Introduction

The relentless growth of modern technology is constantly feeding society’s increasing need for energy usage and consumption. However, the development of energy storage devices has not been able to keep pace with the rapid progress of portable electronics, large-scale usage such as Electric Vehicles (EVs), plug-in hybrid Electric Vehicles (PHEV) and even grid-scale storage1,2,3. Over the course of developing an effective anode material, tin-based anode materials has been recognized as one of the most promising materials to replace the current generation of commercial graphite anode4,5,6. This is due to its salient features such as high theoretical capacity, high energy density owing to low voltage discharge profile and low cost due to its high abundance in nature6,7. Amongst the different tin-based anode materials, SnS2 offers a two-dimensional layered type structure where the CdI2-type layers are loosely bounded by weak van der Waals forces and hence easily susceptible to the intercalation of lithium-ions (LixSnS2)8,9. Further interaction with lithium-ions in a discharge process results in the conversion process of the lithiated SnS2 to form metallic Sn and Li2S8,10. In subsequent charge-discharge cycles, the liberated tin alloys reversibly with lithium forming Li4.4Sn. The above reactions can be described using the following equations:

Simplifying equations (1) & (2),

In subsequent charge-discharge cycles,

The reversible reaction of Sn in SnS2 with 4.4 mols of lithium (equation (4)) translates into a high theoretical capacity of 645 mAh g−1 (compared to graphite: 372 mAh g−1). Despite this, it can be observed that 4 mols of lithium is spent in the irreversible formation of Li2S during the conversion reaction of lithiated SnS2 (equation 2). Hence, the theoretical capacity of SnS2 can potentially be as high as 1231 mAh g−1 (8.4 mol Li+ per mol SnS2) compared to 645 mAh g−1 (4.4 mol Li+ per mol SnS2) if the irreversible reaction in equation 2 is made reversible.

To date, there are several reports of SnS2 with beyond theoretical capacities of 645 mAh g−1 11,12,13,14,15,16,17,18,19,20,21,22,23. However, the origin of these excess capacities are often unclear and attributed to the effects of nano-sized SnS2 and possible synergistic effects with carbonaceous materials such as amorphous carbon11,12, carbon nanotubes13,24, and reduced graphene oxides14,15,16,17,18,19,20,25. The synthesis and formation of nanostructures and nanocomposites is a common strategy to overcome large volume changes that SnS2 experience during lithiation and delithiation that plagues the stability and cyclability of the active material. The carbonaceous materials within the nanocomposite does not only provide volume change buffers for SnS2, but also serves to improve electron conduction. These strategies are also common to transition metal oxides and sulfides based anode materials26,27,28,29,30. However, efforts have been limited to the above strategies and little has been done to properly utilize both conversion (equation 2) and alloying (equation 4) process of SnS2, which can potentially improve capacities significantly.

Herein, we present the successful integration of a small amount of Molybdenum (Mo) nanoparticles catalyst, directly onto SnS2 nanosheets, to promote the active reversible conversion process of Sn to SnS2 and decomposition of Li2S. Mo catalyst, grown onto SnS2 in the form of MoS2, leads to a remarkable enhancement in the performance of the lithium ion batteries. Furthermore, SnS2 nanosheets are grown onto 3-dimensional graphene foam (3DG) to improve conductivity and suppress volume changes during charge and discharge cycles. The as-synthesized SnS2/MoS2/3DG are directly used as binder-free and lightweight electrodes where it exhibits enhanced electrochemical performance in terms of improved specific capacity, high rate capability and cycling stability. Direct evidence of the catalytic effect of Mo is also observed via ex situ transmission electron microscopy (TEM) and X-ray photonspectroscopy (XPS) analysis. This catalytic process is able to overcome the irreversible reaction in the tin-based anode materials and has important implications for future applications for such materials in various energy storage systems.

Results and Discussion

Characterization of Materials

The step-by-step synthesis process and morphology of SnS2/MoS2/3DG nanostructure is detailed in the illustration and the corresponding SEM images in Fig. 1. 3DG was prepared by CVD using ethanol as carbon source and Ni foam as a substrate. The Ni foam substrate was subsequently etched away using 3 M HCl leaving a lightweight 3DG current collector (~0.5 mg cm−2) (Fig. 1b). SnS2 nanosheets (~1.2 mg cm−2) were uniformly grown onto 3DG through a facile solvothermal synthesis process where surfactant (SDS), was used to reduce the size of the nanosheet (Fig. 1c)31,32. This creates a denser loading of SnS2 on 3DG (Figure S1, see Supporting Information). In the second solvothermal synthesis step, trace amounts of MoS2 precursors were used to control the limited growth of MoS2 catalyst on SnS2. This was carried out by placing pieces of the as-synthesized SnS2/3DG in the Teflon-lined autoclave with the dissolved MoS2 precursors. As shown in Fig. 1d, sparse amount of MoS2 nanosheets were grown directly onto the surfaces of SnS2 nanosheets. Instead of directly introduction Mo as a catalyst, MoS2 nanosheets were grown on SnS2 to exploit the 2D nature of MoS2 with large exposed surfaces that will facilitate rapid catalytic action during charge-discharge cycles. Lastly, ethanol solvent was used in both solvothermal syntheses as 3DG is hydrophobic. The final product, SnS2/MoS2/3DG, can be used directly as a lightweight binder-free anode material without needing additional conductive carbon. Figure S2 shows a photograph of the as-synthesized products where the stepwise growth of the SnS2 and MoS2 results in a color change of the electrode, green (SnS2) then black (MoS2). In order to obtain a more detailed morphology of MoS2 growth on SnS2, TEM characterization was carried out as shown in Fig. 2a,b. From the TEM image in Fig. 2a, fine sheets of MoS2 can be clearly observed on both sides of the larger SnS2 nanosheets. This sandwich structure is clearly marked out as shown in the inset of Fig. 2a. In the high resolution TEM image in Fig. 2b, the interplanar spacing of 6.0 Å and 5.8 Å can be assigned to the (002) and (001) crystal planes of MoS2 and SnS2 respectively8,28.

Figure 1
figure 1

(a) Illustration of the growth process of SnS2/MoS2/3DG via stepwise solvothermal synthesis. In increasing high and low (inset) magnification SEM images of (b) pristine etched 3DG, (c) SnS2/3DG, and (d) SnS2/MoS2/3DG.

Full size image
Figure 2
figure 2

HRTEM images of SnS2/MoS2/3DG (a) low magnification with inset showing sandwich MoS2/SnS2/MoS2 and (b) high magnification showing lattices of SnS2 and MoS2. (c) XRD pattern of SnS2/MoS2/3DG and SnS2/3DG. (d) Raman spectra of SnS2/MoS2/3DG, SnS2/3DG, and as-prepared 3DG.

Full size image

The crystalline phases and structures of the as-synthesized SnS2/3DG and the final product SnS2/MoS2/3DG were confirmed via X-ray diffraction (XRD) as shown in Fig. 2c. The major peaks at 15.0°, 28.5°, and 32.2° are respectively indexed to the (001), (100), and (101) of SnS2 (JCPDS No. 23-0677)8,33, while the intense peak at 26.8° arises from the (002) crystal plane of graphitic carbon (JCPDS No. 75-1621)29. On the other hand, the new peaks at 34.3°, 38.0°, and 51.9° of the SnS2/MoS2/3DG can be assigned MoS2 (JCPDS No. 65-3656)34. Raman spectroscopy was carried out to further identify the as-synthesized materials (Fig. 2d). The peak at 314 cm−1 is attributed to the A1g mode of SnS235, while the peaks at 378 and 405 cm−1 arises from the in-plane E12g and out-of-plane A1g modes of MoS226,36. In addition, the peaks at 1573 and 2718 cm−1 corresponds characteristic G and 2D bands of the 3DG37,38. Furthermore, the absence of the D band is indicative of the high quality and low defect of 3DG produced via CVD method39. Thermogravimetric analysis (TGA) was carried out, at a temperature range of 25 °C to 900 °C in dry air at a rate of 5 °C min−1, to quantify the composition and verify the loading of MoS2 on SnS2/MoS2/3DG nanocomposite. From Figure S3, the initial weight loss of ~5% arises from the moisture loss from the surfaces of the samples. The second step between 200 to 500 °C is attributed to the oxidation of SnS2 and MoS228,40. As SnS2 and MoS2 oxidize at roughly the same temperature range, mass loading of MoS2 in SnS2/MoS2/3DG can be deduced by the difference in weight loss (~3%) between the two samples (SnS2/3DG vs. SnS2/MoS2/3DG). Finally, the last weight loss step between 600 to 850 °C corresponds to the combustion of 3DG in air29. Therefore, based on TGA, the composition of SnS2/MoS2/3DG and be deduced to be SnS2: 0.65, MoS2: 0.03, 3DG: 0.32 [See supporting information for detailed derivation].

The chemical state and elemental distribution of SnS2/MoS2/3DG was also confirmed by X-ray photoelectron spectroscopy (XPS) and elemental dispersive X-ray (EDX) analysis, respectively. Figure 3a shows the broad spectrum of the XPS scan on SnS2/MoS2/3DG where the presence of elements Sn, S, Mo, and C are clearly observed. The high resolution XPS spectra of these elements are shown in Figs 3b,c,d, and S4, respectively. The deconvolution of these high resolution spectra produced pairs of peaks at 495.2 and 486.8 eV (Fig. 3b), 162.5 and 161.1 eV (Fig. 3c), and 232.2 and 229.0 eV (Fig. 3d), corresponding to the Sn 3d3/2 and Sn 3d5/2 states in SnS2, S 2p1/2 and S 2p3/2 in MoS2 and SnS2, as well as Mo 3d3/2 and Mo 3d5/2 in MoS2, respectively22,41. The results indicate the oxidation states of Sn and Mo to be +4, which are in good agreement with the phase pure results of SnS2/MoS2/3DG obtained via XRD. The deconvuluted C 1 s spectra in Figure S4 shows the presence of several oxygen functional groups on the surfaces of 3DG such as C-O/C-O-C at 285.2 eV as well as C=O/O-C=O at 288.5 eV42. In addition, EDX analysis in Fig. 3e shows homogenously distributed elements, Sn, Mo, S, and C, which implies the uniform growth of SnS2/MoS2 on the surfaces of 3DG.

Figure 3
figure 3

(a) XPS survey spectrum, (b) Sn 3d spectra, (c) S 2p spectra, and (d) Mo 3d spectra of SnS2/MoS2/3DG. (e) SEM image of SnS2/MoS2/3DG corresponding to the EDX elemental mapping images of Sn, S, Mo, and C showing uniform distribution of the elements.

Full size image

Electrochemical Measurements

The catalytic effect of Mo in SnS2/MoS2/3DG was verified via electrochemical characterization through cyclic voltammetry (CV) and galvanostatic charge-discharge cycling. These experiments were carried out through the assembly of a half-cell battery with SnS2/MoS2/3DG as a binder-free electrode and lithium foil as a counter electrode. Control sample, SnS2/3DG, was tested in the same conditions. Figure 4a shows the CV curves of the SnS2/MoS2/3DG electrodes. In the initial discharge cycle, the cathodic small broad peak centered at around 1.5 V can be attributed to the initial Li insertion into SnS2 and MoS2, without phase transformation, forming LixSnS2 and LixMoS2 9,43. Upon further discharge, peaks at ~0.8 V and ~0.45 V can be ascribed to the conversion reaction of lithiated SnS2 and MoS2 to metallic Sn and Mo, respectively16,43. Towards the end of the discharge cycle, the broad peak from ~0.1–0.3 V corresponds to the alloying reaction between Sn and Li followed by the insertion of Li into 3DG, as well as the formation of the solid-electrolyte interface (SEI) film between the active materials and electrolyte. The formation of the SEI film contributes significantly to the initial loss in discharge capacity9,29,44. In the following charge-discharge cycles, the overlapping redox peaks can be attributed to the highly reversible lithium insertion/extraction, conversion and alloying reactions. The first redox pairs at ~0.1/0.25 V and ~1.2/1.3 V arise from the lithium insertion and extraction in 3DG and MoS2, respectively28,29. Next, the alloying and de-alloying reactions of Sn with Li produces redox peaks centered around 0.25 V and 0.56 V, respectively33. Lastly, the redox pairs at ~1.16/1.86 V and ~1.8/2.26 V correspond to the conversion reactions occurring for both Sn and Mo, respectively14,28. In more details, the anodic peaks represent the reduction of SnS2 and MoS2 to metallic Sn and Mo as well as formation of Li2S. The cathodic peaks represent the reformation of SnS2 and MoS2 with the decomposition of Li2S. In the absence of the catalyst, Mo, the CV curves of SnS2/3DG are similar to that of SnS2/MoS2/3DG with the exception of the peaks arising MoS2, as well as a gradually decaying redox pair ~1.32/1.87 V (Figure S5). The observed weakened redox pair indicates that the conversion reaction is only partially reversible in SnS2/3DG. In contrast, the strong overlapping of the Sn conversion reaction redox pairs for SnS2/MoS2/3DG implies that in the presence of Mo catalyst, the conversion process of Sn becomes highly reversible14. The reactions can be described using the following equations9,28:

Figure 4
figure 4

(a) Cyclic voltammograms of SnS2/MoS2/3DG electrode at a rate of 0.05 mV s−1 in a potential range of 0.01 to 3.0 V. (b) Galvanostatic discharge and charge curves for SnS2/MoS2/3DG and SnS2/3DG electrodes at the 1st cycle at a current density of 50 mA g−1 in the potential range of 0.01 to 3.0 V showing the initial Coulombic Efficiency. (c) Rate capability of SnS2/MoS2/3DG and SnS2/3DG electrodes. (d) Cycling performance of SnS2/MoS2/3DG and SnS2/3DG electrodes at current density of 200 mA g−1, and the corresponding coulombic efficiency of SnS2/MoS2/3DG electrode.

Full size image

Reactions involving SnS 2

In the first discharge,

Simplifying equations (5) & (6),

In subsequent charge-discharge cycles,

Reactions involving MoS 2

Where x and y are the number of moles of Li+/e− involved in the insertion reactions with SnS2 and MoS2 respectively.

Figure 4b shows the initial charge-discharge profile of both SnS2/MoS2/3DG and SnS2/3DG. Similar to the CV curves, both electrodes display similar charge-discharge profile with differences owing to the presence of MoS2 in SnS2/MoS2/3DG. The first discharge begins with a brief slope at ~1.8 V followed by a short plateau at ~1.3 V which is attributed to the insertion of Li ion (equations 5 & 9) into and subsequent conversion of SnS2 and MoS2 (equations 6 & 10), respectively9,43. Thereafter, the curve proceeds with a gentle slope towards 0.05 V followed by small plateau. The sloped region corresponds to the alloying process of Li with Sn (equation 8) as well as the formation of the SEI layer which results in an irrecoverable loss of lithium9,44. The small plateau at 0.05 V arises from the intercalation of Li ions to 3DG29. In the charge cycle, delithiation proceeds with extraction of Li ions from 3DG at ~0.15 V followed by de-alloying of Li4.4Sn at ~0.5 V. Upon charging to higher voltage (1.0–3.0 V), LixMoS2 delithiates (equation 9), Sn oxidizes to SnS2 while Li2S decomposes to S (backward reaction in equation 7)14,43. However, in the case of SnS2/3DG, the oxidation of Sn and decomposition of Li2S is limited, which can be observed in the acute slope between 1.9–3.0 V. Therefore, the catalytic effect of MoS2 in SnS2/MoS2/3DG is evident with the reversible decomposition of Li2S and oxidation of Sn to SnS2 (equation 7). This is in stark contrast to previously reported SnS2 electrodes where Li2S formed during the initial discharge is irreversible or at most partially reversible in the first few cycles9. The reversible decomposition of Li2S in the first cycle leads to an improved initial Coulombic Efficiency (ICE). This can be observed between the MoS2 loaded, SnS2/MoS2/3DG with 81.5% CE (first discharge capacity: 1196.5 mA h g−1, first charge capacity: 973.0 mA h g−1) compared to SnS2/3DG with 60.4% CE (first discharge capacity: 1372 mA h g−1, first charge capacity: 829.7 mAh g−1). The higher first discharge capacity in the SnS2/3DG electrode may be due to a larger formation of SEI layer. The effective catalytic conversion reaction of SnS2 may be attributed to the direct growth of MoS2 on SnS2 nanosheets that facilitate rapid reversible conversion process of Sn and decomposition of Li2S.

Apart from improving the ICE, the reversible decomposition of Li2S increases the total number of moles of Li ion reacting with SnS2 from 4.4 (equation 8) to 8.4 (equations 7 and 8) by utilizing both alloying and conversion processes. In this manner, the theoretical capacity of SnS2 increases significantly from 644 mA h g−1 (4.4 mols of Li+) to 1231 mA h g−1. In the case of SnS2/MoS2/3DG, the theoretical capacity of the nanocomposite, taking into the account of MoS2 and 3DG, should be 940 mAh g−1 (CapacitySnS2/MoS2/3DG = 1231 ∗ 0.65 + 670 ∗ 0.03 + 372 ∗ 0.32). This increase is evident even at different current densities as observed in Fig. 4c where the two different samples are subjected to galvanostatic charge-discharge cycles at increasing current densities. SnS2/MoS2/3DG delivered discharge capacities of 960.5, 837.0, 705.1, 626.2, and 495.6 mA h g−1 at current densities of 50, 200, 500, 800, and 1000 mA g−1, respectively. This is distinctly higher as compared to SnS2/3DG with highest reversible discharge capacities of ~800 mA h g−1 at low current densities of 50 mA g−1 and high current discharge capacity of 441.7 mA h g−1 at 1000 mA g−1. Highly reversible and stable cycling was also observed for the MoS2 loaded SnS2/MoS2/3DG at current density of 200 mA g−1 where capacity retention is 91.5% after 50 cycles (Fig. 4d). Furthermore, the first cycle CE is also consistent with that at 50 mA g−1 of 82% that increased and maintain near 100% in subsequent cycles (Fig. 4d). The high capacity, highly reversible and stable cycling implies the effective catalytic action of MoS2. Furthermore, successful reversible decomposition of Li2S at high current densities of 1000 mA g−1 highlights importance of the sandwiched MoS2/SnS2/MoS2 nanostructure.

Electrochemical impedance spectroscopy (EIS) measurements were also conducted for both SnS2/MoS2/3DG and SnS2/3DG to understand and evaluate the influence of MoS2 loading on SnS2/3DG. The measurements were performed on the samples at a semi-charged state (~2.5 V) across a frequency range of 10 mHz to 1 MHz.

Figure 5a shows the Nyquist plots where the curves from both electrodes display similar profile; a small semi-circle in the high frequency region corresponding to a combination of surface film resistance (Rsf) and charge transfer resistance (Rct) followed by an acute straight line in the low frequency region which arises from the Warburg region (Zw) as well as a Li accumulation element (Cint)30,45. An equivalent circuit consisting of Rsf+ct, Zw, Cint, as well as Re, which is the ohmic resistance arising from the electrolyte and cell components, and CPE(sf+dl), a constant phase element relating to the surface film and double layer, was used to fit the Nyquist plots (inset in Fig. 5a). The fitted values using the equivalent circuit are presented in Table S1. The lower surface film and charge transfer resistance (Rsf+ct) of SnS2/MoS2/3DG (33.65 Ω) compared to SnS2/3DG (64.67 Ω) is an indication that MoS2 loading improves charge kinetics. This improvement may be attributed to the release of conductive metallic Mo, through the conversion process of MoS2 in the SnS2/MoS2/3DG electrode, over charge-discharge cycle. The effects of the enhanced conductivity via MoS2 loading in SnS2/MoS2/3DG can be observed by evaluating the electrode at high current densities where charge kinetics are more demanding. Figure 5b shows the discharge capacities of SnS2/MoS2/3DG and SnS2/3DG at high current densities of 1000 mA g−1 for 50 cycles where the capacity retention was 85.2%. Owing to the enhanced conductivity and charge kinetics, rapid catalytic action of Mo is achievable. This is reflected by the substantially increased capacity of the SnS2/MoS2/3DG (647 mA h g−1) as compared to SnS2/3DG (355.1 mA h g−1) (Fig. 5b). Another key contributing factor to the effective catalytic action of Mo at different current densities is the well-designed SnS2/MoS2/3DG hierarchical nanostructure as illustrated in Fig. 5c. The sandwich MoS2/SnS2/MoS2 nanostructure allows the released metallic Mo to be in direct proximity of Li2S formed during lithiation. In this manner, Mo is able to rapidly catalyze the decomposition of Li2S during delithiation.

Figure 5
figure 5

(a) Electrochemical impedance spectra of SnS2/MoS2/3DG and SnS2/3DG electrodes as well as the corresponding fitting for both electrodes. Inset: Equivalent circuit used for curve fitting. (b) Cycling performance of SnS2/MoS2/3DG and SnS2/3DG electrodes at current density of 1000 mA g−1. (c) Illustration of MoS2/SnS2/MoS2 sandwich structure which is able to efficiently promote catalytic action.

Full size image

Further evidence towards catalytic effect of Mo

In order to further demonstrate the catalytic effect of Mo, control samples of SnS2 and SnS2/MoS2 were synthesized in the same method as SnS2/3DG and SnS2/MoS2/3DG in the absence of 3DG, respectively. Without the 3DG current collectors to act as a platform for ordered growth, the as-synthesized SnS2 nanosheets aggregated into micron sized clusters, which resembled a fully bloomed flower (Figure S6a). When MoS2 was introduced, the nanosheets of MoS2 was also able to grow on the surfaces of SnS2 flowers (Figure S6b). Given the lack of a growth platform, the as-synthesized control samples were in the form of powders. Hence, to obtain an electrode, these powders were made into a slurry form that were subsequently coated on Ni-foam current collectors and finally assembled into coin cells. Similarly, CV and galvanostatic cycling were performed on SnS2 and SnS2/MoS2 as shown in Figure S6c–f. The CV peaks of SnS2 and SnS2/MoS2 are consistent with SnS2/3DG and SnS2/MoS2/3DG, respectively, with the exception of peaks arising from 3DG. In terms of galvanostatic cycling, the catalytic effect could also be observed in SnS2/MoS2 where there is a significant improved of ICE from 55% (SnS2) to 67% (SnS2/MoS2) (Figure S6e) as well as improved discharge capacity from 6745 mAh g−1 (SnS2) to 806 mAh g−1 (SnS2/MoS2) (Figure S6f). However, at higher current densities (500, 800, and 1000 mA g−1), both electrodes delivered roughly the same discharge capacities. This could be due to the instability of the SnS2/MoS2 structure that resulted in the gradual dissociation of MoS2 nanosheets from SnS2 surfaces because of significant volume changes of SnS2 after prolonged cycling. Therefore, these results further supports the presence of the catalytic action of Mo towards the decomposition of Li2S and oxidation of Sn to SnS2 which results in a substantial improvement towards both the ICE and capacities of SnS2 anodes. Furthermore, despite the improvement, the rapid decay in capacities of SnS2/MoS2 highlights the importance of a rational design towards a well-structured composite that is demonstrated in SnS2/MoS2/3DG.

Thus far, there has already been numerous reports citing high discharge capacities in SnS2-based anode beyond its 644 mAh g−1 theoretical capacity14,15,16,17,18,19,20,21. In most of these cases, SnS2 was synthesized as a nanocomposite together with reduced graphene oxide (rGO). For example, Lin et al. reported the synthesis of ultrasmall SnS2 on rGO with high capacities of 1034 mA h g−1 at current density of 64.5 mA g−1 while few-layer SnS2/graphene presented by Chang et al. delivered capacities of 920 mA h g−1 16,46. However, the contributing factors towards the excess capacity are loosely assigned to a synergistic effect between SnS2 and rGO without clear clarification. In a recent report, Qu et al. attributed the high capacities of the SnS2/Graphene nanocomposite to a reversible conversion of SnS2 from Sn through a catalytic effect brought about by Sn nanoparticles, which were released during conversion reactions of SnS2, stacked within graphene matrices14. Unfortunately, there was no physical evidence provided to support this postulation. Furthermore, due to the presence of defects, surface functional groups and other additional Li storage sites, the capacity of rGO varies depending on the quality of the graphene sheets and can be as high as ~1000 mA h g−1 47. As a result, the capacity contribution by rGO in the SnS2/rGO nanocomposites may be more significant than reported.

In an analogous system, the reversible decomposition and formation of Li2O from SnO2 and GeO2 has been shown to be mediated in the presence of nano-sized catalyst, Co and Ge respectively. This led to capacities beyond their theoretical capacities (SnO2/Co3O4/rGO: 1038 mAh g−1 vs SnO2: 782 mAh g−1 and GeO2/Ge/C: 1860 mAh g−1 vs GeO2: 1126 mAh g−1)48,49,50,51,52. The enhanced capacities were attributed to the catalytic role of the respective dopants which facilitated the reversible formation and decomposition of Li2O. Another consequence of this catalytic effect is the improvement of initial Coulombic efficiencies (ICE) as a significant portion of the initial capacity is typically lost due to the formation of irreversible Li2O apart from the formation of surface electrolyte interface (SEI). Therefore, since SnS2 is analogous to SnO2, the effect of the catalyst can be extended to SnS2 based anodes.

The reversible oxidation/reduction of SnS2/Sn and Li/Li2S as described in equation 7 occurs during the discharge/charge cycles, respectively. Hence, evidence of the reversibility of equation 7 can be investigated through the detection of the products at the end of discharge (0.01 V) and charge (3.0 V) process. In this regard, post cycling TEM and XPS was conducted for SnS2/MoS2/3DG electrodes after 20 cycles at current density of 200 mA g−1. Being free from additives such as binder and active carbon, the cycled electrodes were free from stray and unwanted signals in the post cycling characterizations. Two sets of coin cells were cycled with one set ending at full discharge state (0.01 V) and the other at full charge state (3.0 V). The electrodes were extracted from the coin cell in the Ar-filled glove box, soaked overnight with acetonitrile and washed with ethanol several times. The ex-situ HRTEM of the fully discharged and charged electrodes are shown in Fig. 6a,b, respectively. Upon full discharge (lithiation), crystalline lattices corresponding to metallic Mo and Sn can be found (Fig. 6a). This corresponds to the conversion process of MoS2 and SnS2 to their metallic phases as well as the incomplete lithium alloying of Sn51,53,54. On the other hand, lithiated Sn alloy (LiXSn) which are often reported as nanoparticles embedded within the Sn-based electrodes, may be identified as the dark nanoparticles (~3–5 nm) circled out in Figs 6a and S755,56,57. However, crystal lattice of the LixSn nanoparticles could not be clearly observed as they are embedded within the overlapping lattices from Sn and Mo in the complicated SnS2/MoS2/3DG structure. Li2S was not identified due to its amorphous nature at full lithiated state58. In the fully charged state (delithiation), the presence of crystalline lattices which can be assigned to MoS2 and SnS2, implies a reversible conversion process of metallic Mo and Sn as well as LixSn to their intial state (Fig. 6b). Therefore, Sn formed in the discharge state and SnS2 charged state corresponds well with reversibility of equation 7 where the forward reaction corresponds to discharge (formation of Sn) and backward reaction the charge process (formation of SnS2). As an additional evidence towards the reversible conversion of SnS2, ex-situ XPS was also conducted on the cycled electrodes at full discharge and charge states. Figure 6c compares the high-resolution XPS spectra of SnS2/MoS2/3DG electrodes at different states, pristine, discharged to 0.01 V, and charged to 3.0 V. The pristine electrode corresponds to the initial state of SnS2/MoS2/3DG where the XPS spectra peaks corresponds to Sn 3d5/2 and Sn 3d3/2 as discussed in Fig. 3b. Upon discharge, these peaks shift downwards to 493.1 eV and 484.7 eV corresponding the reduction of Sn4+ of SnS2 to metallic tin (Sn0)59. In the charge state, the shifted peaks returns to the initial positions of 495.1 eV and 486.6 eV indicating an oxidation of Sn0 to Sn4+, i.e. the reformation of SnS2 60. Once again, the presence of Sn at discharge states and SnS2 at charged states points towards the reversibility of the reaction in equation 7. Therefore, the results obtained from the ex-situ HRTEM and XPS analyses of the post cycled electrodes at charged and discharge states provides conclusive evidence that the reversible conversion reactions of SnS2 (equation 7) are promoted through the catalytic action of transition metal, Mo, in the active material matrix.

Figure 6: Ex situ analysis of post cycled SnS2/MoS2/3DG electrode.
figure 6

HRTEM image of (a) discharge (0.01 V) state and (b) charge (3.0 V) state. (c) Sn 3d spectra of the electrode in pristine, discharge, and charge state.

Full size image

Conclusion

In summary, a rationally designed hierarchical nanostructured SnS2/MoS2/3DG nanocomposite was synthesized via stepwise solovothermal synthesis of metal sulfides on a lightweight current collector substrate, 3DG. The stepwise synthesis facilitated the growth of self-assembled vertical SnS2 nanosheets on surfaces of 3DG followed by growth of ultrafine MoS2 nanosheets which branches out on both sides of SnS2, creating a sandwich MoS2/SnS2/MoS2 structure. The sandwich structure places MoS2 in very close proximity to SnS2 that facilitates rapid catalytic action of Mo towards decomposition of Li2S and conversion of SnS2. As a catalyst towards the reversible conversion reaction of SnS2, the loading of MoS2 on SnS2/MoS2/3DG was controlled at a low content of 3%. A small amount of Mo proved sufficient towards the catalytic effect reversible conversion of SnS2 that saw a significantly improved ICE as well as increased discharge capacities even up to high current density of 1000 mA g−1. The binder-free SnS2/MoS2/3DG lightweight electrode also facilitated post cycling investigations. Ex situ HRTEM and XPS analysis of charged and discharged electrodes provided substantial evidence towards the catalytic effect of Mo in the reversible conversion of SnS2; metallic Sn and Mo could be found in discharged electrodes while SnS2 and MoS2 identified in charged electrodes. Therefore, the use of catalyst to unlock the potential of SnS2 as an anode that exploits both conversion and alloying reactions creates a new direction towards high capacity metal sulfide anode materials.

Experimental Section

SnS2/MoS2/3DG was prepared by 3 general steps 1, chemical vapor deposition (CVD) to obtain 3DG 2, solvothermal to obtain SnS2 on 3DG (SnS2/3DG), and 3 solvothermal to obtain MoS2 on SnS2/3DG (SnS2/MoS2/3DG).

Preparation of 3DG and removal of Ni Foam

The preparation of the 3DG foam adopts a similar method as previously reported29,33,61. Ni foam (1.6 mm thick, purchased from Alantum Advanced Technology Materials (Shenyang)) was cut into 150 mm × 45 mm pieces. The rectangular Ni foam pieces are then rolled and placed into a 1 inch quartz tube. Before heat was applied, Argon (Ar) gas was allowed to flow for 10 mins to remove residual air in the tube. Thereafter, a heat rate of 50 °C min−1 was applied to heat the quartz tube to 1000 °C. At this temperature, ethanol vapour was mixed with the flowing N2/H2 gas through the bubbling of anhydrous ethanol. The mix N2/H2/ethanol vapour was allowed to flow for 5 mins. Lastly, the quartz tube was allowed to cool to room temperature rapidly (~100 °C/min). The Ni foam on 3DG-Ni was removed in an etchant solution of 3 M HCl at 80 °C. Pure 3DG was washed using deionized (DI) water and ethanol several times before drying at 60 °C. The typical loading of this process yields ~0.5 mg cm−2 of 3DG on Ni foam (3DG-Ni).

Synthesis of SnS2/3DG

The synthesis of SnS2/3DG follows the previously reported method with the addition of SDS33. SnS2 grown on 3DG (SnS2/3DG) was prepared via a simple solvothermal reaction. SnCl4.5H2O (32 mM), thioacetamide (TAA) (80 mM), and Sodium dodecyl sulfate (SDS) (0.07 mM) was weighed and dissolved into 35 mL ethanol. Magnetic stirring and mild heat (50 °C) was applied to ensure homogeneity of the precursors. 6 pieces of 3DG (12 mm diameter) was cut and together with the mixture, transferred into a 50 mL Teflon-line stainless steel autoclave. The reaction proceeded by heating the autoclave at 180 °C for 12 h. After the reaction, as-synthesized SnS2/3DG was collected by rinsing with DI H2O and ethanol, and drying at 60 °C. The loading of SnS2 on 3DG was ~1.2 mg. Control sample, pristine SnS2, was synthesized by the same method in the absence of 3DG.

Synthesis of SnS2/MoS2/3DG

SnS2/3DG synthesized in the previous step was used directly in this step after drying at 60 °C. MoS2 grown on SnS2/3DG, termed as SnS2/MoS2/3DG, was prepared by an L-cysteine assisted solvothermal reaction. L-cysteine (0.25 mM) and Na2MoO4·2 H2O (0.03 mM) was weighed and dissolved in 15 mL DI H2O. 15 mL of ethanol was added to the continuously stirred mixture. 6 pieces of SnS2/3DG was added to the mixture and transferred to a 50 mL Teflon-line stainless steel autoclave. The stainless steel autoclave was heated at 180 °C for 12 h. The as-synthesized SnS2/MoS2/3DG was collected by rinsing with DI H2O and ethanol, and drying at 60 °C. The loading of MoS2 on SnS2/3DG was between 0.05–0.1 mg. Lastly, the SnS2/MoS2/3DG was annealed in Nitrogen (N2) environment at 400 °C for 2 hours at a slow heating rate of 3 °C min−1. Control sample, SnS2/MoS2, was synthesized following the above method using pristine SnS2 in place of SnS2/3DG.

Materials Characterization

The nanostructures and morphology of the as-synthesized samples were imaged under a field-emission scanning electron microscope (FESEM, JSM-7600) and transmission electron microscope (TEM, JEM-2100F). Crystalline phase and structure of the samples were identified by X-ray diffraction (XRD, Bruker D8). Raman spectroscopy was conducted by a confocal Raman setup with a 532 nm lazer excitation (WITec Instruments Corp, Germany). The nanocomposite content breakdown was measured by thermogravimetric analysis (TGA, Shimadzu, DTG-60). The chemical valence states of the samples were investigated with a X-ray photoelectron spectroscope (XPS, PHI Quantera II, Physical Electronics, Adivision of ULCAV-PHI).

Coin cell assembly and Electrochemical measurements

The as-synthesized SnS2/MoS2/3DG was used directly as an anode in the two electrode half-cell configuration where lithium metal acts as the counter electrode (cathode). Control samples of SnS2/3DG, pristine SnS2, and SnS2/MoS2 were also assembled. Pristine SnS2 was synthesized using the procedure described in step 2, in the absence of 3DG whereas SnS2/MoS2 was synthesized using the procedure described in step 3 by using pristine SnS2 instead of SnS2/3DG. Electrodes for the powdered control samples, SnS2 and SnS2/MoS2 were made into a slurry and subsequently coated on Ni foam current collectors (12 mm diameter each). The slurry was prepared by mixing 8 parts of active materials to 1 part of polyvinylidene fluoride (PVDF) binder and 1 part of carbon black in the presence of N-Methylpyrrolidone (NMP) solvent. The SnS2/3DG was used directly as an electrode. All working electrodes were dried at 120 °C for 12 h before assembly. The coin cell were assembled in an Ar-filled glove box using standard CR 2032 as casing. Each cell consisted of a Celgard 2400 membrane sandwiched between a working electrode (active material) and counter electrode (lithium metal). Electrolyte used was a mixture of 1 M LiPF6 solution in Ethylene Carbonate (EC)/Di-Methyl Carbonate (DMC), 1:1 v/v. Electrochemical measurements, cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) were performed using an electrochemical workstation (VMP3, Biologic, France). While galvanostatic charge-discharge cycles were carried out using a battery analyser (Neware, China).

Additional Information

How to cite this article: Huang, Z. X. et al. Unlocking the potential of SnS2: Transition metal catalyzed utilization of reversible conversion and alloying reactions. Sci. Rep. 7, 41015; doi: 10.1038/srep41015 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.