Interface strain in vertically stacked two-dimensional heterostructured carbon-MoS2 nanosheets controls electrochemical reactivity

Two-dimensional (2D) materials offer numerous advantages for electrochemical energy storage and conversion due to fast charge transfer kinetics, highly accessible surface area, and tunable electronic and optical properties. Stacking of 2D materials generates heterogeneous interfaces that can modify native chemical and physical material properties. Here, we demonstrate that local strain at a carbon-MoS2 interface in a vertically stacked 2D material directs the pathway for chemical storage in MoS2 on lithium metal insertion. With average measured MoS2 strain of ∼0.1% due to lattice mismatch between the carbon and MoS2 layers, lithium insertion is facilitated by an energy-efficient cation-exchange transformation. This is compared with low-voltage lithium intercalation for unstrained MoS2. This observation implies that mechanical properties of interfaces in heterogeneous 2D materials can be leveraged to direct energetics of chemical processes relevant to a wide range of applications such as electrochemical energy storage and conversion, catalysis and sensing.

2D graphene and transition metal dichalcogenide (TMDC) materials have captivated researchers in the past decade owing to a set of unique physical and chemical properties that deviate from their bulk analogues and the exploitation of these properties in broad applications 1 . Specifically for electronics, vertical integration of 2D materials enables logic component design with high on-off ratio and promise for three-dimensional (3D) electronics that progress beyond silicon [2][3][4][5][6][7][8] . In semiconductor electronics, engineering 2D TMDC materials using strain applied at an interface has been shown to strongly modulate the bandgap and band structure, which results in modified electrical and optical properties for strained materials [9][10][11][12][13][14] . For monolayer MoS 2 , the bandgap is observed to shift by up to 15 meV under tensile strain of up to 4.8% (ref. 15), and for multilayered WSe 2 the bandgap transitions from indirect to direct under strain of up to 2% (ref. 16). Theoretical efforts have further emphasized strain-enabled broadband absorption and photodetection in MoS 2 , even though this has not yet been experimentally realized 17 . The intersection of strain-engineered properties of 2D materials and heterostructured vertically integrated interfaces of 2D materials presents an engaging research area for next-generation electronics. Nonetheless, such materials and research directions remain elusive for many other applications of 2D materials.
On this front, 2D materials and specifically 2D TMDCs have demonstrated excellent performance in a range of electrochemical applications including lithium-and sodium-ion batteries [18][19][20] , photocatalytic conversion [21][22][23] and biosensing applications [24][25][26] . Unlike electronic devices, these applications require bulk-like quantities of 2D materials-a challenge aided by recent developments in the liquid exfoliation and assembly of layered TMDCs [27][28][29] . However, addressing how properties in heterostructured or complex 2D materials can impact chemical processes responsible for electrochemical applications is hampered by uniform material fabrication routes that can be used on scales required for electrochemical measurements. As a result, the impact of strained interfaces in 2D materials on chemical and electrochemical processes remains virtually unstudied. Only recently has an observation emerged that compressive strain on Pt catalysts can improve the oxygen reduction reaction capability of Pt relevant to fuel cells 30 . In this manner, 2D materials provide an ideal test bed for the understanding of how interfaces and strain can impact electrochemical processes, motivated by pioneering efforts in the field of semiconductor electronics. This is the focus of this report, which documents that interface strain measured over statistical quantities of stacked 2D carbon-MoS 2 heterostructured nanosheets can be directly correlated with distinct differences in chemical processes occurring in 2D materials. In particular, combining optical and electrochemical techniques we demonstrate that strain engineering of interfaces can enable control of the energetic pathway for the chemical conversion of MoS 2 into electrochemically active Mo and Li 2 S n species during reaction with lithium. This highlights the important role that mechanical strain can have in engineering energy storage processes in materials.

Results
Fabrication of interface-strained materials. MoS 2 nanosheets were produced through liquid exfoliation of bulk MoS 2 powders in n-methyl-2-pyrrolidone (NMP) solvents and subsequent centrifugation. Transmission electron microscopy (TEM) of a representative exfoliated MoS 2 nanosheet is shown in Fig. 1a,b. Interlayer spacing of B0.61 nm is observed for MoS 2 nanosheets, with thicknesses ranging from 2-15 atomic layers ( Supplementary Fig. 1, Supplementary Note 1). Ultrathin carbon layers are grown directly on the MoS 2 surface through MoS 2 catalysed decomposition of C 2 H 2 precursors using a temperature ramp chemical vapour deposition process ( Supplementary Fig. 2) that is capable of gram-scale batch processing. This generates vertically stacked architectures where ultrathin carbon layers are formed on both sides of the MoS 2 nanosheets, with a representative TEM image of this architecture shown in Fig. 1c, and corresponding elemental analysis map in Fig. 1d. To produce an appreciable mass of material for electrochemical tests, electrophoretic deposition was used to assemble the vertically stacked carbon-interfaced MoS 2 (C-MoS 2 ) nanosheets into conformal films on metal substrates from NMP dispersions following chemical vapour deposition ( Fig. 1e; Supplementary Fig. 3, Supplementary Note 2). All aspects of synthesis and processing of C-MoS 2 materials are chosen to be compatible with scalable processing, and ongoing research efforts to improve the simplicity and scalability of liquid exfoliation can further improve this 31 . Raman spectroscopy analysis (Fig. 1f) of vertically stacked C-MoS 2 nanosheets indicates that vertical stacks maintain an identical signature of crystalline MoS 2 following the carbon synthesis based on E 2g and A 1g modes (300-500 cm À 1 ). The carbon layers exhibit a significant amount of sp 3 hybridized carbon atoms (B1,320 cm À 1 , or D-band) relative to sp 2 carbon species (B1,580 cm À 1 , or G-band). Whereas this indicates the presence of carbon, X-ray photoelectron spectroscopy (XPS) indicates the emergence of both a peak in the sulfur 2p 3/2 and 2p 5/2 spectra at low-binding energies, as well as shoulders in both Mo 3d 5/2 and 3d 3/2 peaks that represent a modulation to the local bonding environment in the crystal (Fig. 1g). Notably, the existence of the low-energy shoulder in the sulfur 2p and Mo 3d peaks is consistent with the formation of the metallic 1T-phase of MoS 2 produced from the strain-induced distortion of sulfur atoms observed previously ( Supplementary Fig. 4, Supplementary Note 3) 32 . This analysis collectively supports the formation of a distinct carbon-MoS 2 stacked interface, where mismatch of the in-plane lattice spacing between MoS 2 and carbon will induce significant interface strain that propagates into the MoS 2 nanosheet from the C-MoS 2 interface.
To assess strain in the vertically stacked 2D material, statistical Raman spectroscopy mapping comprising over 200 individual scans in separate areas was performed on the E 2g and A 1g modes of MoS 2 , which are highly sensitive to tensile or compressive strain [33][34][35] . (Fig. 2a,b) On the basis of the peak-to-peak analysis of Raman mode distributions, blue-shifts of B0.66 cm À 1 and B0.59 cm À 1 were observed for the E 2g and A 1g modes, respectively. Asymmetry in these modes is expected and attributed to stronger electronic coupling to the A 1g mode 36 . This yields an B0.1-0.2% compressive strain based on relative A 1g mode shifts in accordance with previous studies 9, 34 , supporting the presence of interface-induced compressive strain on the MoS 2 material. This is further confirmed using X-ray diffraction which demonstrates a similar B0.1% compressive strain due to vertical stacking based on analysis of the (100) and (110) low-index planes of MoS 2 (Fig. 2c). We anticipate this measured strain to be present at both the interface and on the interior of the nanosheets as they possess a thickness well below the critical layer thickness at which strain relaxation occurs 37 .
Controlling electrochemical processes using strain. To assess how the vertically stacked architecture and interface strain influences electrochemical processes, we combined electrophoretically assembled vertically stacked 2D C-MoS 2 nanosheets with Li metal electrodes, and a 1.0 M lithium hexafluorophosphate solution in ethylene carbonate and diethyl carbonate electrolyte, and compared the electrochemical properties against similar electrode materials produced with pristine MoS 2 nanosheets (Supplementary Figs 5,6, Supplementary Notes 4,5). Cyclic voltammetry scans of these electrodes and corresponding differential capacity plots based on galvanostatic measurements are shown in Fig. 3. As these two electrode materials differ only by the presence of a vertically stacked C-MoS 2 interface, electrochemical data indicates significant changes to the chemical processes occurring on lithium insertion into the MoS 2 material. For pristine MoS 2 nanosheets, two subsequent reactions are observed, with the first one at B1.1 V and the second at B0.55 V versus Li/Li þ . (Fig. 3a,b) This is consistent with the known pathways for insertion of lithium into pristine MoS 2 , which occurs first through an intercalation reaction (1.1 V) that follows   where MoS 2 undergoes a transition from a semiconducting 2H phase to the metallic 1 T phase due to strain-induced deformation of the crystal from lithium ion insertion. Following this transition, the MoS 2 can undergo a subsequent conversion reaction at lower voltages (0.55 V) that follows In contrast to this, insertion of lithium into the vertically stacked C-MoS 2 nanosheets yields a chemical reaction evident at B2.3 V versus Li/Li þ based on both CV and diferential capacity curves (Fig. 3c,d) which is close to the open-circuit voltage (OCV) of the device. This reaction proceeds in the absence of Li intercalation into MoS 2 and no significant signature of lithium insertion at lower voltages occurs such as in pristine MoS 2 nanosheet electrodes. This highlights the presence of a chemical storage process occurring in vertically stacked C-MoS 2 materials that is not observed in pristine MoS 2 materials (Fig. 4a) and that interface-induced strain can directly trigger chemical conversion at practical voltages. It is important to note that these results distinguish the stacked C-MoS 2 configuration from simple mixtures of carbon nanomaterials and MoS 2 nanosheets which have been reported previously [38][39][40][41][42] , in two key ways, namely that the grown carbon layer induces interface strain not possible in weakly adhered mechanically mixed materials that modulates the chemical reactivity of the MoS 2 nanosheet toward chemical conversion, and that the stacked configuration provides a physical barrier to inhibit polysulfide dissolution into the electrolyte in a manner consistent with a yolk-shell confinement strategy 43 . However, the benefits reported in other studies by combining MoS 2 and carbons through mixing are also maintained in our study as evidenced by galvanostatic studies indicating better rate capability and improved storage capacity at high rates for the C-MoS 2 electrodes ( Supplementary Fig. 7). In addition, whereas the correlation between X-ray diffraction, Raman spectroscopy and XPS indicate the presence of strain due to the mismatch of in-plane lattice parameters of the stacked materials below the critical thickness, charge transfer effects between the C and MoS 2 materials can play a role in the origin of strain measured using these techniques. We also observe the lack of any electrochemical signature of lithium intercalation into the ultrathin carbon coating, implying that stacked C-MoS 2 nanosheets exhibit differences from pristine nanosheets that can be explained by compressive strain imposed on the 2D MoS 2 material.
To address the effect of the C-MoS 2 interface on the electrochemical properties during lithium insertion, Raman spectroscopy of the electrophoretically assembled MoS 2 -based electrodes at different cathodic potentials was carried out (Fig. 4b). This analysis is possible due to the preparation of electrodes in a manner that does not require binder materials often used in conventional battery electrodes that can overwhelm the desired Raman spectroscopic features. At OCV conditions, both electrode materials exhibit only the native Raman modes of MoS 2 . Cathodic scans from OCV conditions to 1.75 V versus Li/ Li þ -an energy below the Li þ insertion reaction observed for vertically stacked C-MoS 2 nanosheets, yields no change for the pristine MoS 2 nanosheets, but the emergence of a distinct Raman peak at 746 cm À 1 is observed for the stacked C-MoS 2 nanosheets. This mode is due to the formation of lithium polysulfides (Li 2 S n for 4rnr8) that can be attributed to chemical conversion of the MoS 2 into reaction products in a manner consistent with equation (2)  A 1g and E 2g modes are observed as expected from conversion ( Supplementary Fig. 8, Supplementary Note 6). This confirms that the high voltage B2.3 V signature in the CV and differential capacity curves for vertically stacked C-MoS 2 nanosheets (Fig. 3c,d) is a chemical conversion process similar to that which is known to occur at low voltages (0.55 V) in pristine MoS 2 . To further support this, cathodic scans were continued down to 0.01 V versus Li/Li þ for the pristine MoS 2 electrode, where the signature of conversion is evident due to the presence of the Raman mode at 746 cm À 1 . Despite the difference in voltage of the conversion reactions between MoS 2 and C-MoS 2 , similar Raman spectroscopic signatures of the polysulfide conversion product and similar post-conversion redox energetics in both cases implies that the resulting chemical state of the material in both conversion reactions is the same. To further support this, we performed in situ electrochemical impedance spectroscopy (EIS, Supplementary Fig. 9, Supplementary Table 1), where EIS analysis was performed at the same voltages as those assessed using Raman spectroscopy. EIS studies elucidate a picture consistent with Raman spectroscopy results that indicates the electrochemical signature of conversion in different voltage regimes for the pristine MoS 2 and the stacked C-MoS 2 , respectively (Supplementary Note 7).

Discussion
On the basis of the combined spectroscopic and electrochemical analysis, a picture emerges emphasizing the role of interface strain on lithium insertion in vertically stacked C-MoS 2 materials. In the pristine semiconducting 2H phase of MoS 2 , conversion of MoS 2 nanosheets cannot occur at voltages above where intercalation occurs (1.1 V versus Li/Li þ ) and can only proceed in a two-step chemical process following an intercalation reaction. As many researchers have discussed the promise of MoS 2 nanosheets for battery cathodes, the requirement of a conversion reaction that is near the reduction potential of lithium is highly impractical in full-cell architectures. In a vertically stacked C-MoS 2 nanosheet, the interface strain due to lattice mismatch in a carbon-MoS 2 solid-solid interface leads to an average B0.1% compressive strain that propagates into the MoS 2 nanosheet lattice and enables control of the energetics of chemical conversion. In this regard, interface strain provides the appropriate energetic landscape to sustain direct conversion at voltages that enable pairing with conventional anode materials such as silicon or graphite which give promise to practical incorporation of this device in a full-cell battery configuration. This distinctive difference in the chemical pathways achieved during lithium insertion as a result of interface strain is illustrated in Fig. 4a, where the chemical state of the 2D materials is drastically different following a cathodic scan to 1 V.
The broader use of controlled interface mechanics to modify chemical pathways in low-dimensional materials presents an unexplored horizon for engineered low-dimensional nanostructures. This utilizes 2D materials as a framework to merge concepts of strain engineering with mechanochemistry to control the energetics of chemical processes relevant to energy storage, energy conversion, catalysis, and sensing applications. Building from our results showing that strained interfaces in vertically stacked C-MoS 2 nanosheets can modulate electrochemical reactivity, we envision a new paradigm for material design that exploits interface strain and mechanics as a versatile toolbox to modulate performance of designer 2D materials in diverse applications.

Methods
MoS 2 nanosheet synthesis. In all, 500 mg of bulk MoS 2 powder (Aldrich, particle sizeo2 mm) was added to 50 ml of 1-methyl-2-pyrrolidinone (Aldrich, 99.5% anhydrous). The solution was sonicated in a bath sonicator for 12 h using 8 Â 90 min intervals. After sonication, the solution was centrifuged at 2,000 r.p.m. for 40 min. The upper 2/3 of the supernatant was removed and subsequently centrifuged for 1 h at 5,000 r.p.m. before the excess supernatant was removed and the accumulated nanosheets were placed under vacuum overnight to completely evaporate the solvent.
Vertically stacked carbon-MoS 2 nanosheet synthesis. MoS 2 nanosheets contained within an alumina boat were placed into a 1-inch tube furnace and the tube was evacuated to 2 mTorr. A gas mixture of 100 standard cubic centimeters per minute (sccm) Ar and 20 sccm H 2 maintained at atmospheric pressure were introduced during ramping to a temperature 750°C. At 750°C, 2 sccm of acetylene (C 2 H 2 ) was introduced for 10 min followed by a temperature ramp to 850°C and an additional 10 min soak, and a final ramp to 950°C for 10 min. After this process, the acetylene was turned off and the furnace cooled back down under a flow of argon and hydrogen, where materials were removed.
Electrophoretic deposition. A total of 20 mg of vertically stacked C-MoS 2 or pristine MoS 2 nanosheets were suspended in 20 ml NMP and sonicated for 30 min. Following this, a stainless steel spacer was placed at 0.5 cm from a 1 Â 1 cm stainless steel counter electrode. A voltage of 30 V was applied for 30 min before carefully removing the coated steel electrode from solution and placing this into a vacuum chamber overnight to dry the sample.
Electrochemical device fabrication and testing. Electrochemical half-cell devices were assembled in an argon glovebox using CR 2,032 stainless steel coin cells purchased from MTI. The coated steel discs were separated from a lithium metal anode by a 2,500 Celgard separator saturated with a 1 M LiPF 6 in 1 g per 1 ml solution of ethylene carbonate and diethyl carbonate. Cyclic voltammetry scans and galvanostatic measurements were performed using a Metrohm autolab multichannel testing system.
Material characterization. Raman measurements were performed using a Renishaw inVia confocal Raman spectrometer. Micro-Raman maps were collected using a 532 nm laser. Before characterizing the electrodes after electrochemical testing, coin cells were disassembled in an Ar filled glovebox and the electrodes washed with an EC/DEC solution. TEM analysis was performed by drop casting a solution of pristine or stacked MoS 2 materials onto amorphous carbon TEM grids and then imaged using an FEI Osiris TEM at a beam voltage of 200 kV. X-ray diffraction was performed on powder samples using a Scintag XGEN 4,000 system with a CuKa (wavelength ¼ 0.154 nm) radiation source. XPS was collected using a PHI 5,000 Versaprobe. Measurements were collected at a 45°takeoff angle with a 100 mm spot size with charge neutralization.
Data availability. The data that support the findings of this study are available from the corresponding author upon request.