Capacity Contribution Mechanism of rGO for SnO₂/rGO Composite as Anode of Lithium-ion Batteries

Lithium-ion batteries have realized large-scale application in electric and hybrid vehicles. In the past few decades, SnO₂ has attracted extensive attention as electrode material. Compared with ordinary graphite anode materials, SnO₂ possesses higher theoretical specific capacity (790 mA·h·g^-1), rich raw materials and low price. While the volume expansion of SnO₂ during lithium-ion extraction/intercalation reaches more than 50%, the high stress caused by volume expansion may lead to the fracture failure of electrode material which makes the specific capacity of the material decrease rapidly, while the rate performance and cycle stability decreasing [1‐3]. Therefore, the key to realize the application of SnO₂ cathode materials in lithium-ion batteries is to find ways to effectively slow down the volume dilatation effect of SnO₂ during the lithium-ion extraction/intercalation process and upgrade the electrochemical performance of electrode materials [4‐7]. Graphene is a 2D material with good conductivity and high specific capacity. Graphene nanosheets can not only effectually prevent the volumetric change and particle aggregation of SnO₂, but also improve the conductivity. As an active material for lithium storage, graphene nanosheets keep the structural integrity of electrode [8‐10]. Chen et al. synthesize a reduced graphene and SnO₂ nanospheres composite and obtain good performance in the application of lithium-ion batteries or sodium-ion batteries [11]. Zhou et al. built a SnO₂/GO structure by hydrothermal method [12]. It’s worth noting that they researched the influence of GO dosage in SnO₂/GO anode and found that the lithium ions storage capability increases with the raise of GO dosage. Lu et al. focused on the drying process after hydrothermal process [13]. By comparing anodes treated by spray-drying and freeze-drying, novel composition of SnO₂ and GO achieved after spray-drying process obtained favorable lithium-ion transmission during charging/discharging process.

The combination of rGO and anode materials of lithium-ion batteries or sodium-ion batteries can increase the capacitance contribution, which has been confirmed by much research [14‐17]. While, for SnO₂, the influence of rGO on capacitance contribution rate has not been studied. In this work, we report to synthesize SnO₂/rGO composite by the sample single step method. We analyze the lithiation/delithiation process and find that the addition of rGO will significantly improve the mobility of lithium ions. When the scan rate comes to 1 mV·s^-1, the capacitive contribution rate raises from 56% to 71% after the addition of rGO. It confirms that the addition of rGO changes the capacity contribution mechanism from diffusion process dominance to surface driven capacitive contribution dominance attributed to high specific surface area, abundant defects and high conductivity of rGO.

For the sake of analyzing the morphology and nanostructure of hydrothermal product, SEM test was carried out on SnO₂ and SnO₂/rGO powders. Figure 1a shows the pure SnO₂ obtained by hydrothermal treatment presents spherical morphology with a diameter of around 1 to 3 microns. Figure 1b illustrates that SnO₂ spheres (with a diameter of approximately 3 μm) are composed of several spherical nanoflowers with a diameter of about 200 nm. As for SnO₂/rGO (Figure 1c and d), spherical SnO₂ are equably absorbed on the surface of rGO nanosheets, meanwhile the SnO₂ spherical structure remains unchanged. It’s obviously that SnO₂ nanoflowers grown on rGO nanosheets, while the size is obviously smaller than that of SnO₂ powders which is about 100 nm. The reduction of the diameter of the nanosphere means the increase of the specific surface area.

TEM spectra (Figure 1e, f) further illustrate the morphology and nanostructure of as prepared materials. It is highly consisted with the results obtained in previous SEM test. SnO₂ nanoparticles are absorbed onto the rGO nanosheets with favorable connection. As for high-resolution TEM spectra (Figure 1f), SnO₂ crystal grains are relatively visible and the rGO nanosheets could be obviously observed and the lattice spacing of 0.33 nm, corresponding to the (110) planes of tetragonal SnO₂. X-ray diffraction (XRD) is tested for the component analysis of SnO₂/rGO and SnO₂ powder (Figure 2a). As for the XRD spectroscopy of SnO₂/rGO, diffraction peaks in the curve could correspond to that of tetragonal tin dioxide with cassiterite structure (JCPDS Card No.41e1445). While, it’s of particular interest that, the XRD spectroscopy of SnO₂ failed to correspond the standard peak one-to-one. This is because the crystallinity of SnO₂ obtained after 6 h of hydrothermal reaction is not high enough, which shifts the peak of 33.8° to the left [22]. Since SnO₂/rGO failed to detect the diffraction peak of rGO, the XRD spectrum results could not prove the existence of rGO.

For the common homonuclear diatomic pairs in carbon-based materials, their Raman activity will be strong, so their Raman peaks can be easily detected in Raman spectra, thus the Raman spectroscopy was used for further composition analysis (Figure 2b). There are two obvious peaks appearing in ~ 1350 cm^-1 and ~ 1580 cm^-1 corresponding to the D and G bands of graphene respectively. The I_D/I_G significantly increases from 0.98 of GO to 1.76 of SnO₂/rGO, which indicates that the sp2 domain of SnO₂/rGO composites is smaller. Meanwhile it illustrates that oxygen content of GO reduced during hydrothermal process [21].

In order to further analyze the components and verify the partial reduction of GO during the hydrothermal process. FT-IR spectra of as-prepared nanocomposites are plotted (Figure 2c). Both spectra show the peak of 3440 cm^-1 and 1629 cm^-1 which correspond to O-H stretching vibrations of absorbed water molecules and O-H bending vibrations, the absorption band of around 1220 cm^-1 corresponding to C-O. After compounding with SnO₂, similar oxygen-containing groups are found in the SnO₂/rGO spectra, which illustrates that GO is partially reduced in hydrothermal treatment, and a strong peak at 669 cm^-1 is attributed to O-Sn-O stretching mode [22, 23]. N₂ adsorption/desorption test is carried out to analyze the specific surface area of SnO₂ particles (purchase from enterprise), SnO₂ nanoflower and SnO₂/rGO composite in Figure 2d. The BET specific surface area of SnO₂ particles is only 23.16 cm²·g^-1, while it comes to 63.17 cm²·g^-1 when the morphology of SnO₂ turns into spherical nanoflower. As for SnO₂/rGO composite, it reaches 126.74 cm²·g^-1, which is much higher than that of SnO₂. It illustrates that the assistance of rGO could significantly increase the specific surface area [13]. The larger specific surface area means that SnO₂/rGO can have better contact with electrolyte and absorb more lithium ions at the same time.

To obtain the exact composition of SnO₂/rGO powders, we tested it with X-ray photoelectron spectroscopy (XPS), since XPS could display the chemical environment of Sn, O and C elements. Figure 3a displays the XPS survey spectra of SnO₂/rGO powders. In C1s fitting patterns (Figure 3b), there are C-C bond at 284.7 eV, C-O bond at 285.6 eV and C=O bond at 286.8 eV, in which C-C bond takes an absolute dominant position. It corresponds to a common rGO chemical bond composition as mentioned in previous study [24]. In O1s spectroscopy (Figure 3c), Sn-O bond at 531.6 eV and C-O-Sn bond at 532.8 eV make different contribution in fitting patterns. The obvious C-O-Sn bond fitting peak indicates that there exists surpassing chemical bonding among SnO₂ and rGO [25]. In Sn3d spectrum (Figure 3d), there are two obvious peaks located at 495.9 eV and 487.6 eV, which correspond to Sn3d 3/2 and Sn3d 5/2 respectively [26]. The peaks generally consist with that reported in pure SnO₂ powder; it illustrates that major constituent of SnO₂/rGO powders is SnO₂.

For the sake of researching the influence on lithium-ion battery performance after compounding with rGO, the electrochemical performance tests were carried out. Figure 4a display the initial three cycle cyclic voltammetry (CV) profiles of the electrode which chose SnO₂/rGO composites as active materials in the voltage range from 0.01 to 3.0 V. In the first cycle, the cathodic peak at around 0.78 V could be observed while it disappeared in the subsequent cycles. This situation occurs due to the irreversible reduction of SnO₂ and the formation of solid electrolyte interface (SEI), as described in Eq. (1) [27]. Furthermore, it will lead to significant capacity loss in the first cycle. An obvious reduction peak appeared at the range of 0.08 and 0.13 V, which correspond to the alloying reaction of Sn and Li (Eq. (2)) [28]. Meanwhile it is relevant to the intercalation of lithium [29]. As for the following two cycles, the obvious cathodic peaks at around 1.2 V correspond to the transformation between SnO₂ and Sn [30]. In the charge curve, the anodic peak at around 0.57 V correspond to the reverse process of the alloying reaction mentioned before, which means the decomposition of Li_xSn [31]. There were two weak oxidation peaks at 1.27 and 1.9 V, which corresponded to reversible reaction of SnO₂ and Li₂O. In comparison of the CV curves for the first three cycles, the crest value of alloying/dealloying reaction showed slight decline as cycle number increased, it indicated that SnO₂ nanoparticles brought out further activation during the test period. The CV curves of later cycles had significant overlap ratio, which indicated that SnO₂/rGO has excellent reversibility [28].

When the specific current reaches 0.1 C, the galvanostatic discharge-charge curves of 1st, 2nd, 5th and 10th for the SnO₂/rGO are shown in Figure 4b. In the first cycle, the discharge and charge capacity of the composite are 1114.3 mA·h·g^-1 and 731.9 mA·h·g^-1, respectively. Compare our study with other SnO₂/rGO materials as the anode for Li ion batteries, the capacitance performance of SnO₂/rGO at 0.1C (1114.3 mA·h·g⁻¹) is superior to SnO₂ NPs/rGO (1100 mA·h·g⁻¹) [11], CGN/SnO₂ composites (1048.7 mA·h·g⁻¹) [32] and SnO₂-RGO (665 mA·h·g⁻¹) [33]. A voltage plateau appears at approximately 1 V in the first cycle, while it can’t be observed in the subsequent cycles. It mainly due to irreversible reduction of SnO₂ and electrolyte decomposition. In the following profiles, the voltage plateaus can be observed in the ranges of 0.01–0.5 V and 0.5–1 V corresponding to combination reaction of Li-Sn alloy. These results are highly consistent with the verdicts obtained in previous CV tests.

The cyclic performance and coulombic efficiency of two materials under current density of 0.1 C is displayed in Figure 4c. The initial cycle of SnO₂/rGO galvanostatic discharge-charge give out 959.2 mA·h·g^-1 for discharge capacity. In contrast, SnO₂ delivers a discharge capacity of 754.6 mA·h·g^-1 with coulombic efficiency of 59.04%. It’s obviously that, SnO₂/rGO composite possess much higher discharge capacity and coulombic efficiency than that of SnO₂ nanoparticles. Besides, the decline of discharge capacity with the increase of cycle number for SnO₂/rGO composite is significantly faster than that of SnO₂ nanoparticles. When the cycle number comes to 100, SnO₂/rGO remains discharge capacity of 469.6 mA·h·g^-1, and SnO₂ nanoparticles reach to 121.1 mA·h·g^-1. The above phenomenon is mainly caused by the following two reasons. Firstly, the SnO₂/rGO composite possess much smaller grain size, which will lead to more active materials participating in the reaction during charge and discharge. Secondly, rGO provides an attached framework for SnO₂. This structure puts out an effective buffer for the volume change in the reaction process. Meanwhile it facilitates the insertion and detachment of lithium ions.

The rate performance of two materials at gradient current densities are displayed in Figure 4d. After 10 cycles charge-discharge process, SnO₂/rGO composite remains capacity of 773.7 mA·h·g^-1 at 0.1 C, while SnO₂ nanoparticles come up to 500.9 mA·h·g^-1 at the same condition. With rate raising up from 0.1 C to 2 C, rate capability diminishes gradually, moreover, the lithium ions storage capability difference of two materials becomes much larger. The reversible capacity of SnO₂/rGO composite come up with 210 mA·h·g^-1, demonstrating admirable reversible capacity and structure stability under large steady current density [12]. When the rate turns back to 0.1 C, the rate capability of SnO₂/rGO composite is basically restored indicating the great reversible capacity. As for Figure 4e, SnO₂/rGO displays great long-cycle stability even at current charge and discharge rate of 1 C. When the cycle number comes to 1000, the discharge capacity still remains about 300 mA·h·g^-1. Obviously, the excellent rate performance and cycle stability are attributed to the addition of rGO. It remarkably improves the electrode material stability and provides more lithium ions storage sites [34].

EIS spectra were performed to determine the electrical conductivity and electrochemical kinetics of two electrodes (Figure 5a). The composition of the equivalent circuit was shown above the fitting plot, which can be used for quantitative analysis. A lower charge transfer resistance (R_ct) value of SnO₂/rGO composite (65.7 ohms) indicates that the electrode gains better kinetics after introducing rGO into SnO₂ nanoparticles. The reason is that rGO can amend the connections among SnO₂ nanoparticles and present sufficient electronic transmission channels which is favorable for electron transfer [12]. For the sake of further exploring the lithium-ion transfer mechanism, linear fitting of Z’ and ω^-1/2 at low frequency were carried out in Figure 5b. Lithium-ion diffusion coefficient is a momentous parameter for evaluating the performance of electrode materials. The following equation is applied to calculate the diffusion coefficients (D_Li) of lithium ions (Eq. (3)).where, R represents gas constant, T stands for Kelvin temperature, A represents the area of electrode, n is the number of electrons transferred during the reaction. F is Faraday constant, σ is the slope of Z’~ω^-1/2 and C is lithium-ion phase concentration [35]. From the date in Table 1, we can find that the D_Li of SnO₂/rGO electrode is much larger than that of SnO₂. It illustrates that the structure of SnO₂/rGO composite is conducive to the electrolyte diffusion and facilitates Li⁺ migration during lithiation/delithiation process.

The lithium storage kinetics of SnO₂/rGO anode were researched by contrasting CV profiles at different scan rates. By estimating the pseudocapacitive contribution, the high rates performance of SnO₂/rGO anode can be investigated. Meanwhile the influence in complexing rGO in SnO₂ would be further discussed. Figure 6a is the CV profiles of SnO₂/rGO composite under the larger current range as scanning rate raises from 0.1 mV·s^-1 to 2 mV·s^-1, the capacity contribution mechanism can be expressed by Eq. (4).where, b value could be acquired by drawing log(i)–log(v) curves and fitting the line to obtain the slope (Eq. 5). Generally, when b = 0.5, the capacity contribution is mainly determined by diffusion process, if b = 1, it represents to the typical capacitive charge storage for the surface faradaic redox reaction [36]. The b value obtained from the fitting line of cathodic and anodic peaks are 0.85 and 0.80 for SnO₂/rGO composite in Figure 6b. It indicates that the capacitance contribution of capacitor driving in electrochemical behavior is larger than that of diffusion process for SnO₂/rGO anode. The b value of SnO₂ anode is much closer to 0.5, suggesting that the capacitive contribution is mainly acquired by diffusion process for SnO₂ anode. The significantly higher b value of SnO₂/rGO anode indicates that much larger specific surface area obtained by adding three-dimensional rGO is the main reason for the obvious increase of capacitive contribution. For the sake of investigating the capacity contribution mechanism, total current (I) is separated into two parts by Eqs. (6) and  (7).

where, k₁v and k₂v^1/2 represents the current attributed to conversion behavior and diffusion-controlled reaction respectively [37]. As shown in Figure 6c and d, when the scan rate comes to 1 mV·s^-1, the capacitive contribution of SnO₂/rGO anode is 71%, which is significantly larger than that of SnO₂ anode (56%). This result shows that among two lithium storage mechanisms in SnO₂/rGO anode, the capacitive drive process takes an absolute dominant position owing to large specific surface area [38]. The capacitive contribution gets bigger and bigger in the wake of scan rate enlargement as shown in Figure 6d, it is in accord with the charge/discharge profiles and rate performance mentioned before. This result indicates that superior capacitive contribution with large current densities is in favor of clipping Li⁺ insertion-extraction reaction [36]. Previous discussions prove that the lithium storage behavior of SnO₂/rGO anode is mainly driven by capacitive behaviors. Meanwhile it is beneficial to cycling stability and rate performance [21]. Besides, it certifies that the structure of rGO could inhibit the aggregation of SnO₂ nanoparticles, which is consistent with the result of SEM.

The charge-discharge mechanism diagram is shown in the Figure 7, and the previous experiments have obtained good electrochemical performance for the following reasons. The addition of rGO provides more loading sites for SnO₂ and can effectively prevent SnO₂ from falling off during the charging and discharging process. rGO has good electrical conductivity, which is conducive to charge transfer. Besides, rGO itself also has the function of storing lithium ions.