Crumpled Transition Metal Dichalcogenide Sheets

Abstract

An electrohydrodynamic (EHD) method results in charge driven droplet fission and volumetric shrinkage of metal dichalcogenide sheet droplets, thus wrinkling individual exfoliated metal dichalcogenide sheets on the nanoscale. For example, the method can be used to activate the basal plane of solution exfoliated, stable 2H reverted sheets of MoS2. In this manner, the basal plane of 2H MoS2 can be strained at an unprecedented scale (3.7%), resulting in charge transport to basal plane defects. The resulting crumpled MoS2 integrates few layer sheets and atomistic defects with nano- and micro-scale ridges and vertices to enable centimeter-length films that elevate the catalytic site count of MoS2 from 1.0×1014 sites/cm2 to 2.93×1017 sites/cm2 (planar 2H vs. crumpled 2H) and a turn-over-frequency (TOF) from 0.016 s−1 to 0.130 s−1.

Description

CROSS-REFERENCE TO RELATED APPLICATION

This application claims the benefit of U.S. Provisional Application No. 62/526,828, filed Jun. 29, 2017, which is incorporated herein by reference.

STATEMENT OF GOVERNMENT INTEREST

This invention was made with Government support under Contract No. DE-NA0003525 awarded by the United States Department of Energy/National Nuclear Security Administration. The Government has certain rights in the invention.

FIELD OF THE INVENTION

The present invention relates to transition metal dichalcogenides and, in particular, to crumpled transition metal dichalcogenide sheets, such as crumpled molybdenum disulfide (MoS₂), that can have catalytic and other applications.

BACKGROUND OF THE INVENTION

The urgent need to transition the global energy infrastructure away from fossil fuels and toward carbon-neutral sources has propelled a resurgence of research in renewable hydrogen generation. Within this context, there is great interest in the development of hydrogen evolution reaction (HER) electrocatalysts that concurrently possess high total electrode activity (e.g., overpotentials, Tafel slope, and exchange current density) at a low catalyst loading, and tailored intrinsic activity (e.g., turnover frequency (TOF) per active site). See J. D. Benck et al., ACS Catal. 4, 3957 (2014). To have widespread impact, such catalytic materials must also be electrochemically stable, thermally robust, and synthetically scalable without the need for specialized processing equipment. While Pt nanoparticles represent the current benchmark for HER electrocatalysis, the elemental scarcity, high cost, and poor stability against nanoparticle coarsening and agglomeration impose practical limitations and motivate the need for a next generation of alternative HER catalysts.

Molybdenum disulfide (MoS₂), a van der Waals solid composed of adhered-2D monolayers, has emerged as a promising alternative to Pt due to its elemental abundance, high catalytic activity, and electrochemical stability. See J. D. Benck et al., ACS Catal. 4, 3957 (2014); J. Greeley et al., Nat. Mater. 5, 909 (2006); B. Hinnemann et al., J. Am. Chem. Soc. 127, 5308 (2005); and Q. Ding et al., Chem. 1, 699 (2016). Despite the various merits of MoS₂, its functional performance remains inferior to that of Pt due to comparatively small active site concentrations as well as its relatively low electrical conductivity. At the root of functional limitations is the inability of MoS₂, in its natural 2H semiconducting phase, to catalyze HER on its inert basal plane, where the Gibbs free energy for hydrogen adsorption (ΔG_H) is exceedingly large. See J. D. Benck et al., ACS Catal. 4, 3957 (2014); T. F. Jaramillo et al., Science 317, 100 (2017); and J. Kibsgaard et al., Nat. Mater. 11, 963 (2012). In contrast, MoS₂ edges possess a favorable ΔG_H, as well as metallic 1D conductivity due to the presence of electronic states in close proximity to the Fermi level. Because of the highly catalytic nature of MoS₂ edges, synthetic approaches that prioritize MoS₂ edges sites over the inert basal plane sites have attracted broad interest. See Q. Ding et al., Chem. 1, 699 (2016); T. F. Jaramillo et al., Science 317, 100 (2017); J. Kibsgaard et al., Nat. Mater. 11, 963 (2012); M.-R. Gao et al., Nat. Commun. 6, 7493 (2015); Z. Lu et al., Adv. Mater. 26, 2683 (2014); and J. Ekspong et al., Adv. Funct. Mater. 26, 6766 (2016). Notable innovations in this area include nanostructured MoS₂ particles, vertically aligned flakes, and wires in conjunction with the engineering of defects, interfacing to conductive scaffolds, and mimicry of the edge structures. See Y. Li et al., J. Am. Chem. Soc. 133, 7296 (2011); D. Kong et al., Nano Lett. 13, 1341 (2013); Z. Chen et al., Nano Lett. 11, 4168 (2011); J. Xie et al., Adv. Mater. 25, 5807 (2013); M. A. Lukowski et al., J. Am. Chem. Soc. 135, 10274 (2013); G. Ye et al., Nano Lett. 16, 1097 (2016); A. J. Smith et al., Adv. Energy Mater. 4, 1400398 (2014); Y. H. Chang et al., Adv. Mater. 25, 756 (2013); and H. I. Karunadasa et al., Science 335, 698 (2012). Yet, despite these innovations, the fundamental inactivity of the MoS₂ basal plane remains a basic material constraint. To enhance catalytic activity, basal plane activation via lithiation and ultrasonication has been used to induce phase transformation into the conductive 1T′ MoS₂ polymorph (sometimes termed “metallic,” “chemically exfoliated” MoS₂, or ce-MoS₂), which is catalytically active on the basal plane. See M. A. Py and R. R. Haering, Can. J. Phys. 61, 76 (1983); S. S. Chou et al., Nat. Commun. 6, 8311 (2015); and D. Voiry et al., Nat. Mater. 12, 850 (2013). The phase engineering strategy provides a new pathway to drastically enhance the overall HER metrics of 1T′ MoS₂ that has been recently extended to WS₂ and MoSe₂. See M. a Lukowski et al., Energy Environ. Sci. 7, 2608 (2014); and Y. Yin et al., Adv. Mater. 29, 1700311 (2017). However, this strategy comes at the cost of reduced robustness, since the 1T′ phase is metastable and temperature sensitive. See M. Chhowalla et al., Nat. Chem. 5, 263 (2013).

Beyond phase transformation, basal plane activation via strain engineering has been explored. This approach takes advantage of the shift of the d-band electronic structure with strain, which facilitates a semiconducting-to-metallic phase transformation. See L. Hromadova et al., Phys. Rev. B: Condens. Matter Mater. Phys. 87, 1 (2013). However, retention of the metallic phase requires continual application of pressure (≈35 GPa), which is impractical for technological use. See Y.-C. Lin et al., Nat. Nanotechnol. 9, 391 (2014); A. P. Nayak et al., Nat. Commun. 5, 1 (2014); and Z. H. Chi et al., Phys. Rev. Lett. 113, 1 (2014). A hybrid approach using strain on plasma-treated MoS₂ has also been reported, with localized substrate strain plus S vacancies from the plasma treatment shifting its ΔG_H towards thermal neutral (≈0 eV), thus activating the basal plane and enabling HER. See H. Li et al., Nat. Mater. 15, 48 (2016); H. Li et al., Nat. Commun. 6, 7381 (2015); C. C. Cheng et al., Nano Energy 30, 846 (2016); and G. Li et al., J. Am. Chem. Soc. 138, 16632 (2016). However, as strain here is derived from lithographically defined substrate topography, it is potentially too costly for production. See H. Li et al., Nat. Commun. 6, 7381 (2015). Furthermore, because the product is inherently planar, it has limited prospects for integration in multilayer systems, which is a practical need for many applications and constrains device architectures. Most recently, the basal plane of MoS₂ has also been “activated” by improving substrate/electron transport to defects within MoS₂ sheets. See D. Voiry et al., Nat. Mater. 15, 1003 (2016). However, this method similarly requires lithography and patterning, rendering scalability a challenge. It should also be noted that the 1T′, the strained S-vacancy, and the electronic coupling approaches to basal plane activation are all limited to extremely thin specimens to 2 layers), thus limiting the number of active sites and making scalable integration difficult.

SUMMARY OF THE INVENTION

The present invention is directed to a method of crumpling transition metal dichalcogenide sheets, comprising forming an electrostatically charged liquid droplet comprising a plurality of planar sheets of a transition metal dichalcogenide; applying a sufficient electric field to the droplet to cause the droplet to rupture into a plurality of fission droplets by an electrohydrodynamic process, wherein each fission droplet comprises at least one planar sheet of the transition metal dichalcogenide; and evaporating the remaining liquid from the plurality of fission droplets, thereby inducing spontaneous self-crumpling of the at least one planar sheet of the transition metal dichalcogenide in each of the fission droplets.

The invention provides a facile and high throughput method for obtaining electrochemically stable, thermally robust, and highly efficient HER catalysts via electrohydrodynamic (EHD) induced dimensional transition of 2D planar ce-MoS₂ sheets into 3D structurally deformed MoS₂ nanostructures, namely crumpled MoS₂ (c-MoS₂). The material, obtained from EHD-assisted crumpling, electrostatically isolates and then transforms individual ce-MoS₂ sheets into 3D crumpled architectures, provides highly accessible and electrochemically active surface-area (310.8+/−0.5 m² g⁻¹ and catalytic site density of 5.86×10¹⁷ sites cm⁻²) that allows infiltration of electrolytes to the fully exposed active sites on the activated basal plane and edges while allowing efficient charge transfer between electrolyte, activated basal plane and edges, as confirmed by a substantially reduced charge-transfer resistance (R_ct). Meanwhile, the hierarchically strained conformational features, including facets, folds, ridges, vertices, and wrinkles, can substantially reduce the series resistance (R_s), closely resembling the orbital effect induced by regions of atomically localized curvature in nanocarbon allotropes and textured MoS₂. See S. Luo et al., Nanotechnology 26, 105705 (2015); and D. Srivastava et al., J. Phys. Chem. B 103, 4330 (1999). In addition, the synergistic combination of high strain load emanating from structurally deformed morphology and spatially distributed S vacancies intrinsic to the chemical exfoliation process collectively tailors the 2H basal plane with a catalytically and thermodynamically favorable environment, as confirmed by the improved TOF of 0.130 H₂ molecules per second. See D. Srivastava et al., J. Phys. Chem. B 103, 4330 (1999). The net result is much enhanced HER activity with a Tafel slope of 64 mV per dec, exchange current density of 10.4 μA cm⁻², high cathodic current density (>100 mA cm⁻² at less than −300 mV), and an overpotential (η₁₀) of −191 mV to achieve a current density of −10 mA cm⁻² at an exceedingly low aerial mass loading of only 13 μg cm⁻². The significantly enhanced intrinsic and total electrode HER characteristics compare favorably with state-of-art MoS₂ catalysts while using a fully scalable and easy to integrate material process. See J. Kibsgaard et al., Nat. Mater. 11, 963 (2012); M.-R. Gao et al., Nat. Commun. 6, 7493 (2015); Y. Li et al., J. Am. Chem. Soc. 133, 7296 (2011); D. Kong et al., Nano Lett. 13, 1341 (2013); M. A. Lukowski et al., J. Am. Chem. Soc. 135, 10274 (2013); G. Ye et al., Nano Lett. 16, 1097 (2016); H. Li et al., Nat. Mater. 15, 48 (2016); and D. Voiry et al., Nano Lett. 13, 6222 (2013). More importantly, the structurally deformed MoS₂, with in situ strain fields propagating on the thermodynamically favored 2H phase, demonstrates consistent HER characteristics under continuous operations of 5000 cycles or annealing at high temperatures (up to 300° C.), enabling effective integration into fuel cell and other applications.

BRIEF DESCRIPTION OF THE DRAWINGS

The detailed description will refer to the following drawings, wherein like elements are referred to by like numbers.

FIG. 1(a) is a schematic illustration the chemical exfoliation of bulk MoS₂ powder. FIG. 1(b) is a transmission electron microscope (TEM) image of the exfoliated MoS₂, showing internal edges which increase overall edge density. FIG. 1(c) is a high-angle annular dark-field scanning transmission microscope (HAADF-STEM) image, showing spatially distributed atomistic point defects, including single and double sulfur vacancies (V_S1, V_S2), Mo anti-site substitutions (Mo_S2) and S anti-site substitutions (S2_Mo).

FIG. 2 is a schematic illustration of EHD-assisted crumpling of exfoliated MoS₂, showing that chemical exfoliation also facilitates stable colloidal dispersions of ce-MoS₂ within the EHD-generated droplets that can be viewed as charged, noncolloidal reactors. Under a high electric field, the reactant, 2D ce-MoS₂, within these nanoreactors undergoes stages of (i) electrostatically induced separation, (ii) fission, and (iii) capillarity-induced self-crumpling into 3D c-MoS₂.

FIG. 3(a) is a transmission electron microscope (TEM) image of the 3D c-MoS₂. FIG. 3(b) is a high-resolution transmission electron microscope (HRTEM) image showing extremely thin layers because of electrostatic forces and fission. FIG. 3(c) is an HRTEM image showing sharp vertices and facets with radii of curvatures at nanoscale. FIG. 3(d) is an HRTEM image showing propagating ridges.

FIG. 4 is a graph of Raman spectra showing a progressive red shift of the E¹_2g with increased deposition temperature, indicating a progressively strained morphology.

FIG. 5(a) shows SEM and HRTEM images of MoS₂ deposited at 100° C., revealing rugged island like morphology and relatively flat, featureless basal plane with slightly corrugated folds around the edges when isolated. FIG. 5(b) shows SEM and HRTEM images of EHD deposition at 200° C. coincides with emergence of wrinkled sheets. FIG. 5(c) shows SEM and HRTEM images of MoS₂ deposited at 250° C., showing fully realized individual crumples with porous structures, and highly rippled edges were observed.

FIG. 6 is an XPS spectra of the Mo 3d (left) and S 2p (right) peaks showing mixed 1T′ and 2H XPs signatures in samples deposited at 100° C. and pure 2H phase at 200° C. and 250° C.

FIGS. 7(a) and 7(b) are SEM images showing the uniform and full coverage of c-MoS₂ supported on CC using EHD deposition at different scales. FIG. 7(c) is an HRSEM image providing a close-up view of densely packed crumpled MoS₂ assemblies with porous morphology and highly curved edges.

FIG. 8 is a graph of linear sweep voltammogram (LSV) curves for the carbon cloth (CC) scaffolds, a planar 2H MoS₂ sheet, crumpled MoS₂ deposited at 200° C., and crumpled MoS₂ prepared at 250° C. Overpotentials required to reach current densities of −10 mA/cm² are labeled.

FIG. 9 is a graph of Tafel plots corresponding to the LSV curves. The plots are labeled with Tafel slopes and exchange current densities, showing much improved total electrode HER characteristics at an exceedingly low mass loading of crumpled MoS₂.

DETAILED DESCRIPTION OF THE INVENTION

FIG. 1(a) is a schematic illustration of the synthesis of single layer ce-MoS₂ by a modified Li-intercalation followed by ultrasonic exfoliation. See M. B. Dines, Mater. Res. Bull. 10, 287 (1975); and G. Eda et al., ACS Nano 6, 7311 (2012). In the example described below, a prolonged reaction time of 96 h resulted in the conversion of 64% of trigonal prismatic coordination (2H phase) into octahedral coordination (1T′ phase). High resolution transmission electron microscopy (HRTEM) revealed that the basal plane of the resultant ce-MoS₂ can be readily addressed with significant tears in the 10-nm scale that affords edge-like structures, as shown in FIG. 1(b). Additionally, as shown in FIG. 1(c), high-angle annular dark field (HAADF) and aberration-corrected scanning transmission electron microscopy (STEM) images reveal atomically resolved point defects including S vacancies (V_S), Mo anti-site substitution (Mo_s), and Mo vacancies (V_Mo), mono in tandem with bi-sulfur vacancies (V_S1, V_S2) and antisite substitutions within MoS₂. See J. Xie et al., Adv. Mater. 25, 5807 (2013). Detailed analysis of the MoS₂ basal plane with lateral dimensions of 6.75 nm×6.75 nm (estimated from the lattice distance of 3.16 Å) revealed that the large majority of the defects consist of V_S with a total number about 79 (out of 440 Mo atoms, ≈9% S-vacancies). Theoretical calculations indicate that S vacancies as low as ≈3% lowers ΔG_H from more than 2 eV down to nearly 0.2 eV. See H. Li et al., Nat. Mater. 15, 48 (2016). The average S-vacancy percentage over 8 such areas was calculated to be 9.3+/−1.7%, thus creating a catalytically active basal plane. Although continuous modulation of ΔG_H toward thermal neutral can be made possible by mounting up the number of S-vacancies, the overall structural integrity of ce-MoS₂ sheets is at the mercy of fragmentation.

To further modulate the ΔG_H of ce-MoS₂ sheets while preserving the structural integrity, an EHD process was adapted, which is typically employed for generating ultrafine droplets in high throughput thin film production. See H. Ishihara et al., Sci. Rep. 6, 38701 (2016); H. Ishihara et al., Adv. Mater. Interfaces 3, 1 (2016); and H. Ishihara et al., J. Mater. Chem. A 4, 6989 (2016). As shown in FIG. 2, under a high electric field, the liquid medium deforms and then disintegrates into self-dispersing, charged droplets, providing electrostatically stabilized colloidal systems that inhibit ce-MoS₂ sheets from restacking. The EHD deposition is characterized by a droplet fission process, in which surface charge density drastically increases during solvent-evaporation-induced volumetric shrinkage, ultimately reaching the Rayleigh limit. At this point, electrostatically charged droplets rupture into smaller ones with a diameter distribution in the sub-micrometer range. Thus, each fine droplet contains single- to few-layered ce-MoS₂ sheets before proceeding to a final stage of self-crumpling. In addition to being a dispersing medium, these individualized, charged colloidal droplets can be considered soft templates that impose capillarity-induced geometric constraints to induce the crumpling of ce-MoS₂ sheets upon evaporation. Therefore, ce-MoS₂ within these charged droplets will first separate into approximately individual sheets before being compressed into crumpled nanostructures.

The method of EHD-assisted crumpling can generally be applied to the crumpling of any transition metal dichalcogenide sheet. Transition metal dichalcogenide sheets are atomically thin semiconductors of the type MX₂, where M is a transition metal atom (e.g., Mo, W) and X is a chalcogenide atom (e.g., S, Se, Te). The sheets comprise a layer of M atoms sandwiched between two layers of X atoms, typically in a hexagonal structure. Bulk crystals of transition metal dichalcogenides comprise monolayer sheets that are bound to each other via weak Van-der-Waals attraction. Therefore, MX₂ monolayers can be obtained by exfoliation of the bulk crystal, either by mechanical abrasion or by liquid-phase exfoliation. Alternatively, monolayer MX₂ can be grown by chemical vapor deposition or molecular beam epitaxy. These MX₂ sheets can be crumpled following the same EHD process described above.

EXAMPLE

Lithium Intercalation and Exfoliation and MoS₂ Crumpling by EHD

As an example of the invention, Li-intercalation was accomplished by immersing 2 g of MoS₂ powder in 15 mL of 0.8 m n-butyl lithium in hexane. The mixture was stirred vigorously in an Ar-filled glovebox for 96 h. Next, the intercalated compound was transferred to DI-H₂O and sonicated to yield exfoliated monolayers. Upon contact with the DI-H₂O, copious gas evolved at the interlayer interfaces, thus yielding high concentrations of monolayers and forming a black, opaque solution. After ultrasonication, the compound was washed over Whatman filter paper with 500 mL of hexane. 200 mL of DI-H₂O was then added to the mixture. With the filter paper removed, the mixture was then sonicated for 90 min. The resulting solution was then aliquoted into 45 mL centrifuge tubes with solution depths of 2-3 cm and centrifuged at 5200 rcf for 12 h, decanted, and then resuspended in 100 mL of DI-H₂O. After resuspension, the centrifugation procedure was repeated twice, followed by resuspension in 50 mL and, later, 25 mL of water. The resulting ce-MoS₂ was further diluted to 250 μg mL⁻¹ in a mixture of DI water and isopropyl alcohol (DI-H₂O: IPA=7:3, v/v) for subsequent EHD crumpling.

To prepare c-MoS₂ with the highest strain (≈3.7%), a hot plate was pre-annealed at 250° C. for 60 min prior to deposition. Next, aqueous solutions of ce-MoS₂ (250 μg mL⁻¹) were fed to a spinneret (gauge 18 TW needle) by a programmable syringe pump. An external electric field of 1.25 kV cm⁻¹ was generated with a high-power supply. Computerized multi-pass deposition onto pre-cleaned carbon cloth (CC) HER electrodes was achieved through the integration of an x-y translational stage at a linear stage speed of 2.5 mm s⁻¹. The flow rate and surface temperature were carefully maintained at 25 μL min⁻¹ and 250° C., respectively, to achieve the high yield of c-MoS₂. The amount of deposited c-MoS₂ catalyst was ≈13 μg cm⁻². Deposition yield was found to linearly scale with the concentration of ce-MoS₂ dispersions, flow rate, and duration of EHD process.

Scanning electron microscopy (SEM) and atomic force microscopy (AFM) images revealed largely separated and spatially distributed ce-MoS₂ sheets with average thickness in the range of nm when the threshold electric field reached 1.25 kV cm⁻¹ for the inception of the fission process. This result confirms that electrostatic stabilization prevents unwanted aggregation. Furthermore, ce-MoS₂ was found to spontaneously crumple into 3D porous, structurally deformed nanostructures with many wrinkles, ridges, and corrugated edges upon reaching both the threshold electric field of 1.25 kV cm⁻¹ and a deposition temperature of 250° C., as suggested in the TEM image shown in FIG. 3(a). HAADF and STEM images collectively showed that the resulting c-MoS₂ deposited at 250° C. comprises a predominant 2H phase because the deposition temperature exceeded the temperature for transformation of the metastable 1T′ phase to the 2H phase. See G. Eda et al., ACS Nano 6, 7311 (2012); and S. S. Chou et al., J. Am. Chem. Soc. 137, 1742 (2015). The impact of the EHD-assisted crumpling process on the structure of c-MoS₂ at the nanoscale was further probed using HRTEM. The HRTEM image in FIG. 3(b) pinpoints areas near edges. FIGS. 3(c) and 3(d) show areas on the basal plane. The edges of c-MoS₂ consist of only few layers, as determined by the clear steps in the relatively flat region, whereas propagating ridges along with sharp vertices, and facets can also be clearly identified, thus creating localized conformational strain with radii of curvature of nanometer scale.

Raman spectroscopy substantiates the c-MoS₂ structures by tracing the evolution of two distinct peaks at 385 and 407 cm⁻¹, assigned to the E¹_2g and A¹_g phonon modes of the 2H planar ce-MoS₂, respectively, as a function of processing temperatures, shown in FIG. 4. See S. S. Chou et al., J. Am. Chem. Soc. 137, 1742 (2015). At higher deposition temperatures, a much more pronounced out-of-plane Mo—S A¹_g phonon was observed, along with a significant red shift of the in-plane Mo—S E¹_2g phonon mode. In the former, the enhancement of the A¹_g mode is characteristic of exposed edges that preferentially align with the polarization of the exciting laser. See D. Kong et al., Nano Lett. 13, 1341 (2013). Additionally, the relatively small E¹_2g to A¹_g ratio (≈0.5) is an indication of both small domain size and textured morphology. This is not surprising since the crumpling process concurrently induces volumetric shrinkage of ce-MoS₂ and surface texturing of the resultant film. The E¹_2g phonon mode, however, did not shift toward a lower wavenumber relative to the planar 2H ce-MoS₂ even though the overall morphology appears to be heavily wrinkled at a deposition temperature of 100° C. Instead, a pronounced red shift of the E¹_2g mode toward lower wavenumbers only took place in the specimens synthesized at temperatures higher than 200° C. The shift of the E¹_2g phonon mode provides an index for quantifying surface strain load and has been shown to shift by ≈1.7 cm⁻¹ per % strain on the MoS₂. See A. Castellanos-Gomez et al., Nano Lett. 13, 5361 (2013). Here, c-MoS₂ synthesized at 100° C. did not yield a significant E¹_2g shift, and thus strain, whereas the E¹_2g peak of c-MoS₂ synthesized at 200° C. was red shifted by 2 cm⁻¹ (≈1.2% of strain). The E¹_2g phonon mode further redshifted for c-MoS₂ formed at 250° C., generating an average strain load of ≈3.7%. For reference, this strain is comparable to that reported by Zheng and co-workers for modulating S-deficient MoS₂ basal planes. See H. Li et al., Nat. Mater. 15, 48 (2016).

To better correlate morphology with the observed strain induced E¹_2g shift, the morphological evolution of c-MoS₂ deposited on carbon cloth (CC) was analyzed as a function of deposition temperature. As described above, ce-MoS₂ sheet precursors aggregated into discrete, rugged islands at the low deposition temperature of 100° C. Nevertheless, TEM image reveals that individual ce-MoS₂ sheets display a relatively flat and featureless morphology with slightly corrugated folds around the edges, as shown in FIG. 5(a). Comparatively, aggregated clumps begin to disintegrate into largely buckled structures with the emergence of thin, vertically protruding walls, meandering ridges, curled edges, and folds when processed at 200° C., as shown in FIG. 5(b), matching the onset of the red shift in the E¹_2g phonon mode and the corresponding moderate strain load. At 250° C., almost all the ce-MoS₂ sheets were transformed and compressed into 3D structurally deformed nanostructures with petal-like, curved morphology, as shown in FIG. 5(c), and thus the high magnitude of strain, an important feature necessary for further modulating the ΔG_H. See J. D. Benck et al., ACS Catal. 4, 3957 (2014); and H. Li et al., Nat. Mater. 15, 48 (2016). In particular, the hierarchically crumpled MoS₂ at 250° C. reveals a highly strained c-MoS₂ that comprises corrugated edges and extremely thin layers, narrow folds converged to form propagating ridges, and sharp vertices and facets with radii of curvatures at nanoscale. Rich and spatially distributed defects are throughout.

In parallel, the chemical states of Mo and S were investigated by X-ray photoelectron spectroscopy (XPS) on c-MoS₂ synthesized at different temperatures, as shown in FIG. 6. Deconvolution of the Mo 3d core level region from c-MoS₂ synthesized at 100° C. showed additional peaks at 228.1 eV (Mo 3d_5/2) and 231.3 eV (Mo 3d_3/2), relative to pure 2H MoS₂, which exhibited peaks at 229.1 eV (Mo 3d_5/2) and 232.2 eV (Mo 3d_3/2). The presence of these additional core level photoemission peaks indicates the presence of the 1T′ phase and integrated peak area ratios reveal a 2H phase fraction of 76% for c-MoS₂ synthesized at 100° C. See S. S. Chou et al., J. Am. Chem. Soc. 137, 1742 (2015); and J. Heising and M. G. Kanatzidis, J. Am. Chem. Soc. 121, 11720 (1999). Images obtained through HRTEM, along with false-colored HAADF, confirm the coexistence of both phases at a deposition temperature lower than that of phase transformation (100° C.), which is consistent with the spectroscopic interpretation. In contrast, a phase transition to pure 2H material was observed when c-MoS₂ was processed at 200° C. The predominant 2H to 1T′ ratio remained constant at 250° C., demonstrating stable 2H crumpling. As shown in the right-hand panels in FIG. 6, complementary analysis of S 2p core-level spectra corroborate these results.

These c-MoS₂ films maintained mesoporosity, as indicated by Brunauer-Emmett-Teller (BET) isotherms (surface area of 310.8+/−0.5 m² g⁻¹). This is notable because the combination of electrostatic stabilization and subsequent fission effectively suppresses the formation of large clumps, enabling 3D integration with HER accessible surfaces and micropores. Moreover, in a similar fashion to a paper ball, the deformed ridges made of several overlapping folds form structural supports that prevent the c-MoS₂ from unfolding or collapsing when subjected to external stimuli. Indeed, structurally robust c-MoS₂ can sustain common material processing conditions, such as solution processing, high temperature annealing, or the combination of both, and still maintain a porous, nanotextured morphology necessary for preserving efficient transport of H⁺ and H₂ to the edge sites and activated basal plane under continuous electrochemical operation or elevated temperate. Together with the strained S vacancies, these structural features, revealed by the Raman, BET, SEM, and TEM characterizations described above, indicate that c-MoS₂ can be used as a catalytically active, electrochemically and thermally stable component for HER.

The use of an EHD process offers the distinct advantage to directly integrate c-MoS₂ into diverse device architectures, as well as onto a variety of substrates, including thin films of carbon nanotubes (CNTs), silicon, and even nanostructured polystyrene, in both monolithic and a selective fashion. To evaluate the catalytic performance of c-MoS₂ for HER, electrodes were formed by depositing thin films of c-MoS₂ onto CC via EHD deposition, as described above. The substrate had an area of 1 cm² and the total aerial mass loading was ≈13 μg cm⁻². In FIGS. 7(a)-(b), it can be seen that densely packed c-MoS₂ assemblies are uniformly distributed throughout the entire carbon clothes (CC). FIG. 7(c) further suggests that mesoporous c-MoS₂ structures remains intact and retains a high degree of crumpling even upon continuous deposition. In addition, corresponding energy-dispersive X-ray (EDX) mapping of relevant elements (not shown), including Mo and S, closely overlap the contours and areas of c-MoS₂ and underlying CC substrates.

The HER cathodes made of c-MoS₂ supported on CC were examined using a standard three-electrode electrochemical configuration in 0.5 m H₂SO₄ electrolyte de-aerated with Ar. Following deposition and prior to measurement, all samples were annealed at 250° C. for 1 h to ensure complete conversion of remnant 1T′ phases to 2H phases. This provides a well-defined system to unambiguously explore the correlation between the impact of 3D nanostructures, strain-induced modulation, and the catalytic activity in phase pure c-MoS₂. To this end, linear sweep voltammograms (LSVs) were measured from thin films of planar 2H ce-MoS₂ standard (S vacancies, strain: 0%, and heavily aggregated morphology), c-MoS₂ synthesized at 100° C. (S vacancies, strain: 0%, and rippled morphology), c-MoS₂ synthesized at 200° C. (S vacancies, strain: 1.2%, and partially crumpled morphology), and c-MoS₂ synthesized at 250° C. (S vacancies, strain: 3.7%, and 3D fully crumpled morphology), along with the bare CC support as a reference, as shown in FIG. 8. CC electrodes did not show any pronounced cathodic current, and remained inactive during LSV scans. The planar 2H ce-MoS₂ standard electrodes showed very poor HER characteristics even with a high density of S vacancies that have shown to activate the basal plane, delivering an exceedingly high η₁₀ of ≈−364 mV. This poor HER performance can be attributed to the lack of electronic coupling between the substrates and active sites in combination with a heavily aggregated morphology that limits the exposure of activated basal plane toward HER. See H. Li et al., Nat. Mater. 15, 48 (2016). Meanwhile, increasing the deposition temperature to 100° C. led to the formation of a rippled morphology and η₁₀ was reduced to −340 mV. The η₁₀ of the c-MoS₂ synthesized at 200° C. was significantly reduced to −309 mV due to the presence of wrinkled morphology and moderate strain load (≈1.2%). Importantly, c-MoS₂ synthesized at 250° C., which was characterized by a combination of high strain load (≈3.7%) and 3D, fully crumpled nanostructures, displayed substantially improved total electrode metrics, including η₁₀ of −191 mV, exchange current densities of 10.4 μA cm⁻², and high cathodic current densities.

Further insights into function were obtained by extracting slopes from the Tafel plots shown in FIG. 9. Tafel slopes were found to decrease with increasing magnitude of crumpling, and were correlated with the strain load. The lowest Tafel slope of 64 mV per dec was obtained for the c-MoS₂ synthesized at 250° C., suggesting a Heyrovsky mechanism at the surface of c-MoS₂ with the hydrogen desorption reaction as the limiting step: H_ads+H₃O⁺+e⁻→H₂+H₂O. See J. D. Benck et al., ACS Catal. 4, 3957 (2014); M. A. Lukowski et al., J. Am. Chem. Soc. 135, 10274 (2013); B. E. Conway and B. V. Tilak, Electrochim. Acta 47, 3571 (2002); and C. Tsai et al., Surf. Sci. 640, 133 (2015). Note that total electrode activities are typically commensurate with the catalyst loading, as shown in Table 1. See J. D. Benck et al., ACS Catal. 4, 3957 (2014). Thus, from a practical perspective, the ability to deliver compelling total electrode characteristics on a HER electrode with very low loadings of c-MoS₂ catalysts per geometric area (≈13 μg cm⁻²) may be transformative and economically viable.

The active site numbers were further estimated using the underpotential deposition (UPD) method. See A. Y. Lu et al., Small 12, 5530 (2016); and C. L. Green and A. Kucernak, J. Phys. Chem. 2, 11446 (2002). The density of active sites was determined to be approximately 1×10¹⁴ sites cm⁻² for planar 2H ce-MoS₂, increased to 5.7×10¹⁶ sites cm⁻² for planar 1T′ ce-MoS₂ polymorph, and further increased to 5.86×10¹⁷ sites cm⁻² for 2H c-MoS₂, thus corroborating that the advent of the hierarchically structured, 3D morphology of c-MoS₂ from the intrinsically planar morphology of ce-MoS₂ and enabling infiltration of electrolyte for exposure to both activated basal plane and edges. See J. Kibsgaard et al., Nat. Mater. 11, 963 (2012); D. Kong et al., Nano Lett. 13, 1341 (2013); and H. Wang et al., ACS Nano 8, 4940 (2014). The effective electrochemically active surface area, deduced from the electric double-layer capacitance (EDLC), increased by 12× when the c-MoS₂ was deposited at 100° C., 24× when the c-MoS₂ was deposited at 200° C., and by 125x when deposited at 250° C. relative to that of planar 2H ce-MoS₂ annealed at 250° C. Meanwhile, electrochemical impedance spectroscopy (EIS) also suggested that the 3D porous morphology facilitates efficient transport across the electrode/electrolyte interfaces as indicated from the substantially reduced R_ct (13Ω of 2H c-MoS₂) relative to that of the planar 2H ce-MoS₂ (R_ct=275Ω, Table 1). Additionally, the measured R_s was significantly reduced from 38Ω (planar 2H ce-MoS₂) to 1.5Ω (2H c-MoS₂). This is unexpected, since the 2H c-MoS₂ does not comprise locally transformed phases and atomically sharp interfaces to lower the contact resistance. See D. Voiry et al., Nat. Mater. 15, 1003 (2016); and D. Voiry et al., Nano Lett. 13, 6222 (2013). Even with the much enhanced electrochemically active surface area, the electronic coupling between the substrate and active sites will be sluggish unless selectively forming the electrically addressable interfaces through heterogeneous doping. Alternatively, both experimental demonstration and theoretical prediction have confirmed that regions of atomically localized curvature lead to a loss of spatial overlap of the d orbitals with neighboring Mo atoms and a shift in hybridization of Mo—S bonding from mixed ionic-covalent bonding between Mo and S atoms. See L. Yu et al., Nano Lett. 16, 2444 (2016); and M. Mavrikakis et al., Phys. Rev. Lett. 81, 2819 (1998). Notably, under applied strain, depletion of charges residing on Mo atoms is increased, thus leading to excess charges redistributed around S atoms. The net result of these orbital effects is a drastic increase in energy locally and reduced bonding-antibonding splitting. Consequently, the bandgap decreases, with the valence band maximum (VBM) and conductive band minimum (CBM) crossing the Fermi level, leading to a reduction of overall R_s similar to the semiconductor-to-metal transition and thus improved electronic coupling. See A. P. Nayak et al., Nat. Commun. 5, 1 (2014); and H. Li et al., J. Am. Chem. Soc. 138, 5123 (2016).

In parallel, using the measured active sites and exchange current densities, the TOF was calculated to infer the intrinsic catalytic activity per active sites. TOF was determined to be ≈0.130 s⁻¹ at −200 mV vs. RHE for 2H c-MoS₂ synthesized at 250° C., an order of magnitude higher than the 2H planar counterpart (0.016 s⁻¹). These results suggest that the improvement of the HER performance comes from the combined effects of dimensional transition. First, the 3D porous morphology not only effectively exposes the initially buried active sites on both edges and basal plane but also facilitates effective transport across various interfaces between 2H c-MoS₂, electrolyte and current collecting CC substrate. Second, the high strain load modulates the ΔG_H and tailors the TOF at each active site. Specifically, the TOF achieved is comparable to or better than values measured for strain-vacancy engineered MoS₂, edge sites of MoS₂, vertical MoS₂ flakes, 1T′ MoS₂ nanoparticles, and [Mo₃S₄]⁴⁺ cluster of amorphous MoS₂. See H. Li et al., Nat. Mater. 15, 48 (2016); T. F. Jaramillo et al., Science 317, 100 (2017); D. Kong et al., Nano Lett. 13, 1341 (2013); H. Wang et al., ACS Nano 8, 4940 (2014); and T. F. Jaramillo et al., J. Phys. Chem. C 112, 17492 (2008). Although not quite as active as the molecular thiomolybdate [Mo₃S₁₃]²⁻, c-MoS₂ has advantages to synergistically tailor both intrinsic and total electrode activities in tandem with readily scalable synthesis, assembly, and processing routes. See J. Kibsgaard et al., Nat. Chem. 6, 248 (2014). Meanwhile, the loading of c-MoS₂ was increased in an attempt to further improve the HER metrics through increasing numbers of active sites. However, HER metrics drastically decreased such that the thickest film in fact exhibited the poorest performance. From the comparison of all three specimens of different thicknesses, it is clear that the major limiting factors are the high η₁₀ and low current densities. The high resistance may arise from the discontinuous pathways where outer layers of c-MoS₂ are loosely attached to each other, as suggested by EDX mapping. The limited contacts at the top of the films serve as constriction points, thus diminishing electron accessibility and adversely affecting the overall HER performance. This limitation can be potentially mediated by adding conductive adducts, such as graphene, CNTs, or even 1T′ ce-MoS₂, to simultaneously minimize charge transport distance and series resistance, while maintaining the accessible surface area.

Another important aspect for practical implementation of MoS₂-based HER catalyst is the long-term stability under continuous operation and elevated temperatures. A durability study on the c-MoS₂-based electrodes was first carried out by employing continuous LSV cycles at a scan rate of 5 mV s⁻¹. The HER catalytic activity remains almost unchanged in the first 2500 cycles and displays a negligible decay after 5000 cycles. See M. A. Lukowski et al., J. Am. Chem. Soc. 135, 10274 (2013); and Y. Yin et al., J. Am. Chem. Soc. 138, 7965 (2016). In parallel, the electrochemical stability of the c-MoS₂ catalyst was also evaluated by prolonged electrolysis at constant potentials. In accordance with the durability tests, the current density of c-MoS₂ remained stable in 0.5 m H₂SO₄ for more than 100 h and showed a negligible change after 110 h. The uniformity and crumpled nanostructure of c-MoS₂ was also examined under SEM and TEM after continuous linear potential sweeps, and appeared to be intact after the degradation measurements. In addition, when subjecting to thermal annealing at 300° C. under ambient conditions, unlike the temperature sensitive 1T′ counterparts, 2H c-MoS₂, with in situ strain fields propagating on the thermodynamically favored 2H phase, remained catalytically active, revealing an unparalleled operational window up to 300° C. When the thermal annealing temperature was further increased to 350° C., a transition from MoS₂ to MoO₃ was observed as confirmed by XPS spectra. Nevertheless, the unprecedented combination of electrochemical and thermal stabilities not only suggests the preservation of both catalytic activity and structural integrity of c-MoS₂, consistent with the structural robustness of c-MoS₂, but also renders c-MoS₂ more versatile for long-term utilization in energy conversion devices. Meanwhile, the chemically and mechanically robust nature of c-MoS₂ provides flexibility in using chemical and mechanical means to further improve catalytic properties for HER. See Z. W. She et al., Science 355, 146 (2017).

In summary, the invention enables the formation of 3D structurally deformed c-MoS₂ catalysts for electrochemically stable, thermally resistant, and highly efficient HER. The use of electrostatically charged droplets as nanoreactors facilitates colloidal dispersion, fission, and capillarity-induced-self-crumpling of 2D ce-MoS₂, ultimately transforming it into 3D c-MoS₂. As summarized in Table 1, electrochemical characterization indicates that HER electrodes composed of c-MoS₂ catalysts synergistically integrate advantageous features of morphologically, chemically, electronically, and mechanically engineered MoS₂ catalysts. See J. Kibsgaard et al., Nat. Mater. 11, 963 (2012); D. Kong et al., Nano Lett. 13, 1341 (2013); H. Li et al., Nat. Mater. 15, 48 (2016); D. Voiry et al., Nat. Mater. 15, 1003 (2016); H. Wang et al., ACS Nano 8, 4940 (2014); and Y. Yin et al., J. Am. Chem. Soc. 138, 7965 (2016). These include high surface area, efficient transport at interfaces, reduced series resistance, exposed basal plane with increased density of active sites, improved TOF per active site, chemical and mechanical robustness. The invention simultaneously achieves state-of-art metrics while addressing long-term stability under continuous operation as well as exposure to high temperature treatment while using a fully scalable and easy to integrate process. Given the wide variety and availability of chemically exfoliated 2D functional materials, including graphene, transition metal dichalcogenide (TMD), and the emerging MXenes, structurally robust, thermally stable, electronically heterogeneous, and catalytically active, multifunctional hybrid nanocomposites can now be readily, rapidly, and rationally assembled through this template free and scalable nanomanufacturing route. See J. Luo et al., ACS Nano 9, 9432 (2015). This enables applications beyond HER, including electrochemical capacitors, battery, catalysis, reactors and separation, drug delivery, biocompatible scaffolds, sensing and high complexity composites.

The present invention has been described as method to form crumpled transition metal dichalcogenide sheets. It will be understood that the above description is merely illustrative of the applications of the principles of the present invention, the scope of which is to be determined by the claims viewed in light of the specification. Other variants and modifications of the invention will be apparent to those of skill in the art.

TABLE 1
Summary of HER characteristics of MoS2 catalysts with various nanostructures, strain S vacancy, edge and 1T′ phase engineering.
η10, j0, Rs and Rct stand for overpotentials, exchange current densities, series and charge-transfer resistances, respectively.
Total electrode properties	Intrinsic properties
	Tafel		Surface	Mass	Active
	η10	slope	J0	Rs + Rct	Capacitance	area	loading	sites	TOF	Stability
	(mV)	(mV/dec)	(μA/cm2)	(Ω)	(mF/cm2)	(m2/g)	(μg/cm2)	(sites/cm2)	(s−1)	Cycles	Temp
2H c-MoS2 (300° C.)	−193	66	10.4	Rs = 1.7	11.0	310.5	13	5.7 × 1017	0.128	5000	300° C.
				Rct = 13
2H c-MoS2 (250° C.)	−191	64	10.4	Rs = 1.5	11.4	310.8	13	5.8 × 1017	0.130	5000	300° C.
				Rct = 13
2H c-MoS2 (200° C.)	−309	92	5.2	Rs = 17	2.2	48.6	13	2.0 × 1015	0.080	2000	300° C.
				Rct = 50
2H c-MoS2 (100° C.)	−340	107	4.5	Rs = 29	1.1	37.2	13	4.7 × 1014	0.052	1000	N/A
				Rct = 150
2H planar MoS2	−364	108	0.7	Rs = 38	0.091	19.0	15	1.0 × 1014	0.016	<1000	N/A
				Rct = 275
*2H porous MoS2	−218	62	10.5	Rct = 280	8.2	N/A	140	2.3 × 1014	N/A	1000	N/A
**Vertically aligned 2H	−400	105-120	2.2	Rs = 2.3	N/A	N/A	8.5	5.4 × 1014	0.013-0.016	1000	N/A
MoS2				Rct > 20
***Double gyroid MoS2	−206	50	0.69	N/A	4.8	N/A	60	N/A	N/A	N/A	N/A
****Strained 2H-MoS2	−170	60	50-60	N/A	N/A	N/A	N/A	3.1 × 1014	0.08-0.31	N/A	N/A
*1T′ Porous MoS2	−153	43	15.8	Rct = 16	63.1	N/A	140	5.8 × 1016	0.5	1000	N/A
nanosheets									(η = 15
									3 mV)
*****1T′ MoS2	−118	66	50	Rs = 18	345	N/A	3,400	6.4 × 1018	0.10	>5000	N/A
nanoparticle				Rct > 5
******1T′ MoSe2	−152	52	N/A	Rct = 16	25.2	N/A	140	N/A	N/A	1000	N/A
*******1T′ WS2	−142	70	N/A	Rct = 5	48	N/A	1000	N/A	N/A	<500	N/A
*Y. Yin et al., J. Am. Chem. Soc. 138, 7965 (2016);
**D. S. Kong et al., Nano Lett. 13, 1341 (2013);
***J. Kibsgaard et al., Nat. Mater. 11, 963 (2012);
****H. Li et al., Nat. Mater. 15, 48 (2016);
*****H. Wang et al., Acs Nano 8, 4940 (2014);
******Y. Yin et al., Adv. Mater. 29, 1700311 (2017);
*******M. a Lukowski et al., Energy Environ. Sci. 7, 2608 (2014).

Claims

1. A method of crumpling transition metal dichalcogenide sheets, comprising:

forming an electrostatically charged liquid droplet comprising a plurality of planar sheets of a transition metal dichalcogenide;

applying a sufficient electric field to the droplet to cause the droplet to rupture into a plurality of fission droplets by an electrohydrodynamic process, wherein each fission droplet comprises at least one planar sheet of the transition metal dichalcogenide; and

evaporating the remaining liquid from the plurality of fission droplets, thereby inducing spontaneous self-crumpling of the at least one planar sheet of the transition metal dichalcogenide in each of the fission droplets.

2. The method of claim 1, wherein the transition metal comprises molybdenum or tungsten.

3. The method of claim 1, wherein the chalcogenide comprises sulfur, selenium, or tellurium.

4. The method of claim 1, wherein the transition metal dichalcogenide comprises molybdenum disulfide.

5. The method of claim 4, wherein the planar sheets of molybdenum disulfide comprise 2H phase sheets having basal plane defects and whereby the self-crumpling activates the basal plane defects.

6. The method of claim 1, wherein the planar sheets of the transition metal dichalcogenide comprise exfoliated sheets.

7. The method of claim 1, wherein the forming an electrostatically charged liquid droplet comprises feeding a solution of the planar sheets of the transition metal dichalcogenide through a spinneret.

8. The method of claim 1, wherein the sufficient electric field is greater than approximately 1.25 kV/cm.

9. The method of claim 1, wherein the evaporating comprises depositing the plurality of fission droplets on a substrate above a deposition temperature.

10. The method of claim 9, wherein the substrate comprises a hydrogen evolution reaction substrate.

11. The method of claim 9, wherein the substrate comprises carbon or silicon.

12. The method of claim 9, wherein the deposition temperature is greater than 200° C.

13. The method of claim 9, wherein the deposition temperature is greater than 250° C.

14. A method for electrochemical hydrogen evolution, comprising exposing an aqueous electrolyte to a crumpled transition metal dichalcogenide in an electrochemical cell at a sufficient cell voltage to catalyze the electrolysis of water and evolve hydrogen.