (3) 1.5B₂O₃–SnO_x/CNFs, and (4) 3B₂O₃–SnO_x/CNFs.

Fig. 3 shows the X-ray photoelectron spectroscopy (XPS) analysis of the 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs samples. As shown in Fig. 3a, the XPS general spectra present the signals of C, N, O and Sn elements for both 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs. In particular, B1s peak was detected at the binding energy around 190 eV on 1.5B–SnO_x/CNFs. Fig. 3b displays the detailed B1s XPS spectra of the 1.5B₂O₃–SnO_x/CNFs, which are evident for the existence of boron heteroatoms in the forms of B–C (binding energy of 190 eV) and B–O (binding energy of 192 eV) bonds.³⁷ The B–C bonds in the 1.5B₂O₃–SnO_x/CNFs indicates that boron enters the carbon lattice by substituting for carbon at the trigonal sites,³⁸ which can acts as electron acceptor as it has three valence electrons, causing the improvement in electrical conductivity of the CNFs matrix.³⁹ And the B–O bonds shows the boron maybe modify the structure of SnO_x. The Sn3d_5/2 spectrum in Fig. 3c was decomposed to three peaks at binding energy of 487.5, 486.7, and 484.7 eV, which are assigned to the Sn⁴⁺, Sn²⁺ and Sn⁰, respectively.⁴⁰ Moreover, the proportions of Sn⁰, Sn²⁺ and Sn⁴⁺ for 1.5B₂O₃–SnO_x/CNFs are similar to those for 0B₂O₃–SnO_x/CNFs, demonstrating that boron addition has no significant effect on the valence state of Sn. Fig. 3d shows the XPS O1s spectra of 0.5B₂O₃–SnO_x/CNFs, 1.5B₂O₃–SnO_x/CNFs and 3B₂O₃–SnO_x/CNFs. Four fitted peaks appeared in Fig. 4d: the two peaks at 530.7 and 532.3 eV are attributed to Sn–O and C–O, while the other two peaks at 532.9 and 531.5 eV are assigned to the bridging oxygen (BO: B–O–B) and the non-bridging oxygen (NBO: B–O⋯Sn).^29,33 By calculating, the ratios of NBO to BO were 0.57, 1.07 and 0.73 for 0.5B₂O₃–SnO_x/CNFs, 1.5B₂O₃–SnO_x/CNFs and 3B₂O₃–SnO_x/CNFs, respectively. It can be seen that the 1.5B₂O₃–SnO_x/CNFs shows the highest NBO/BO ratio, which should be closely related to the electrochemical performance. Obviously, the chemical bonding in B–C, non-bridging oxygen in B–O⋯Sn, and the bridging oxygen in B–O–B reveal the chemical bonding interactions between B₂O₃ and SnO_x/CNFs, forming the B₂O₃–SnO_x glasses network in CNFs matrix by boron addition. The formed B₂O₃–SnO_x glasses network is beneficial for compensating the large volume change of SnO_x and prohibiting the aggregation of Sn particles by the bonding interactions of NBO during charged/recharged cycles.³³

Fig. 3 (a) XPS survey spectra of 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs; (b) B1s XPS spectra of 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs; (c) Sn3d_5/2 XPS spectra of 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs webs; (d) O1s XPS spectra of 0.5B₂O₃–SnO_x/CNFs, 1.5B₂O₃–SnO_x/CNFs and 3B₂O₃–SnO_x/CNFs composites. Inset: schematic structure of the binding conditions of B, Sn and O in composite showing BO (B–O–B) and NBO (B–O⋯Sn), indicated by solid black rings.

Fig. 4 Charge-discharge curves of (a) 0B₂O₃–SnO_x/CNFs and (b) 1.5B₂O₃–SnO_x/CNFs between 0.005 V and 3.000 V at 200 mA g⁻¹; (c) cycle performance of all samples between 0.005 V and 3.000 V at a current density of 200 mA g⁻¹ and (d) rate performance of 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs at different current densities.

To investigate the electrochemical behavior of the B₂O₃–SnO_x/CNFs electrodes, the 1st, 2nd, 10th, and 20th cycle charge/discharge voltage profiles of the 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs electrodes were measured at a current density of 200 mA g⁻¹, as shown in Fig. 4a and b. The initial charge and discharge capacities for 0B₂O₃–SnO_x/CNFs are 1106.4 and 753.7 mA h g⁻¹. In comparison, the initial charge and discharge capacities for 1.5B₂O₃–SnO_x/CNFs are 1126.2 and 852.2 mA h g⁻¹, corresponding to a columbic efficiency of 75.7%, which is higher than that of 0B₂O₃–SnO_x/CNFs (68.1%). Generally, the initial irreversible capacity is mainly attributed to the formation of solid electrolyte interphase (SEI) layer and the incompletely reversible reaction of SnO_x to Sn and Li₂O.^41,42 Compared with 0B₂O₃–SnO_x/CNFs, the 1.5B₂O₃–SnO_x/CNFs electrode shows a higher initial columbic efficiency and specific capacity, suggesting that the reversibility of the reaction of SnO_x to Sn and Li₂O can be improved by boron addition. After the initial cycle, the columbic efficiency of 1.5B₂O₃–SnO_x/CNFs electrode increases to about 100%, demonstrating the excellent cycling stability. Fig. 4c shows the cycling performance for all the samples in the potential range of 0.005–3 V at a current density of 200 mA g⁻¹. After 100 cycles, the specific capacities for 0B₂O₃–SnO_x/CNFs, 0.5B₂O₃–SnO_x/CNFs, 1.5B₂O₃–SnO_x/CNFs and 3B₂O₃–SnO_x/CNFs are 565.7, 633.4, 670.2 and 580.4 mA h g⁻¹, respectively. It is clear that the 1.5B₂O₃–SnO_x/CNFs electrode exhibits the highest specific capacity of 670.2 mA h g⁻¹ in the 100^th cycle, which is about 20% higher than that of 0B₂O₃–SnO_x/CNFs, and also better than those of other Sn-based composite anode materials such as Sn/SnO_x/CNFs (510 mA h g⁻¹@30 mA g⁻¹, 40^th),¹³ Sn/SnO₂/MWCNT nanocomposites (374 mA h g⁻¹@75 mA g⁻¹, 100^th),⁴¹ carbon coated Sn/graphene nanocomposites (566 mA h g⁻¹@75 mA g⁻¹, 100^th),⁴³ carbon-supported SnO₂ nanowire arrays (403 mA h g⁻¹@160 mA g⁻¹, 60^th),⁴⁴ and graphene-wrapped SnO₂ nanotubes (SnO₂–NTs/G) (422.4 mA h g⁻¹@100 mA g⁻¹, 200^th).⁴⁵ Moreover, the rate capability of the 1.5B₂O₃–SnO_x/CNFs has also been tested at various current densities from 100 to 2000 mA g⁻¹ (Fig. 4d). As observed, the reversible capacities vary along with the discharging rates, and the values changes from 766.5 mA h g⁻¹ (100 mA g⁻¹) to 696.6 mA h g⁻¹ (200 mA g⁻¹), 557.3 mA h g⁻¹ (500 mA g⁻¹) 435.0 mA h g⁻¹ (1000 mA g⁻¹), and 300.2 mA h g⁻¹ (2 A g⁻¹), and then reversible back to 728.1 mA h g⁻¹ (100 mA g⁻¹), demonstrating the high rate capability of the 1.5B₂O₃–SnO_x/CNFs.

To disclose the mechanism for the enhanced electrochemical performances, the cyclic voltammetry (CV) measurement was performed on 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs at a scanning rate of 0.1 mV s⁻¹ with the potential range of 0.005–3 V, as shown in Fig. 5a and b. The first cathodic peak observed at 0.6 V can be ascribed to the formation of SEI film and reduction of SnO_x to Sn and Li₂O, which should be responsible for the large irreversible capacity loss.⁴⁶ Another peak appeared at around 0.01 V due to the insertion of Li⁺ into the CNFs. Two oxidation peaks appeared at around 0.55 V (peak 1) and 1.25 V (peak 2), as described by the following equations:³¹

Sn + xLi⁺ + xe⁻ ↔ Li_xSn, (0 ≤ x ≤ 4.4) (1)

SnO_x + 2xLi⁺ + 2xe⁻ → Sn + xLi₂O, (x ≤ 2) (2)

Fig. 5 CV curves of initial three cycles scanned between 0.005–3 V at a rate of 0.1 mV s⁻¹ for (a) 0B₂O₃–SnO_x/CNFs and (b) 1.5B₂O₃–SnO_x/CNFs; (c) CV curves of 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs electrodes scanned at 0.5, 1, 3 and 5 mV s⁻¹; (d) anodic peak (peak 2) current versus square root of the scan rate for 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs.

The 0.55 V (peak 1) is related to the alloying and dealloying Li_xSn (0 ≤ x ≤ 4.4) according to the eqn (1). An oxidation peak is also observed at about 1.25 V (peak 2) in the first cycle, which is mostly likely to be the formation of metallic Sn and Li₂O as in eqn (2).^34,46 Fig. 5c shows the CV curves of 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs electrodes at different scan rates (0.5, 1, 3, and 5 mV s⁻¹). On the basis of these CV curves, the peak current (I_p, the peak near 1.25 V) is plotted as a function of the root of scan rate (v^1/2) in Fig. 5d, which follows the Randles–Sevcik equation, I_p = (2.69 × 10⁵)n^3/2D^1/2C₀v^1/2, where n is the number of electrons per molecule during the intercalation, C₀ is the concentration of lithium ions, D is the diffusion coefficient, and v is the scan rate.³³ As can be seen from Fig. 5d, the liner correlations between I_p and v^1/2 are obtained for both 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs, and the 1.5B₂O₃–SnO_x/CNFs electrode shows a higher slope than 0B₂O₃–SnO_x/CNFs, indicating that 1.5B₂O₃–SnO_x/CNFs electrode has a larger Li⁺ transfer coefficient compared with 0B₂O₃–SnO_x/CNFs. The significant enhancement of diffusion coefficient (D) in 1.5B₂O₃–SnO_x/CNFs electrode could be attributed to the 3D network in the B₂O₃–SnO_x glasses formed by boron addition, which facilitates ionic charge transfer in electrodes, and thus causing the enhanced rate capability.

The kinetic information for 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs electrodes is also provided by conducting the electrochemical impedance spectroscopy (EIS) measurements after 50 charge–discharge cycles, as shown in Fig. 2S (seen in the ESI†). An equivalent circuit used for fitting the impedance spectra is inserted in Fig. 2S.† Both the two curves consist of a semicircle and a straight line, standing for the overall resistance at interface (R_ct) and diffusion of Li⁺ in the bulk electrode.³³ The value of R_ct for 0B₂O₃–SnO_x/CNFs electrode (109.4 Ω) is larger than that of 1.5B₂O₃–SnO_x/CNFs electrode (71.3 Ω), while the Li⁺ diffusion coefficient (straight line) of 0B₂O₃–SnO_x/CNFs (1.513 × 10⁻³) is lower than that of 1.5B₂O₃–SnO_x/CNFs (2.593 × 10⁻³). These results attest that boron addition into SnO_x can enhance the ionic conductivity with decreasing charge transfer resistance, which is beneficial for improving the rate capability of SnO_x/CNFs electrodes.

The morphologies of 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs electrodes after 100 cycles at a current density of 200 mA g⁻¹ were investigated by HR-TEM. For the 0B₂O₃–SnO_x/CNFs electrode (Fig. 6a and c), after 100 cycles many large particles with diameters in the range of 200 to 300 nm can be clearly found to locate randomly in the CNFs. However, in comparison, no obvious particles can be observed for the 1.5B₂O₃–SnO_x/CNFs electrode (Fig. 6b and d) due to the formation of B₂O₃–SnO_x glasses network in CNFs matrix, which can effectively buffer the large volume change of SnO_x and prohibit the aggregation of Sn particles during charged/recharged cycles,^47–49 causing the high cycle stability performance of the 1.5B₂O₃–SnO_x/CNFs electrode.

Fig. 6 The low resolution TEM images of the electrodes after 100 cycles for (a) 0B₂O₃–SnO_x/CNFs, and (b) 1.5B₂O₃–SnO_x/CNFs; the (c) and (d) are higher resolution for the 0B₂O₃–SnO_x/CNFs and 1.5B₂O₃–SnO_x/CNFs.

Such good rate capability and reversibility of B₂O₃–SnO_x/CNFs could be attributed to the following reasons:

(1) Boron is a unique element as substitution for carbon and Sn to promote the properties of electrical behaviour, leading to high reversibility.

(2) The B₂O₃–SnO_x glasses network by the form of the bridging oxygen (BO: B–O–B) and the non-bridging oxygen (NBO: B–O⋯Sn) in CNFs matrix can provide a great number of active sites for the storage of lithium ions, leading to high capacity behaviours.

(3) The glasses network in the B₂O₃–SnO_x can further effectively buffer the large volume change of SnO_x and prohibit the aggregation of Sn particles during charged/recharged cycles.^47,48

4. Conclusions

B₂O₃–SnO_x/CNFs nanohybrids webs have been successfully prepared via electrospinning and subsequent thermal treatments. Boric acid in precursor solution is transformed to B₂O₃ during the carbonization process, forming the B₂O₃–SnO_x glasses network by non-bridging oxygen (NBO) (B–O⋯Sn) in CNFs matrix. By controlling the addition of boric acid in the precursor solution, an optimal 1.5B₂O₃–SnO_x/CNFs electrode (the approximate weight percent of boric acid to PAN is 1.5 wt%) has been obtained, which shows the highest reversible capacity of 670.2 mA h g⁻¹ at 200 mA g⁻¹ after 100 cycles, and a rate capacity of 300 mA h g⁻¹ at the high current density of 2 A g⁻¹. This excellent lithium-ion storage performance is associated with the 3D SnO_x–B₂O₃ glasses network, which can not only accommodate the enormous volume change associated with lithium insertion/extraction and prevent the SnO_x particles from agglomeration, but also provide enhanced diffusion for Li⁺ and charge transfer. The boron adding to SnO_x/CNFs may be an effective route to develop the Sn-based anode materials with superior lithium-ion storage performance.

Acknowledgements

This work was financially supported by National Natural Science Foundation of China (No. 51072013, 51272021, 51142004 and 51402010) and Natural Science Foundation of Jiangsu Province (No. BK20131147 and BK20140270).

Notes and references

1. J. B. Goodenough and Y. Kim, Chem. Mater., 2010, 22, 587.
2. W. Zhang, J. Power Sources, 2011, 196, 13.
3. B. Scrosati and J. Garche, J. Power Sources, 2010, 195, 2419.
4. V. Etacheri, R. Marom, R. Elazari, G. S. Salitra and D. Aurbach, Energy Environ. Sci., 2011, 4, 3243.
5. Y. Wang, Z.-S. Feng, J.-J. Chen and C. Zhang, Mater. Lett., 2012, 71, 54.
6. Y. Wang, Z.-S. Feng, C. Zhang, L. Yu, J.-J. Chen, J. Hu and X.-Z. Liu, Nanoscale, 2013, 5, 3704.
7. D. Saikia, T.-H. Wang, C.-J. Chou, J. Fang, L.-D. Tsai and H.-M. Kao, RSC Adv., 2015, 5, 42922.
8. L. Zhan, S. Wang, L.-X. Ding, Z. Li and H. Wang, J. Mater. Chem. A, 2015, 3, 19711.
9. J. Zhu, L. Chen, Z. Xu and B. Lu, Nano Energy, 2015, 12, 339.
10. Z. Sun, C. Cao and W.-Q. Han, RSC Adv., 2015, 5, 72825.
11. D. Larcher, S. Beattie, M. Morcrette, K. Edstrom, J.-C. Jumas and J.-M. Tarascon, J. Mater. Chem., 2007, 17, 3759.
12. Y. Zhang, Y. Liu and M. Liu, Chem. Mater., 2006, 18, 4643.
13. L. Zou, L. Gan, R. Lv, M. Wang, Z. Huang, F. Kang and W. Shen, Carbon, 2011, 49, 89.
14. H. Wang, P. Gao, S. Lu, H. Liu, G. Yang, J. Pinto and X. Jiang, Electrochim. Acta, 2011, 58, 44.
15. X. Zhou, L.-J. Wan and Y.-G. Guo, Adv. Mater., 2013, 25, 2152.
16. X. Niu, H. M. Zhou, Z. Y. Li, X. H. Shan and X. Xia, J. Alloys Compd., 2015, 620, 308.
17. X. Lou, J. Chen, P. Chen and L. A. Archer, Chem. Mater., 2009, 21, 2868.
18. I. Meschini, F. Nobili, M. Mancini, R. Marassi, R. Tossici, A. Savoini, M. L. Focarete and F. Croce, J. Power Sources, 2013, 226, 241.
19. B. Jang, S. Park and W. Lee, J. Alloys Compd., 2013, 574, 325.
20. Y. Wang, M. Wu, Z. Jiao and J. Y. Lee, Chem. Mater., 2009, 21, 3210.
21. S.-M. Paek, E. J. Yoo and I. Honma, Nano Lett., 2009, 9, 72.
22. P. Lian, J. Wang and D. Cai, J. Alloys Compd., 2014, 604, 188.
23. W. Zhang, J. Hu, Y.-G. Guo, S.-F. Zheng, L.-S. Zhong, W.-G. Song and L.-J. Wan, Adv. Mater., 2008, 20, 1160.
24. Y. Yu, Q. Yang, D. Teng, X. Yang and S. Ryu, Electrochem. Commun., 2010, 12, 1187.
25. L. Zou, L. Gan, F. Kang, M. Wang, W. Shen and Z. Huang, J. Power Sources, 2010, 195, 1216.
26. Y. Wang and T. Chen, Electrochim. Acta, 2009, 54, 3510.
27. C. Li, W. Wei, S. Fang, H. Wang, Y. Zhang, Y. Gui and R. Chen, J. Power Sources, 2010, 195, 2939.
28. D. Kim, D. Lee, J. Kim and J. Moon, ACS Appl. Mater. Interfaces, 2012, 4, 5408.
29. Y. Liu, X. Yan, J. Lan, D. Teng, Y. Yu and X. Yang, Electrochim. Acta, 2014, 137, 9.
30. X. Liu, D. Teng, T. Li, Y. Yu, X. Shao and X. Yang, J. Power Sources, 2014, 272, 614.
31. R. Li, D. Li, D. Tian, G. Xia, C. Wang, N. Xiao, N. Li, N. H. Mack, Q. Li and G. Wu, J. Power Sources, 2014, 251, 279.
32. G. Xia, N. Li, D. Li, R. Liu, N. Xiao and D. Tian, Mater. Lett., 2012, 79, 58.
33. A. Hayashi, M. Nakai, M. Tatsumisago, T. Minami, Y. Himei, Y. Miura and M. Katada, J. Non-Cryst. Solids, 2002, 306, 227.
34. J. Chen, Y. Tan, C. Li, Y. Cheah, D. Luan, S. Madhavi, F. Y. C. Boey, L. A. Archer and X. Lou, J. Am. Chem. Soc., 2010, 132, 6124.
35. R. Trócoli, S. Franger, M. Cruz, J. Morales and J. Santos-Pena, Electrochim. Acta, 2014, 135, 558.
36. X. Bai, B. Wan, H. Wang and J. Jiang, J. Alloys Compd., 2015, 628, 407.
37. Z. Wu, W. Ren, L. Xu, F. Li and H. Cheng, ACS Nano, 2011, 5, 5463.
38. C. E. Lowell, J. Am. Ceram. Soc., 1967, 50, 142.
39. D.-W. Wang, F. Li, Z.-G. Chen, G. Q. Lu and H.-M. Cheng, Chem. Mater., 2008, 20, 7195.
40. B. Zhang, Y. Yu, Z. Huang, Y. He, D. Jang, W. Yoon, Y. Mai, F. Kang and J. Kim, Energy Environ. Sci., 2012, 5, 9895.
41. M. Alaf, D. Gultekin and H. Akbulut, ACS Appl. Mater. Interfaces, 2013, 275, 244.
42. X. Xia, X. Wang, H. Zhou, X. Niu, L. Xue, X. Zhang and Q. Wei, Electrochim. Acta, 2014, 121, 345.
43. D. Wang, X. Li, J. Yang, J. Wang, D. Geng, R. Li, M. Cai, T. K. Sham and X. Sun, Phys. Chem. Chem. Phys., 2013, 15, 3535.
44. Z. Yang, S. Zhao, W. Jiang, X. Sun, Y. Meng, C. Sun and S. Ding, Electrochim. Acta, 2015, 158, 321.
45. P. Wu, X. Xu, Q. Zhu, X. Zhu, Y. Tang, Y. Zhou and T. Lu, J. Alloys Compd., 2015, 626, 234.
46. Y. Xu, Q. Liu, Y. Zhu, Y. Liu, A. Langrock, M. R. Zachariah and C. Wang, Nano Lett., 2013, 13, 470.
47. X. Wu, Z. Wang, L. Chen and X. Huang, Solid State Ionics, 2004, 170, 117.
48. H. Xiang, S. Fang and Y. Jiang, Solid State Ionics, 2002, 148, 35.
49. T. Yang and B. Lu, Phys. Chem. Chem. Phys., 2014, 16, 4115.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra19387g

‡ These authors contributed equally to this work.

---