Sword-sheath structured CNT@SnO₂@CNx boosting the electrochemical performance in lithium-ion batteries

1. Introduction

With the fast-growing economy, exploiting sustainable energy, such as solar, hydrogen, and wind energy, has become a hot topic [1]. Lithium-ion batteries (LIBs) have aroused great interest because of their light weight, long lifespan, and excellent energy density [2,3]. However, anode materials are still the bottleneck for LIBs due to the low capacity of commercial graphite. It is challenging to design advanced anode materials with satisfying capacity and cycle life [4]. Recently, tin oxide (SnO₂) has become a hot star material as an anode material in LIBs, arising from its high theoretical capacities (781 mAh g^-1), low voltage window, and abundance on Earth [5,6]. However, the large high volume change (>300%) and serious agglomeration of SnO₂ nanoparticles during the lithiation/delithiation process will lead to the electrical contact loss and aggregation, resulting in a poor electrochemical performance. In addition, the lower conductivity of SnO₂ will cause larger charge transfer resistance and lower the rate performance of LIBs [7,8]. To overcome these drawbacks, many strategies have been developed to resolve these problems. In a response, reducing SnO₂ to the nanoscale with various nanostructures such as nanorods [9], nanosheets [10,11], nano-flower [12], and nanoflakes [13] can greatly improve the capability in LIBs. These materials with nanoscale features can offer a shorter pathway for Li⁺ ions and electrons, as well as increase the contact interface that allows a better electrochemical performance to be achieved. However, nanoscale SnO₂ is not sufficient to avoid the capacity decay due to the unavoidable volumetric expansion after longer cycling [14].

Meanwhile, incorporating SnO₂ with higher conductivity nanocarbons such as graphene, carbon nanotubes (CNTs), and a carbon coating layer has been proven to be an effective strategy to enhance the performance in LIBs [15-17]. The contribution of carbonaceous materials in the composites can be divided into two types: 1) carbon matrix and 2) carbon coating layer. The carbon matrix plays an important role for SnO₂, which could protect the SnO₂ from aggregation and pulverization during the electrochemical process. For the carbon coating layer, the conductivity of this carbon layer was employed as a protection layer to modify the surface of SnO₂ for improving the surface chemistry of the composites, especially the heteroatom-doped carbon layer [18,19]. Thus, SnO₂-carbon-based composites with unique nanostructure materials are considered as potential candidates for LIBs when serving as anode materials. However, simply directing the incorporation of SnO₂ with a carbon matrix will not solve the volumetric effect induced by SnO₂.

Based on the above descriptions, new strategies to address the above problems still need to be explored. In the present work, we designed a unique sword-sheath structure with a void space between CNTs and SnO₂, which includes an N-doped carbon coating layer as an anode material for LIBs. Within these unique nanostructures, the void space located between the conductivity skeleton CNTs and SnO₂ can mostly prevent the volume expansion of tin oxide during lithiation/delithiation reaction. These significant advances will endow the nanostructure with enhanced battery behavior in LIBs.

2. Materials and Methods

3. Results and Discussion

As shown in Figure 1(a), first, CNT@m-SiO₂ (mesopore-SiO₂) structure was fabricated by the sol-gel method using CNTs as the conductivity skeleton. The mesopore-SiO₂ not only acted as the template for the void space but also provided more active sites for the following SnO₂ deposition process. Then the SnO₂ nanoparticles were encapsulated in mesoporous silica to form CNT@m-SiO₂@SnO₂ via the hydrothermal method. After removing m-SiO₂ from CNTs, glucose and dopamine were used as carbon sources to coat CNT@SnO₂ to protect the whole composite from the electrolyte and improve the conductivity. TEM and HRTEM analyses were performed to investigate the morphological evolution and structural characteristics of the samples throughout the preparation process. It can be discerned from Figure 1(b) and Figure S1(a) that the original CNTs with a smooth surface are coiled with each other to form the conductive network with a diameter of 40 nm. After the sol-gel reaction, mesoporous silica (m-SiO₂) is well embraced on the CNTs with a thickness of around 20 nm (Figure 1c and Figure S1b), and the uncoated CNT tips were marked by white arrows (Figure S1). To confirm the porosity of the CNTs and CNT@m-SiO₂, N₂ adsorption/desorption isotherms and pore size distribution (PSD) plots were investigated in Figures S2 (a and b). Typically, Nitrogen adsorption analysis reveals that CNT@m-SiO₂ displays typical type-I isotherm characteristics: strong adsorption at low pressures (P/P₀ < 0.1) due to micropore filling; a well-defined adsorption knee representing mesopore capillary condensation; and a subsequent plateau. The bimodal pore size distribution quantitatively confirms this hierarchical porosity, showing distinct peaks for micropores (<1 nm) and mesopores (2-4 nm). The porosity properties of CNT@m-SiO₂ are 1330 cm² g^-1, which is much higher than that of CNTs (95.5 cm² g^-1), indicating that m-SiO₂ provided more active sites for SnO₂ deposition.

Close

Figure 1.

(a) Schematic illustration for the synthesis of sword-sheath structured CNT@SnO2@C and CNT@SnO₂@CNx. TEM images of (b) original CNTs; (c) CNT@m-SiO₂; (d) CNT@m-SiO₂@SnO₂; (e) CNT@SnO₂; (f) CNT@SnO₂@C, (g) CNT@SnO₂@CNx.

Export to PPT

The TEM image of CNT@m-SiO₂@SnO₂ has been shown in Figure 1(d). It can obviously be seen that the surface became rough and SnO₂ is uniformly and completely deposited on the CNT@m-SiO₂. After removing the m-SiO₂, the void space is observed between CNTs and the SnO₂ layer, which is marked with a red arrow in Figure 1(e). To further confirm the presence of the void space, SEM and EDX Mapping of CNT@SnO₂ were obtained. As the EDX mappings shown in Figures S1(e-g), there is a uniform dispersion of the elements of C, O, and Sn elements in CNT@SnO₂ without Si elements, indicating that the NaOH etching process can completely remove the m-SiO₂ template. The final products (CNT@SnO₂@C and CNT@SnO₂@CN _x) are also confirmed by TEM images in Figures 1(f,g), in which each one can observe the carbon layer attached to the CNT@SnO₂ with a thickness of around 20 nm. Corresponding HRTEM images showed the clear lattice fringe spacing of the crystallites, 0.32 nm and 0.34 nm, matching the (110) and (001) planes of SnO₂ in the inside figure of Figures 1(f,g)[23]. These results declare that sword-sheath CNT@SnO₂@C and CNT@SnO₂@CN _x nanostructures with a void space and a carbon coating layer were successfully prepared.

After that, the crystal structure and phase of each process sample were investigated by XRD, including CNTs, CNT@m-SiO₂, CNT@SnO_2, and CNT@SnO₂@C and CNT@SnO₂@CN _x. As shown in Figure S3, the CNTs exhibit a sharp diffraction peak at 26.5°, which can be assigned to the (002) plane of graphite, and the broad diffraction peak at about 23° is ascribed to the (011) plane of the amorphous m-SiO₂. These diffraction peaks disappeared in the CNT@SnO₂ after NaOH etching and were replaced by the diffraction angles at 2θ = 26°, 33°, and 51° correspond to the planes of (110), (101), and (211) belong to SnO₂ (JC-PDS:41-1445), implying that m-SiO₂ was completely etched away effectively and SnO₂ still remained in the sample [24]. As shown in Figure 2(a), for both of CNT@SnO₂@C and CNT@SnO₂@CN _x samples, the XRD characteristics are essentially similar to CNT@SnO₂, indicating that the carbothermal reduction did not occur between SnO₂ and carbon. It is difficult to distinguish the carbon coating layer in XRD analysis due to its amorphous structure related to the lower treatment temperature (450°C). Because of the strong crystalline nature of SnO₂, its diffraction peaks appear to dominate the pattern, making the CNT peaks indistinguishable. TGA measurement was carried out to determine the loading contents of SnO₂ under air flow from 25°C to 900°C, which means that the mass weight of SnO₂ is 57.23 wt% and 58.61 wt% for CNT@SnO₂@C and CNT@SnO₂@CN _x, respectively. And there were two, both similar in a weight loss range of CNT@SnO₂@C and CNT@SnO₂@CN _x.

Close

Figure 2.

(a) XRD patterns of CNT@SnO₂@C and CNT@SnO₂@CNx; (b) TGA curves of each step samples; (c) XPS survey spectrum of CNT@SnO₂@C and CNT@SnO₂@CNx; High-resolution XPS spectra of de-convoluted (d) O 1s; (e) Sn 3d; and (f) N 1s of CNT@SnO₂@CNx.

Export to PPT

Figure 2b shows the thermal stability of all the sample in air. The chemical bonds and compositions of CNT@SnO₂@C and CNT@SnO₂@CN _x were analyzed by XPS to determine their surface chemical composition and electronic states, as shown in Figures 2(c-f) and Figures S4(a-d). It’s clear to demonstrate from the full spectra that the existence of C1s (285 eV), O1s (533 eV), and Sn 3d (487 eV) elements in both samples, particularly, the N species (399 eV), was only exhibited in CNT@SnO₂@CN _x with the content of 3.38 at% (Table S1). As shown in Figures S4(a and b), the C1s peak was resolved into four contributions: C-C (283.29 eV), C-OH (284.74 eV), C=O (286.22 eV), and COOH (288.53 eV), respectively [25]. The strong peaks at 531.94eV and 533.20 eV could be attributed to the covalent bonding of Sn-O and -C-OH in the high-resolution O_1s XPS spectrum (Figure 2d) [6,12], and another peak located at 486.78eV and 495.12 eV confirmed the presence of Sn 3d_5/2 and Sn 3d_3/2 related to SnO₂ (Figure 2e) [8]. Combining the high resolution XPS spectra of Sn 3d and O 1s, it can be concluded that the presence of SnO₂, which was confirmed by XRD. Furthermore, the high-resolution XPS spectrum of N1s (Figure 2f) could be resolved into 397.384 eV and 399.748 eV; the observed spectral signatures correspond to pyridinic (N-6) and pyrrolic (N-5) nitrogen coordination environments, respectively [26]. These results clearly suggested that nitrogen atoms were successfully doped into the carbon matrix. Nitrogen functional groups (pyridinic/pyrrolic) create defect-rich domains in the carbon matrix and increase the electronic conductivity of the carbon matrix, enabling enhanced electrochemical performance in LIBs.

Based on the TEM and XRD results, CNT@SnO₂@C and CNT@SnO₂@CN_x hybrid composites were successfully prepared. Above all, the unique structure CNT@SnO₂@C and CNT@SnO₂@CN_x nanocomposites exhibit superior electrochemical performance for LIB applications through three synergistic mechanisms: Interstitial voids buffer SnO₂ volume changes (∼300%) during lithiation; reduced Li⁺ diffusion distances enable fast kinetics; continuous CNT networks maintain electrical connectivity throughout cycling. Secondly, CNTs as the framework can offer high conductivity and mechanical strength for the whole composite. Moreover, the outer carbon coating layer would provide a protective screen for the inner material, in favor of the formation of a stable solid electrolyte interphase (SEI) layer.

To evaluate the lithium storage performance of the SnO₂-based materials, half-cell electrochemical tests were systematically conducted. Figure 3a and Figure 3b. presents the cyclic voltammetry curves of CNT@SnO₂@C and CNT@SnO₂@CNₓ during the first three cycles at a scan rate of 0.5 mV s^-1 between 0.005-3.0 V, showing characteristics consistent with reported SnO₂-carbon composites [27]. The initial cathodic scans reveal broad peaks at 0.76 V for CNT@SnO₂@C and 0.95 V for CNT@SnO₂@CNₓ, corresponding to solid electrolyte interphase formation, with the positive voltage shift in the nitrogen-doped sample suggesting modified interfacial reactions. Subsequent cycles demonstrate stabilized redox activity with characteristic peaks around 0.1/0.6V, indicating reversible lithium alloying/dealloying in the tin-based components. The improved cycling stability observed after the initial cycle reflects enhanced electrode-electrolyte interface compatibility in these composite architectures. Two reduction peaks at 1.1V and 0.4V correspond to SnO₂ conversion to Sn and subsequent Li-Sn alloying, respectively. (SnO₂ + 4Li⁺ + 4e^- ↔ Sn + 2Li₂O; Sn + xLi⁺ xe^- ↔ Li _xSn ($0\le \text{x}\le 4.4$) [6,28]. The cathodic scan peak at around 0.07V can be found, which corresponds to the Li⁺-insert carbonaceous material to form Li _xC. (C+ xLi⁺ ↔ Li _xC) [29]. During charging, the distinct 0.6 V peak signifies Li extraction from LiₓSn, while the broader 1.2 V feature indicates Sn reoxidation to SnO₂, respectively (Sn + 2Li₂O ↔ SnO₂ + 4Li⁺ + 4e^-; Li _xSn ↔ Sn + xLi⁺ xe^- ($0\le \text{x}\le 4.4$) [28]. In addition, these oxidation/reduction peaks in the following scan curves show a good repeatability after the first scan, indicating its excellent reversibility and the stability of the SEI on the electrode [30].

Close

Figure 3.

CV curves (a) CNT@SnO₂@C and (b) CNT@SnO₂@CNx at a scan rate of 0.5 mV s^-1. (c) Galvanostatic first charge/discharge curves at a current density of 0.1 A g^-1; (d) rate performance at various current density, and (e) cycling performance at 0.1 A/g of CNT@SnO₂ and CNT@SnO₂@C.

Export to PPT

The first-cycle electrochemical performance of the three composites (CNT@SnO₂, CNT@SnO₂@C, and CNT@SnO₂@CNₓ) was evaluated through galvanostatic charge/discharge testing at 0.1 A g^-1 (Figure 3c and Figure S5). The discharge curves show a clear plateau in all the samples, which are in good agreement with the peaks in the CV curves. The charge and discharge capacity values of CNT@SnO₂, CNT@SnO₂@C, and CNT@SnO₂@CN _x are 887.3/1678.4 mAh g^-1, 1000/1356.5 mAh g^-1, and 827.9/1403.8 mAh g^-1, achieving the Coulombic efficiency of 52.87%, 73.72%, and 58.90%, respectively. The measured Coulombic efficiency reached 52.87%, reflecting considerable capacity consumption in SEI formation [31]. In addition, CNT@SnO₂@C and CNT@SnO₂@CN _x show a higher Coulombic efficiency than CNT@SnO₂, indicating that the carbon layer allowed the formation of a stable SEI film and protected the SnO₂ from coarsening successfully during battery cycling. Moreover, the value of the first discharge of CNT@SnO₂@CNx is higher than that in CNT@SnO₂@C, indicating the N-doped carbon layer could provide more sites for Li⁺ storage. And CNT@SnO₂ exhibits the highest discharge capacity among all the samples, which is a common phenomenon and related to the direct exposure under the electrolyte [32].

With outstanding rate capability, the CNT@SnO₂@CN _x electrode achieved a stable average reversible capacity of 956.4 mAh g^-1 which is much higher than SnO₂ (Figure S6).

Figure 3(d) represents the rate performance of three samples within a voltage range of 0.005-3V and current density from 0.1 A g^-1 to 5 A g^-1. With outstanding rate capability, the CNT@SnO₂@CN _x electrode achieved a stable average reversible capacity of 956.4 mAh g^-1, 919.0 mAh g^-1, 817.6 mAh g^-1, 740.9 mAh g^-1, 655.6 mAh g^-1, 533.6 mAh g^-1, and at different current densities. In addition, when the current density was restored to 0.1 A g^-1 following multi-rate cycling at varying current densities, the electrode recovered its initial capacity of 856.2 mAh g^-1, demonstrating an excellent rate performance. It is quite clear that nitrogen-functionalized carbon outer layers in the sword-sheath configuration synergistically improve electronic conduction and create a shortened Li⁺ transport pathway [33]. For those reasons, CNT@SnO₂@CN _x shows a preferable rate capacity than that of CNT@SnO₂@C. Moreover, CNT@SnO₂@C and CNT@SnO₂@CN _x show higher rate capacities than CNT@SnO₂, which shows a reversible capacity of 708.5 mAh g^-1 and 342.7 mAh g^-1 at 0.1 and 5 A g^-1, respectively.

Cycling performance and corresponding Coulombic efficiency of CNT@SnO₂, CNT@SnO₂@C, and CNT@SnO₂@CN _x at 0.1 A g^-1 were tested and shown in Figure 3(e). It can be seen under the identical experimental conditions, CNT@SnO₂@CN _x sustains 1087.5 mAh g^-1 after 100 cycles (∼100% Coulombic efficiency post-first cycle), outperforming CNT@SnO₂@C (964.8 mAh g^-1). As for CNT@SnO₂, a rather high capacity of 913.6 mAh g^-1 is recorded initially. After 40 cycles, the capacity rapidly falls to 669.1 mAh g^-1. The SEM image of CNT@SnO₂@CN _x after cycling has been shown in Figure S7. The poor cycling performance originates from the absence of a carbon sheath related to the large volume expansion/contraction and SnO₂ particle aggregation during several repeated charge and discharge processes. These results demonstrate that sword-sheath structures with N-doped carbon sheath could effectively maintain the integrity of architectures and significantly improve the cycling stability of SnO₂ nanoparticles. Compared with previously published results with carbon-based SnO₂ electrode for LIBs (Table S2), the designed sword-sheath architecture endows CNT@SnO₂@CNₓ with unprecedented reversible capacity and cycling stability, surpassing previous records in the literature.

Electrochemical impedance spectroscopy (EIS) was utilized to examine the electrode reactions. A representative equivalent circuit (inset, Figure 4a) was constructed, and its components were adjusted to obtain an optimal fit to the measured spectra. In Figure 4, all the Nyquist plots have three components, including an Rₛ-dominated intercept at high frequencies (Z’ axis), while the low-frequency response exhibits a slope influenced by Rₜ (charge transfer resistance) and Z_w (Warburg impedance due to Li⁺ diffusion in the electrode) [7].

Close

Figure 4.

(a) Nyquist plots with equivalent circuit pattern as insert and (b) the relationship plots between Z’ and ω-1/2 CV in the low frequency region of CNT@SnO₂@C and CNT@SnO₂@CNx.

Export to PPT

Following the equivalent circuit in Figure 4(a), CNT@SnO₂@CN _x (3.1Ω) exhibited a smaller value of R_s than that of CNT@SnO₂@C (14.5Ω), indicating Li⁺ showed a faster diffusion rate within the SEI layer. In addition, the R_ct of CNT@SnO₂@CN _x (57.7 Ω) is much lower than that of CNT@SnO₂@C (75.9Ω), much higher than CNT@SnO₂ (Figure S8), revealing an enhanced conductivity that is in good agreement with its excellent electrochemical performance [34]. Based on EIS results, the N-doped carbon coating significantly boosts electrode conductivity, accelerating charge transfer while creating more Li⁺ storage sites [35]. According to the previous report, the Warburg factors (σ) are an important parameter to evaluate the ions diffusion ability (D) of the electrodes, which can be calculated from the relationship between Z’ and ω-1/2 at low frequencies. Based on equations (1) and (2), a smaller Warburg factor (σ) means a faster ion diffusion rate (D). As shown in Figure 4(b), the smaller σ of CNT@SnO₂@CN _x indicated the larger D and faster Li⁺ storage and release reaction during long-term cycles.

Based on the above experimental results, when served as the anode material of LIBs, the sword-sheath structured CNT@SnO₂@CN _x is capable of effectively buffering the volume variation in SnO₂ particles in the cycling process. Simultaneously, the existence of CNT and carbon coating layer can enhance both electrical conductivity and structural integrity of the hybrid composite. All these advantages endow the sword-sheath structured CNT@SnO₂@CN _x with excellent electrochemical performance in LIBs.

4. Conclusions

In conclusion, a facile method was demonstrated for the synthesis of sword-sheath nanostructured CNT@SnO₂@C and CNT@SnO₂@CN_x, the void space within CNT@SnO₂@C and CNT@SnO₂@CN _x nanostructures enables avoiding large volume change of SnO₂ and fastens the Li⁺ diffusion pathway, leading to excellent electrochemical performance achieved. Moreover, within this nanostructure, the N-doped carbon layer and CNT can not only enhance the electrochemical stability but also effectively enhance the conductivity of the hybrid composite for enhanced Li⁺ ion storage. Benefiting from these properties, the unique sword-sheath nanostructures allow for greatly improved performance of LIBs. When evaluated as the anode materials for LIBs, CNT@SnO₂@CN _x showed super cycle and rate performance, delivering a specific capacity of 1087.5 mAh g^-1 at a current density of 0.1 A g^-1 after 100 cycles. It is believed that the smart architecture designed for sword-sheath structured CNT@SnO₂@CNx can provide practical guidance for a new generation of electronic devices.