Tungsten diselenide (WSe₂) as a high capacity, low overpotential conversion electrode for sodium ion batteries

Abstract

Tungsten diselenide (WSe₂) is demonstrated as an efficient electrode for sodium ion batteries for the first time. A high reversible capacity above 200 mA h g⁻¹ is observed at 20 mA g⁻¹ rate, with over 250 mA h g⁻¹ capacity measured in the first sodium extraction. Assessment of electrolyte and binder materials on performance was examined and an EC/DEC electrolyte with CMC binder emerges to yield the highest capacity and cycling retention. Comparison between WS₂ and WSe₂ distinguishes WSe₂ to exhibit superior performance due to more efficient energetics bearing a small overpotential <0.30 V. Ex situ analysis and imaging after cycling confirms a sodium-mediated conversion reaction that yields isolated domains of W metal or Na_xSe and reformation of WSe₂ upon sodium extraction, enabling insight into the chemical storage pathway. This work highlights the promise of WSe₂ compared to other conversion-based transition metal dichalcogenides as a practical material for sodium ion batteries.

---

Lithium ion batteries (LIBs) have become the prominent energy storage technology in modern applications such as electric vehicles and portable electronics.¹ As battery demand increases, alternative earth-abundant battery materials are required to achieve progress both in performance and cost. Sodium ion batteries (NIBs) are a viable alternative to lithium ion batteries due to the extreme abundance, low cost, and similar reactivity of Na compared to Li.² In many cases, this enables direct application of knowledge gained from LIB research to be readily applied in NIB systems since electrode materials, such as transition metal oxides, have Na equivalents.^3,4 The main drawback with NIBs is the larger size of Na⁺ (1.02 Å) compared to Li⁺ (0.76 Å). This poses a challenge in designing materials to exhibit both high storage capacity and appropriate cycling performance. As an example, graphite – which is the benchmark anode material for LIBs, stores over 10× less Na ions per carbon atom than Li.^4,5 This underlines the need for the exploration of new electrode materials for NIBs that can be competitive in performance with commercially available lithium-ion cells.

The transition metal dichalcogenides (TMDs) are a family of materials that have recently gained significant attention for their ability to store metal ions such as Li, Na, and Mg.^6–8 TMDs (MX₂ where M = Mo, W and X = S, Se, Te) have a lamellar structure (space group P6₃/mmc) similar to graphite but with larger interlayer spacing more appropriate for Na⁺ intercalation. Of all TMDs, MoS₂ has been the most widely studied for sodium ion batteries,^9,10 where storage has been identified to occur through both a high voltage intercalation reaction (products: Na_xMoS₂ (x < 0.5)) and a low voltage conversion reaction (products: Na₂S + Mo),⁷ the former of which is highly reversible.¹¹ The intercalation reaction shows a 2H (semiconductor) to 1T (metallic) transition while the conversion reaction produces amorphous Mo and a sodium/sulfide species.¹²

Research in TMD-based electrodes for NIBs outside of MoS₂ is limited despite research in other energy-related applications correlating TMD composition to catalytic or chemical performance,¹² due to physical reactivity tuned by the metal to chalcogen composition.¹³ For NIB batteries, MoSe₂ yolk shell microspheres and MoSe₂ nanoplates have shown high capacities and good cyclability,^14,15 and WS₂ crystals distributed on graphene sheets have indicated high rate capabilities.¹⁶ However, no studies have been performed to date to assess the ability for WSe₂ to store sodium ions.

In this study, we investigate WSe₂ as a conversion electrode for NIBs for the first time. Bulk WSe₂ is used instead of synthesized nanostructures in order to assess the material properties as opposed to structural effects (SEM of WSe₂ Fig. S1†). Whereas TMDs containing heavier elements such as W or Se may not intuitively be expected to exhibit high specific capacities, our work demonstrates superior energetics of WSe₂ compared to WS₂, and comparable or better specific capacity in comparison to other TMDs containing lighter elements. We highlight the chemical nature of the conversion reaction, speculating on similarities and differences between WSe₂ and heavily studied MoS₂ materials that give promise to WSe₂ as a candidate material for next-generation NIB platforms.

To produce NIB devices from TMD materials, we combined TMDs, super P carbon, and a binder (polyvinylidene fluoride (PVDF) or carboxymethyl cellulose (CMC)) into either n-methyl-2-pyrrolidone (NMP) for PVDF or water for CMC to form a slurry. The slurry was then cast onto a stainless steel disk electrode. The conductivity of the films with CMC, PVDF, and without binder were 1.2, 2.1, and 3.78 × 10⁻⁴ S cm⁻¹ respectively. This electrode was then combined with a glass fiber separator, 1 M NaPF₆ in various electrolytes (4 : 6 v/v ethylene carbonate/diethyl carbonate (EC/DEC), propylene carbonate (PC), and diglyme), and a sodium metal electrode and pressed into a coin cell for testing. Full experimental details are available in the ESI.†

Previous studies have indicated that PVDF with carbonate electrolytes can compromise the integrity of the electrode through unwanted side reactions.^17,18 Recently, Wang et al. studied the effect of binders on microscale MoS₂ in EC/DEC and found that CMC had the highest capacity and also degraded much less than PVDF.¹⁹ In this manner, our first efforts aim to identify WSe₂ electrode, binder, and electrolyte combinations that yield the best battery performance (Fig. 1). Fig. 1a–c and 1d–f show differential capacity plots and cycling data for each electrolyte combined with PVDF and with CMC respectively (charge–discharge profiles Fig. S2†). Differential capacity data is obtained by differentiation of galvanostatic charge–discharge profiles. These provide the same information as the charge–discharge profiles but more obviously highlights the energetics of the reaction. In each case, the sodiation and desodiation reactions remain invariant across the different electrolytes, even in the case of diglyme where solvent co-intercalation reactions can occur that drastically change the energetics.²⁰ Galvanostatic charge–discharge cycling tests were carried out for 30 cycles at 100 mA g⁻¹ (Fig. 1b and e, ESI S2 and S4†), which distinguish the stability of different electrolyte–binder combinations in this system. In all cases, the capacity of CMC-based electrodes is comparable to or greater than the equivalent system with PVDF. For both binders, diglyme electrolytes exhibit higher capacity for the first few cycles but the capacity quickly fades by over 50% after ∼10 cycles with a final capacity of 118 mA h g⁻¹ with PVDF and 117 mA hg⁻¹ with CMC after 30 cycles. PC electrolytes combined with both CMC and PVDF binders also exhibit rapid capacity fade over 30 cycles, while EC/DEC electrolytes lead to a stable capacity through 30 cycles following the first few cycles. We attribute the adverse cycling performance of both PC and diglyme electrolytes to irreversible side reactions occurring between the electrolyte and binder. Such irreversible side reactions are also apparent for PC electrolytes through the low coulombic efficiency between charge and discharge. EC/DEC samples have the highest capacities at the end of 30 cycles for both PVDF (155 mA h g⁻¹) and CMC (190 mA h g⁻¹), with stable cycling performance. This allows us to establish the best electrolyte–binder combination for WSe₂ electrodes, which is EC/DEC combined with CMC binders. This combination shows virtually no capacity fade over the last 25 cycles and exhibits the highest reversible capacity of all electrolyte–binder combinations following 5 cycles of consecutive cycling. Even after 60 cycles this best battery composition has a capacity around 100 mA h g⁻¹ (Fig. S3†). Therefore, this binder/electrolyte combination was used for the rest of the testing.

Fig. 1 (a) dQ/dE for PVDF binder with 1 M NaPF₆ in EC/DEC, diglyme, and PC (b) cycling capacity at 100 mA g⁻¹ for PVDF with all 3 electrolytes (c) capacity retention based on PVDF samples at 100 mA g⁻¹ (d) dQ/dE for CMC binder with 1 M NaPF₆ in EC/DEC, diglyme, and PC (e) cycling capacity at 100 mA g⁻¹ for CMC with all 3 electrolytes (f) capacity retention based on CMC samples at 100 mA g⁻¹.

Cyclic voltammetry (CV) curves for WSe₂ taken at a scan rate of 0.5 mV s⁻¹ are shown in Fig. 2a for the 1^st, 2^nd, and 3^rd cycles with sodiation and desodiation processes labeled. The first sodiation involves a reaction peak at low voltages (<0.25 V vs. Na/Na⁺) without any signature of higher voltage chemical processes. The 2^nd cycle also shows this low voltage signature combined with two higher voltage reaction peaks between 1.2 and 1.85 V vs. Na/Na+. In the 3^rd cycle, the low voltage peak is absent, and only a series of two peaks between 1.2 and 1.85 V are observed. Based on this result, we infer that this low voltage (<0.25 V vs. Na/Na⁺) signature is evidence of Na⁺ intercalation into WSe₂, which upon sodiation goes through a reversible conversion reaction. After the 2^nd CV cycle, the absence of this low voltage peak indicates a fully converted WSe₂ electrode material. The conversion reaction proposed for TMDs, which we later confirm with STEM EDS mapping, produces a Na/chalcogenide conversion product. In the case of WSe₂, the reaction would be

WSe₂ + 2xNa⁺ + 2xe⁻ ↔ 2Na_xSe + W (1)

Fig. 2 (a) CV scans for the first three cycles from WSe₂ devices with CMC and EC/DEC (b) CV scans comparing the energetics of chemical storage in WSe₂ and WS₂ during the 3^rd cycle (c) dQ/dE of a WSe₂ electrode at 10 mA g⁻¹, inset charge discharge curves at 10 mA g⁻¹ (d) galvanostatic rate study at 20, 40, 100, 200, and 400 mA g⁻¹.

The 2^nd and 3^rd cycles display a cathodic peak at 1.25 V and anodic peak at 1.85 V that correspond to a coupled reaction centered near 1.6 V vs. Na/Na⁺. At scan rates of 0.5 mV s⁻¹, the energetics of this redox process are far better than that observed during chemical conversion of WS₂ (Fig. 2b). WS₂ displays an anodic peak also near 1.8 V but a cathodic peak centered near 0.4 V, yielding a much higher overpotential that would bottleneck the energy efficiency of a full cell device. The chemical conversion reaction is further isolated through galvanostatic measurements, where charge and discharge profiles are differentiated with respect to voltage to yield plots of dQ/dE (Fig. 2c). The charge–discharge curves for the dQ/dE can be found in the inset. Analogous to CV curves, a low overpotential (0.26 V) conversion reaction is isolated at currents of 10 mA g⁻¹. Low overpotentials are critical to the operation of an electrode in a full cell configuration where energy efficiency is a benchmark for performance, and our work distinguishes the low overpotential for the WSe₂ conversion reaction from both WS₂ as well as other studied TMDs such as MoS₂ which has an overpotential of 1.3 V for the intercalation reaction and 0.4 V for the conversion reaction.⁷

In addition to the cyclability and energetics another important characteristic of an electrode material is the rate capability (Fig. 2d). WSe₂ was tested at 20, 40, 100, 200, and 400 mA g⁻¹ (charge–discharge curves Fig. S4†). Capacities of WSe₂, which range from over 200 mA h g⁻¹ at the lowest rate to near 135 mA h g⁻¹ at the highest rates, are higher than the best results reported for bulk MoS₂, the most studied TMD in NIBs.^7,21 This is notable since both W and Se are heavier elements than Mo and S, which implies WSe₂ to have superior volumetric storage capacity compared to MoS₂ – an important concept for many emerging energy storage applications. WSe₂ displays a large average extraction capacity of 228 mA h g⁻¹ at 20 mA g⁻¹ and maintains over 60% capacity (130 mA h g⁻¹) at a 20× faster charging rate of 400 mA g⁻¹. The average capacity of the final 5 cycles at 20 mA g⁻¹ show only 2% degradation compared to the average capacity of the initial 5 cycles.

Whereas the premise of this work thus far is that WSe₂ exhibits a conversion reaction in a manner consistent with that observed in MoS₂ materials, ex situ analysis of sodiated WSe₂ materials was performed to assess the signature of the conversion products. A combination of electron dispersive X-ray (EDS) analysis in the transmission electron microscope (TEM), XRD, and Raman spectroscopy was utilized to analyze products following the 5^th discharge for EDS, 5^th discharge and 5^th charge for XRD, and the 1^st discharge for Raman. The 5^th cycle was chosen to characterize for EDS and XRD to ensure full conversion of the WSe₂. Analysis of the 5^th discharge products with XRD indicates the presence of numerous crystalline products that are not associated with the initial WSe₂/PVDF/carbon black electrode (Fig. 3a). The low angle peaks at 11.8° and 17.0° corresponding to an interlayer spacing of 7.48 Å and 5.21 Å are likely attributed to a hexagonal Na_xSe material that forms upon sodiation. The improved crystallinity of WSe₂ conversion products compared to MoS₂ conversion products could lead to lower resistance within the electrode. Additional peaks at 27.9° and 32.4° are consistent with Na₂O (Fig. S2,† JCPDS card no. 03-1074), likely formed due to the air exposure during the measurement. No peaks related to W metal are observed, indicating that W products may form amorphous or poorly crystalline domains.¹² XRD after the 5^th charge cycle reveals a change in the low angle peak compared to the previous sodiated state (Fig. 3b). The peaks at 11.8° and 17.0° are no longer present but the main WSe₂ (002) peak at 13.6° is apparent. The intensity of the WSe₂ peak is lower than the initial state due to a loss in crystallinity but it is obvious that WSe₂ is reformed after sodium extraction.

Fig. 3 (a) XRD pattern of the initial WSe₂, the Na_xSe product formed after the 5^th discharge, and the less crystalline WSe₂ formed after the 5^th charge (b) the same spectrum in (a) but zoomed into only the lower angle peaks.

Ex situ Raman spectroscopy (Fig. 4) supports the formation of a new crystalline product upon sodiation. The initial spectrum has 2 main peaks at 251 and 254 cm⁻¹ corresponding to the E_2G and A_1G WSe₂ peaks (inset Fig. 4).²² After the 1^st discharge, the WSe₂ peaks are no longer present and new peaks arise at 247 cm⁻¹, 329 cm⁻¹, 816 cm⁻¹, and 930 cm⁻¹. We attribute these to Na_xSe conversion products formed during sodiation, even though these peaks are not due to cubic Na₂Se (165 cm⁻¹, 225 cm⁻¹)²³ or Se (235 cm⁻¹).²⁴

Fig. 4 Raman spectra of the initial electrode and the electrode after the 1^st discharge. The inset shows the 2 peak fit of the WSe₂ Raman modes.

Analysis of EDS data from TEM analysis (Fig. 5a–c) indicates a picture consistent with reports on conversion products in MoS₂ materials.¹² Composite EDS maps of the material before conversion shows a homogenous mix of W and selenide species throughout the sample. EDS maps after the 5^th discharge elucidates the formation of isolated W domains adjacent to regions that contain selendies (Fig. 5b). The Na species is located in the same spatial region as the selenides, supporting the formation of Na_xSe (Fig. 5c). The separation of the domains is likely due to W diffusion through the lattice.²⁵ This supports the storage of Na through the conversion reaction discussed in eqn (1), where domains of Na_xSe and W reversibly interchange between sodiated (W + Na_xSe) and mostly desodiated (Na_xWSe₂) states. Since the capacities we measure are comparable to MoS₂ despite the heavy nature of both W and Se compared to Mo and S, we speculate that the improved conductivity of metal selenides compared to metal sulfides could play a role to improve the electrical connectivity of sodium-storage converted products in bulk matrices of TMD materials.²⁶

Fig. 5 (a) EDS composite map of a WSe₂ electrode before cycling, scale bar = 100 nm (b) EDS composite map of a WSe₂ battery electrode after the 5^th discharge showing W and Se, scale bar = 100 nm (c) the same EDS spectrum as (b) but with the Na added (d) schematic illustrating the storage mechanism of Na in WSe₂. Upon Na⁺ insertion segregated domains of Na_xSe and W form. After the Na⁺ has been removed, WSe₂ is reformed but it is less crystalline than in its original state.

Therefore, based on both electrochemical data and the direct assessment of sodiated products through EDS in the TEM, Raman spectroscopy, and XRD, we propose a mechanism of eqn (1) where crystalline WSe₂ is converted into sodiated products of Na_xSe and poorly crystalline or amorphous W metal. Upon removing the sodium, WSe₂ is reformed but is less crystalline than the original sample. This mechanism is illustrated in Fig. 5d. This general conversion reaction is consistent with previous studies for MoS₂ materials, even though WSe₂ exhibits an efficient conversion reaction characterized by a crystalline product, low overpotential, good cyclability even in bulk, and improved specific capacity despite the use of a compound with heavier elements. This distinguishes WSe₂ as a strong candidate for high efficiency sodium ion battery electrodes mediated by a sodium–selenium conversion reaction yielding comparable specific capacity and better volumetric capacity compared to widely studied MoS₂ materials.

In closing, WSe₂ is demonstrated as an efficient electrode material for a sodium ion battery for the first time. 6 different binder/electrolyte combinations have been studied and CMC binder with a NaPF₆ in EC/DEC electrolyte has been shown to have the highest capacity (190 mA h g⁻¹) and best capacity retention (72%) after 30 cycles with virtually negligible capacity change over the last 25 cycles. Directly comparing the energetics of WSe₂ to WS₂ electrodes, we observe a significantly reduced overpotential of 0.26 V, making the energetics of this conversion reaction practical for high energy efficiency devices. WSe₂ electrodes exhibit over 200 mA h g⁻¹ reversible specific capacity at 20 mA g⁻¹ rates and maintain 60% capacity at 400 mA g⁻¹. Based on performance alone, this material exhibits storage capability comparable to previous reports on MoS₂, despite the heavier W and Se elemental components – hence improved volumetric capability. Ex situ analysis including EDS in the TEM, XRD, and Raman spectroscopy indicate the conversion mechanism as the formation of poorly crystalline or amorphous W domains alongside Na_xSe products, which we speculate is enhanced for WSe₂ due to the more conductive nature of Se-based conversion products compared to S-based products. This work highlights WSe₂ as a highly promising and practical electrode material for NIB applications, with a multitude of routes such as nano/microstructuring,²⁷ surface modification through ALD passivation,^28,29 or forming graphene-WSe₂ composites,^30–32 to allow for further improving and optimizing the already promising performance for broad impact of this TMD in next-generation battery technology.

Conflict of interest

The authors declare no competing financial interests.

Acknowledgements

We thank A. S. Westover, A. Douglas, and N. Muralidharan for useful discussions related to this work. This work was supported in part by NSF grant CMMI 1334269, Vanderbilt start-up funds, K. S. and A. P. C. are supported by NSF GFRP fellowships grant # 1445197, and J. L. was supported by NSF REU support.

References

1. J.-M. Tarascon and M. Armand, Issues and Challenges Facing Rechargeable Lithium Batteries, Nature, 2001, 414, 359–367.
2. V. Palomares, P. Serras, I. Villaluenga, K. B. Hueso, J. Carretero-González and T. Rojo, Na-Ion Batteries, Recent Advances and Present Challenges to Become Low Cost Energy Storage Systems, Energy Environ. Sci., 2012, 5, 5884–5901.
3. M. D. Slater, D. Kim, E. Lee and C. S. Johnson, Sodium-Ion Batteries, Adv. Funct. Mater., 2013, 23, 947–958.
4. P. Ge and M. Fouletier, Electrochemical Intercalation of Sodium in Graphite, Solid State Ionics, 1988, 28, 1172–1175.
5. M. M. Doeff, Y. Ma, S. J. Visco and L. C. de Jonghe, Electrochemical Insertion of Sodium into Carbon, J. Electrochem. Soc., 1993, 40, 169–170.
6. H. Liu, D. Su, G. Wang and S. Z. Qiao, An Ordered Mesoporous WS₂ Anode Material with Superior Electrochemical Performance for Lithium Ion Batteries, J. Mater. Chem., 2012, 22, 17437–17440.
7. J. Park, J.-S. Kim, J.-W. Park, T.-H. Nam, K.-W. Kim, J.-H. Ahn, G. Wang and H.-J. Ahn, Discharge Mechanism of MoS₂ for Sodium Ion Battery: Electrochemical Measurements and Characterization, Electrochim. Acta, 2013, 92, 427–432.
8. B. Liu, T. Luo, G. Mu, X. Wang, D. Chen and G. Shen, Rechargeable Mg-Ion Batteries Based on WSe₂ Nanowire Cathodes, ACS Nano, 2013, 7, 8051–8058.
9. S. Zhang, X. Yu, H. Yu, Y. Chen, P. Gao, C. Li and C. Zhu, Growth of Ultrathin MoS₂ Nanosheets with Expanded Spacing of (002) Plane on Carbon Nanotubes for High-Performance Sodium-Ion Battery Anodes, ACS Appl. Mater. Interfaces, 2014, 6, 21880–21885.
10. Y.-X. Wang, K. H. Seng, S.-L. Chou, J.-Z. Wang, Z. Guo, D. Wexler, H.-K. Liu and S.-X. Dou, Reversible Sodium Storage Via Conversion Reaction of a MoS₂–C Composite, Chem. Commun., 2014, 50, 10730–10733.
11. S. D. Lacey, J. Wan, A. V. W. Cresce, S. M. Russell, J. Dai, W. Bao, K. Xu and L. Hu, Atomic Force Microscopy Studies on Molybdenum Disulfide Flakes as Sodium-Ion Anodes, Nano Lett., 2015, 15, 1018–1024.
12. X. Wang, X. Shen, Z. Wang, R. Yu and L. Chen, Atomic-Scale Clarification of Structural Transition of MoS₂ Upon Sodium Intercalation, ACS Nano, 2014, 8, 11394–11400.
13. X. Chia, A. Ambrosi, Z. Sofer, J. Luxa and M. Pumera, Catalytic and Charge Transfer Properties of Transition Metal Dichalcogenides Arising from Electrochemical Pretreatment, ACS Nano, 2015, 9, 5164–5179.
14. Y. N. Ko, S. H. Choi, S. B. Park and Y. C. Kang, Hierarchical MoSe₂ Yolk–Shell Microspheres with Superior Na-Ion Storage Properties, Nanoscale, 2014, 6, 10511–10515.
15. H. Wang, X. Lan, D. Jiang, Y. Zhang, H. Zhong, Z. Zhang and Y. Jiang, Sodium Storage and Transport Properties in Pyrolysis Synthesized MoSe₂ Nanoplates for High Performance Sodium-Ion Batteries, J. Power Sources, 2015, 283, 187–194.
16. D. Su, S. Dou and G. Wang, WS₂ @ Graphene Nanocomposites as Anode Materials for Na-Ion Batteries with Enhanced Electrochemical Performances, Chem. Commun., 2014, 50, 4192–4195.
17. L. O. Vogt, M. El Kazzi, E. Jämstorp Berg, S. A. Pérez Villar, P. Novák and C. Villevieille, Understanding the Interaction of the Carbonates and Binder in Na-Ion Batteries: A Combined Bulk and Surface Study, Chem. Mater., 2015, 27, 1210–1216.
18. M. A. Muñoz-Márquez, M. Zarrabeitia, E. Castillo-Martinez, A. Eguía-Barrio, T. Rojo and M. Casas-Cabanas, Composition and Evolution of the Solid–Electrolyte Interphase in Na₂Ti₃O₇ Electrodes for Na-Ion Batteries: XPS and Auger Parameter Analysis, ACS Appl. Mater. Interfaces, 2015, 7, 7801–7808.
19. X. Wang, Y. Li, Z. Guan, Z. Wang and L. Chen, Micro-MoS₂ with Excellent Reversible Sodium-Ion Storage, Chem.–Eur. J., 2015, 21, 6465–6468.
20. B. Jache and P. Adelhelm, Use of Graphite as a Highly Reversible Electrode with Superior Cycle Life for Sodium-Ion Batteries by Making Use of Co-Intercalation Phenomena, Angew. Chem., Int. Ed., 2014, 53, 10169–10173.
21. G. S. Bang, K. W. Nam, J. Y. Kim, J. Shin, J. W. Choi and S.-Y. Choi, Effective Liquid-Phase Exfoliation and Sodium Ion Battery Application of MoS₂ Nanosheets, ACS Appl. Mater. Interfaces, 2014, 6, 7084–7089.
22. H. Sahin, S. Tongay, S. Horzum, W. Fan, J. Zhou, J. Li, J. Wu and F. Peeters, Anomalous Raman Spectra and Thickness-Dependent Electronic Properties of WSe₂, Phys. Rev. B: Condens. Matter Mater. Phys., 2013, 87, 165409.
23. C. Herwig, M. Schnell and J. Becker, Raman Band of Matrix Isolated Na_mSe_n Clusters, Chem. Phys. Lett., 2004, 385, 462–466.
24. G. Lucovsky, A. Mooradian, W. Taylor, G. Wright and R. Keezer, Identification of the Fundamental Vibrational Modes of Trigonal, α-Monoclinic and Amorphous Selenium, Solid State Commun., 1967, 5, 113–117.
25. M. T. McDowell, Z. Lu, K. J. Koski, J. H. Yu, G. Zheng and Y. Cui, In Situ Observation of Divergent Phase Transformations in Individual Sulfide Nanocrystals, Nano Lett., 2015, 15, 1264–1271.
26. M. Chhowalla, H. S. Shin, G. Eda, L.-J. Li, K. P. Loh and H. Zhang, The Chemistry of Two-Dimensional Layered Transition Metal Dichalcogenide Nanosheets, Nat. Chem., 2013, 5, 263–275.
27. M. Xu, F. Yi, Y. Niu, J. Xie, J. Hou, S. Liu, W. Hu, Y. Li and C. M. Li, Solvent-Mediated Directionally Self-Assembling MoS₂ Nanosheets into a Novel Worm-Like Structure and Its Application in Sodium Batteries, J. Mater. Chem. A, 2015, 3, 9932–9937.
28. W.-H. Ryu, J.-W. Jung, K. Park, S.-J. Kim and I.-D. Kim, Vine-Like MoS₂ Anode Materials Self-Assembled from 1-D Nanofibers for High Capacity Sodium Rechargeable Batteries, Nanoscale, 2014, 6, 10975–10981.
29. B. Ahmed, D. H. Anjum, M. N. Hedhili and H. N. Alshareef, Mechanistic Insight into the Stability of HfO₂-Coated MoS₂ Nanosheet Anodes for Sodium Ion Batteries, Small, 2015, 11, 4341–4350.
30. J. Wang, C. Luo, T. Gao, A. Langrock, A. C. Mignerey and C. Wang, An Advanced MoS₂/Carbon Anode for High-Performance Sodium-Ion Batteries, Small, 2014, 11, 473–481.
31. L. David, R. Bhandavat and G. Singh, MoS₂/Graphene Composite Paper for Sodium-Ion Battery Electrodes, ACS Nano, 2014, 8, 1759–1770.
32. X. Xie, Z. Ao, D. Su, J. Zhang and G. Wang, MoS₂/Graphene Composite Anodes with Enhanced Performance for Sodium-Ion Batteries: The Role of the Two-Dimensional Heterointerface, Adv. Funct. Mater., 2015, 25, 1393–1403.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details, additional galvanostatic charge–discharge measurements during cycling, and SEM characterization of WSe₂. See DOI: 10.1039/c5ra19717a

---