CATALYSTS FOR HYDROGEN EVOLUTION REACTION INCLUDING TRANSITION METAL CHALCOGENIDE FILMS AND METHODS OF FORMING THE SAME

TECHNICAL FIELD

The present invention generally relates to catalysts for hydrogen evolution reaction and methods of forming and using the same.

BACKGROUND

The hydrogen evolution from water stands as an appealing strategy for energy storage. It may store electrical or solar energy in format of chemical fuels (hydrogen) that can be delivered at will and consumed with negligible impact to environment. However, the implementation of this energy storage strategy has been delayed by the imperfection of the catalysts that are required to drive the reaction. Ideal catalysts would feature high catalytic activities and low cost. Noble metals such as Pt can provide excellent catalytic activities for the hydrogen evolution reaction (HER), but are too expensive and scarce to be useful for massive application. Molybdenum disulfide (MoS₂) has been considered to be a promising low-cost alternative (Laursen, et al., Energy Environ Sci 2012, 5, 5577; Voiry, et al., Nat Mater 2013, 12, 850; Merki, et al., Energy Environ Sci 2011, 4, 3878). Molybdenum disulfide (MoS₂) bears particular implications for the storage of solar energy due to its capability to efficiently absorb solar radiation and fast interfacial charge transfer (Huang, et al., ACS Nano 2016; Cao, MRS Bulletin 2015, 40, 592; Yu, et al., Nano Letters 2014, 15, 486). However, the catalytic activity of MoS₂is inferior to that of Pt, and thus improvement of the activity of MoS₂may be beneficial.

SUMMARY

According to some embodiments of the present invention, catalysts for hydrogen evolution reaction are provided. The catalysts may include a metal chalcogenide film including chalcogen atom vacancies, and a density of the chalcogen atom vacancies may be from about 5% to about 15%. The metal chalcogenide film may be a monolayer film or a film including less than 10 layers.

According to some embodiments of the present invention, catalysts for hydrogen evolution reaction may include a substrate including nickel, titanium, silver, zinc, and/or platinum and a metal chalcogenide film extending on the substrate.

According to some embodiments of the present invention, catalysts for hydrogen evolution reaction may include a substrate and a metal chalcogenide film extending on the substrate. The metal chalcogenide film may include a first surface facing the substrate and a second surface opposite the first surface. The catalysts may further include hydrogen ions disposed on the first surface of the metal chalcogenide film or on the second surface of the metal chalcogenide film.

According to some embodiments of the present invention, catalysts for hydrogen evolution reaction may include a metal chalcogenide film including dopants. The metal chalcogenide film may be a monolayer film or a film including less than 10 layers, and the dopants may include nickel atoms and/or cobalt atoms.

According to some embodiments of the present invention, methods of forming a catalyst for hydrogen evolution reaction are provided. The methods may include providing a metal chalcogenide film on a substrate. The metal chalcogenide film may include a first surface facing the substrate and a second surface opposite the first surface. The methods may further include disposing hydrogen ions on the first surface of the metal chalcogenide film or on the second surface of the metal chalcogenide film.

According to some embodiments of the present invention, methods of forming a catalyst for hydrogen evolution reaction may include forming a metal chalcogenide monolayer film including dopants by performing a chemical vapor deposition (CVD) process using a first precursor including metal atoms of the metal chalcogenide monolayer film, a second precursor including chalcogen atoms of the metal chalcogenide monolayer film, and a third precursor including dopant atoms of the metal chalcogenide monolayer film.

According to some embodiments of the present invention, methods of generating hydrogen from water may include contacting the water with the catalysts according to some embodiments of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS

Further aspects of the present invention will be readily appreciated upon review of the detailed description of its various embodiments, described below, when taken in conjunction with the accompanying drawings.

FIG. 1A is an optical image of an as-grown monolayer MoS₂film. A scratch is intentionally introduced to show the contrast between the film and the substrate (sapphire). The inset is an optical image of a film transferred onto a glassy carbon substrate in size of around 1×1 cm. FIG. 1B is an optical image of discrete monolayer MoS₂flakes. FIG. 1C is a graph of the magnetic measurement results of monolayer MoS₂films (upper) and flakes (lower). The straight line indicates the diamagnetic moment of the substrate, and the curved center represents the ferromagnetic moment of the flakes. FIG. 1D shows polarization curves of monolayer MoS₂films and flakes. The polarization curves of glass carbon substrates and Pt are also given as shown. The inset depicts stability test results of monolayer MoS₂films. The initial result represents the result collected from the film when its performance appears to be stable after pre-testing cycles (see FIG. 8). FIG. 1E shows Tafel plots derived from the results given in FIG. 1D. The dashed lines serve to illustrate the Tafel slope and the current density at 0V overpotential.

FIGS. 2A and 2B show structural and compositional characterizations of monolayer MoS₂films and flakes. FIG. 2A shows Raman spectra of monolayer MoS₂film and flakes before and after the catalytic reaction. The spectra in low wavenumbers are included to indicate no characteristic peaks of 1 T MoS₂. FIG. 2B shows XPS results of as-grown MoS₂film and flakes.

FIG. 3A is a schematic illustration of a process of repairing sulfur vacancies in MoS₂. FIG. 3B depicts the XPS results of the film before (solid) and after (dashed) the repair of sulfur vacancies. FIG. 3C shows the polarization curves of the MoS₂film before and after the repair of sulfur vacancies. The polarization curve of the flake-merged MoS₂film with little sulfur vacancies is also given (dashed). FIG. 3D shows polarization curves of the MoS₂flakes before and after the repair of sulfur vacancies.

FIGS. 4A-4D show quantitative evaluation for the catalytic activities of each active site. FIG. 4A shows XPS results for monolayer MoS₂films with different densities of sulfur vacancies. FIG. 4B shows polarization curve of the MoS₂films with different densities of sulfur vacancies. FIG. 4C shows exchange current densities (upper) and Tafel slopes (lower) of the films as a function of the density of sulfur vacancies. The dashed lines serve to guide the vision. FIG. 4D shows exchange current densities (upper) and Tafel slopes (lower) of the MoS₂flakes as a function of the edge length per unit area. The dashed lines serve to guide the vision.

FIGS. 5A-5D show images of catalytic active sites with the electrochemical deposition of Cu nanoparticles. FIG. 5A is an SEM image of the Cu particles deposited on MoS₂films. Inset is an image for the Cu deposition on the MoS₂film, which is pre-treated with the repair of sulfur vacancies. The scale bar is 10 SEM images of the Cu particles deposited on MoS₂flakes (FIG. 5B), MoS₂flakes with mirror boundaries (FIG. 5C), and MoS₂flakes with tilt boundaries (FIG. 5D) are shown. The tilt boundaries are marked with dashed ovals. The dashed lines in FIG. 5D mark the edges of the flake. The insets in FIG. 5C and FIG. 5D schematically illustrate the mirror and tilt boundaries.

FIG. 6 shows typical EIS measurement results collected from MoS₂films.

FIGS. 7A-7D show AFM characterizations of monolayer MoS₂films and flakes. FIG. 7A is an AFM image of a typical as-grown triangle flake on sapphire substrates. FIG. 7B is a height profile of the white dash line shown in FIG. 7A. FIG. 7C is an AFM image of a typical as-grown monolayer MoS₂film on sapphire substrates. FIG. 7D is a height profile of the white dash line shown in FIG. 7C.

FIG. 8 shows pre-test cycling of monolayer MoS₂film and flakes. Cyclic Voltammetry (CV) were performed at the film and flakes transferred onto glassy carbon in the range of 0- −0.5 (vs. RHE) till the catalytic performance appears to be stable. The polarization curves were collected from a typical MoS₂film (upper) and typical MoS₂flakes (lower) with different cycles as indicated.

FIGS. 9A and 9B show XPS results of monolayer MoS₂film and flakes before and after the catalytic reactions. The peaks of Mo and S are labeled as shown.

FIG. 10 shows Raman spectra of as-grown monolayer MoS₂film and flakes. The A_1gpeak at 405 cm⁻¹and the E¹_2gpeak at 384.8 cm^−1.The intensity is normalized to the intensity of the Raman peak of sapphire substrates at 420 cm⁻¹.

FIG. 11 shows XPS results of monolayer MoS₂flakes before and after the treatment of sulfur vacancy repair.

FIG. 12 shows Tafel plots of MoS₂films. These Tafel plots are derived from the polarization curves given in FIG. 3C.

FIG. 13 is a graph of current density as a function of the density of grain boundaries. The exchange current densities are extracted from the polarization curves measured at the defect-repaired films and the flake-merged films.

FIG. 14 shows measured XPS results and corresponding fitting for monolayer MoS₂film with different densities of sulfur vacancies (4%, 7%, 2%, 10%, 12% and 14% from the top plane).

FIG. 15 shows Raman spectra of monolayer MoS₂films with different densities of sulfur vacancies. The density of sulfur vacancies is labeled as shown.

FIGS. 16A-16D show dependence of the catalytic activity of sulfur vacancies on crystalline quality. FIG. 16A shows Raman spectra, FIG. 16B shows PL spectra, FIG. 16C shows XPS results, and FIG. 16D shows polarization curves of two monolayer MoS₂films. The two films show reasonably comparable densities of sulfur vacancies as indicated by the XPS results, but have different amount of grain boundaries as evidenced by the Raman and PL measurement.

FIGS. 17A and 17B show structural and compositional characterization of monolayer MoS₂films. FIG. 17A shows Raman spectra of the film before pre-testing cycling, after pre-testing cycling, and after an treatment of Ar plasma for 2 s. The arrows are associated with defects created by the Ar plasma treatment. FIG. 17B shows XPS measurement of the film before pre-testing cycling, after pre-testing cycling, and after an treatment of Ar plasma for 2 s.

FIGS. 18A and 18B show the effect of Ar plasma treatment on the catalytic activities. FIG. 18A shows polarization curves of the monolayer MoS₂films at different stage: initial, after being cycled (in the range of 0- −0.5 V vs. RHE) for 5000 times, treated by Ar plasma for 2 s, and after treated by the process of repairing sulfur vacancies. FIG. 18B shows polarization curves of the monolayer MoS₂films at different stage: initial, treated by Ar plasma for 2 s, and after cycling (in the range of 0- −0.5 for 5000 times) for 5000 cycles.

FIGS. 19A-19C show the effect of the treatment time on the catalytic activities. FIG. 19A shows Raman spectra of the films treated by Ar plasma for 2 s, 10 s, and 15 s. FIG. 19B shows polarization curves of the monolayer MoS₂films before and after the Ar plasma treatment. The repair of sulfur vacancies also was performed at the film for 15 s. The polarization curve after the sulfur vacancy repair is also given. FIG. 19C shows SEM images of the films treated by Ar plasma for 2 s (top), 10 s (middle), and 15 s (bottom). The black areas are holes with MoS₂materials missing.

FIGS. 20A-20E show measurements during the activation process. FIG. 20A shows polarization curves, and FIG. 20B shows Tafel plots during the activation process. FIG. 20C shows PL, FIG. 20D shows XPS, and FIG. 20E shows Raman spectrum during before and after activation process.

FIGS. 21A-21E show electrochemical catalytic performance, PL, XPS and Raman spectrum change with TFSI treatment. FIG. 21A shows polarization curves, and FIG. 21B shows Tafel plots of different TFSI treatment times. FIG. 21C shows PL, FIG. 21D shows XPS and FIG. 21E shows Raman spectrum with increasing TFSI treatment times.

FIGS. 22A-22D show the electrochemical catalytic performance, PL and Raman spectrum of bilayer MoS₂during TFSI treatment and electrochemical cycling. FIG. 22A shows polarization curves of bilayer MoS₂during TFSI and electrochemical cycling. FIG. 22B shows PL spectrum changes during TFSI and electrochemical cycling. FIG. 22C shows Raman spectrum changes in large scale including E_2gand A_1gmodes during TFSI and electrochemical cycling. FIG. 22D shows Raman spectrum changes in small scale only including A_1gmode shift during TFSI and electrochemical cycling.

FIG. 23A shows polarization curves during the activation process of MoS₂nanoflakes for hydrogen evolution. FIG. 23B shows SEM images before and after electrochemical cycling showing no change on the morphology and structures of MoS₂nanoflakes. FIG. 23C shows Raman spectrum before and after the activation process showing no change on the crystalline quality and phase transition. FIG. 23D shows XRD spectrum before and after electrochemical cycling.

FIGS. 24A and 24B show activated MoS₂monolayer film usage in neutral and base solutions. FIG. 24A shows polarization curves in neutral and base solutions using monolayer without and with electrochemical activation in acid. FIG. 24B shows stability test of activated monolayer MoS₂in base and neutral solution.

FIG. 25A shows polarization curves of thick MoS₂nanosheets as a function of electrochemical cycling in acidic electrolyte. The inset shows a SEM image for a typical MoS₂nanosheets. FIG. 25B shows Raman spectra of the thick MoS₂nanosheets before and after the electrochemical cycling. FIG. 25C shows XRD spectra of the thick MoS₂nanosheets before and after the electrochemical cycling.

FIG. 26A shows the polarization curves of monolayer MoS₂on different substrates including nickel, gold, glassy carbon (GC), and copper for hydrogen evolution, and FIG. 26B shows the Tafel plots of monolayer MoS₂on different substrates including nickel, gold, glassy carbon, and copper for hydrogen evolution.

FIG. 27 shows the polarization curves collected from the monolayer MoS₂film on Pt substrates under pH values of 1, 7 and 14.

FIG. 28 shows Tafel plots corresponding to the polarization curves of FIG. 27.

FIG. 29 shows polarization curves of the MoS₂films with different surfer vacancy densities on Pt substrates.

FIG. 30A shows polarization curves of the MoS₂films having different numbers of layers on Pt substrates. FIG. 30B shows exchange current densities of the MoS₂films having different numbers of layers on Pt substrates and on glassy carbon (GC) substrates. FIG. 30C shows Tafel slopes of the MoS₂films having different numbers of layers on Pt substrates and on glassy carbon (GC) substrates.

FIGS. 31A and 31B show polarization curves of the MoS₂films having different numbers of layers on Pt substrates and on glassy carbon (GC) substrates, respectively.

FIG. 32A shows Raman spectra of monolayer MoS₂on a Pt substrate and on a glassy carbon substrate. FIG. 32B shows XPS spectra of the monolayer MoS₂on a Pt substrate and on a glassy carbon substrate.

FIG. 33A shows polarization curves collected from monolayer MoS₂films on thin Pt layers deposited on nickel substrates in acidic media (pH=1). FIG. 33B shows stable catalytic performance of monolayer MoS₂films on Pt substrates with continuous reaction under with a density of >20 mA/cm²for more than two months in acidic media (pH=1).

FIG. 35A shows catalytic activity according to Ni precursor amount used in synthesis. FIG. 35B catalytic performance of a bare pt plate and Ni doped MoS₂on a Ni substrate.

FIGS. 36A, 36B, 37A, 37B and 38 show schematic diagrams of a catalyst according to some embodiments of the present invention.

DETAILED DESCRIPTION

Before the present invention is described in greater detail, it is to be understood that this invention is not limited to particular embodiments described herein, and as such may, of course, vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting.

All publications and patents cited in this specification are herein incorporated by reference in a matter consistent with the present application as if each individual publication or patent were specifically and individually indicated to be incorporated by reference and are incorporated herein by reference to disclose and describe the methods and/or materials in connection with which the publications are cited. The citation of any publication is for its invention prior to the filing date and should not be construed as an admission that the present invention is not entitled to antedate such publication by virtue of prior disclosure. Further, the dates of publication provided could be different from the actual publication dates that may need to be independently confirmed.

Functions or constructions well-known in the art may not be described in detail for brevity and/or clarity. Embodiments of the present invention will employ, unless otherwise indicated, techniques of nanotechnology, organic chemistry, material science and engineering and the like, which are within the skill of the art. Such techniques are explained fully in the literature.

It should be noted that ratios, concentrations, amounts, and other numerical data can be expressed herein in a range format. It is to be understood that such a range format is used for convenience and brevity, and thus, should be interpreted in a flexible manner to include not only the numerical values explicitly recited as the limits of the range, but also to include all the individual numerical values or sub-ranges encompassed within that range as if each numerical value and sub-range is explicitly recited. To illustrate, a numerical range of “about 0.1% to about 5%” should be interpreted to include not only the explicitly recited values of about 0.1% to about 5%, but also include individual values (e.g., 1%, 2%, 3%, and 4%) and the sub-ranges (e.g., 0.5%, 1.1%, 2.2%, 3.3%, and 4.4%) within the indicated range. Where the stated range includes one or both of the limits, ranges excluding either or both of those included limits are also included in the disclosure, e.g., the phrase “x to y” includes the range from ‘x’ to ‘y’ as well as the range greater than ‘x’ and less than ‘y’. The range can also be expressed as an upper limit, e.g., ‘about x, y, z, or less’ and should be interpreted to include the specific ranges of ‘about x’, ‘about y’, and ‘about z’ as well as the ranges of ‘less than x’, less than y’, and ‘less than z’. Likewise, the phrase ‘about x, y, z, or greater’ should be interpreted to include the specific ranges of ‘about x’, ‘about y’, and ‘about z’ as well as the ranges of ‘greater than x’, greater than y’, and ‘greater than z’. In some embodiments, the term “about” can include traditional rounding according to significant figures of the numerical value. In addition, the phrase “about ‘x’ to ‘y’”, where ‘x’ and ‘y’ are numerical values, includes “about ‘x’ to about ‘y’”.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the specification and relevant art and should not be interpreted in an idealized or overly formal sense unless expressly defined herein.

The articles “a” and “an,” as used herein, mean one or more when applied to any feature in embodiments of the present invention described in the specification and claims. The use of “a” and “an” does not limit the meaning to a single feature unless such a limit is specifically stated. The article “the” preceding singular or plural nouns or noun phrases denotes a particular specified feature or particular specified features and may have a singular or plural connotation depending upon the context in which it is used. As used herein the term “and/or” includes any and all combinations of one or more of the associated listed items. Like reference numbers refer to like elements throughout.

It will be understood that “monolayer” or “monolayer film” refers to a single layer film. A MoS₂monolayer film refers to a single layer film including a single plane of molybdenum atoms is sandwiched by planes of sulfide atoms and also may be referred to as a two-dimensional material.

Metal Chalcogenide Films

One challenge for improving the catalytic activity of MoS₂may be the lack of unambiguous understanding for its catalytically active sites. The common theory believes only the edge sites of crystalline MoS₂to be catalytically active, while the basal plane is inert for hydrogen evolution (Jaramillo, et al., Science 2007, 317, 100). As a result, one major strategy explored to improve the catalytic activity is increasing the number of edge sites (Karunadasa, et al., Science 2012, 335, 698; Kibsgaard, et al., Nat Mater 2012, 11, 963; Kong, et al., Nano Letters 2013, 13, 1341; Shi, et al., ACS Nano 2014, 8, 10196; Xie, et al., Journal of the American Chemical Society 2013, 135, 17881; Ye, et al., Nano Letters 2016, 16, 1097; Gao, et al., Nat Commun 2015, 6). Some recent studies have demonstrated other ways to enhance the catalytic activity of MoS₂for hydrogen evolution (Bench, et al., ACS Catalysis 2014, 4, 3957; Faber, et al., Energy & Environmental Science 2014, 7, 3519). For instance, MoS₂with the 1T structural phase may show improved catalytic activities, even with substantial oxidation at the edge sites (Merki, et al., Energ Environ Sci 2011, 4, 3878; Voiry, et al., Nano Letters 2013, 13, 6222; Lukowski, et al., J. Am. Chem. Soc. 2013, 135, 10274; Wang, et al., P. Natl. Acad. Sci USA 2013, 110, 19701; Yin, et al., Journal of the American Chemical Society 2016, 138, 7965). It has also been demonstrated that the catalytic activity of MoS₂films turns to increase with the thickness decreasing and can be even higher than that of edge-rich pyramid MoS₂nanoplates (Yu, et al., Nano Letters 2014, 15, 486; Tan, et al., Advanced Materials 2014, 26, 8023). These results suggest that the catalytic activity of MoS₂for hydrogen evolution is more complicated that what has been commonly believed.

The present inventors demonstrated that, besides the edge sites, the sulfur vacancy of MoS₂provides another major catalytically active site for the HER. The grain boundary may show some catalytic activity as well. The turnover frequency (TOF) of the edge sites, sulfur vacancies, and grain boundaries are quantitatively evaluated to be 7.5±1.5 s⁻¹, 3.2±0.4 s⁻¹, and around 0.1 s⁻¹, respectively. And the typical Tafel slopes are 65-75 mV/dec, 65-85 mV/dec, and 120-160 mV/dec for the edge sites, sulfur vacancies, and grain boundaries. Unlike the linear dependence on the length of the edge sites and grain boundaries, the catalytic activity is relatively high when the density of sulfur vacancies in a range of 5-15%, and in some embodiments in a range of 7-10%. A density of sulfur vacancies (or sulfur atom vacancies) refers to a number of sulfur atom sites not occupied by sulfur atoms per 100 sulfur atom sites. For example, if there are 100 sulfur atom sites that sulfur atoms are supposed to occupy, and 5 sulfur atom sites are not occupied by sulfur atoms, a density of sulfur vacancies is 5%.

The present inventors demonstrated that the catalytic activity of the sulfur vacancies is also related with the crystalline quality at the proximity of the vacancies as higher crystalline quality at the proximity may enable higher catalytic activity at the vacancies. Monolayer MoS₂having the optimal density of sulfur vacancies may show high catalytic activity, and MoS₂having high crystalline quality may show high catalytic activity. It is worthwhile to point out that a very recent work has also claimed catalytic activities at the sulfur vacancies of MoS₂for the HER, but the experimental results used to support the claim are misinterpreted and actually cannot support the claim (Li, et al., Nat Mater 2016, 15). In stark contrast with the previous studies, the present inventors demonstrated that the sulfur vacancies created by Ar plasma treatment are not catalytically active, which is likely due to an unfavorable crystalline structure at the proximity of the plasma-created vacancies (Li, et al., Nat Mater 2016, 15).

A variety of metal chalcogenide films are provide. The metal chalcogenide films according to some embodiments of the present invention may be formed using methods discussed in U.S. Pat. No. 9,527,062, which is herein incorporated by reference in a matter consistent with the present application, and may continuously extend. The metal chalcogenide films can be thin, e.g., having a thickness of about 15 nm, about 10 nm, about 8 nm, about 5 nm, about 2 nm, about 1 nm, or less. In some embodiments, the metal chalcogenide films are atomically thin, e.g., having a thickness of about 30 Å, about 10 Å, about 8 Å, about 6 Å, or less. In some embodiments, the metal chalcogenide films may be a monolayer or bilayer films. In some embodiments, the metal chalcogenide films may include less than 10 layers (e.g., 10 monolayers of metal chalcogenide). The metal chalcogenide films may contain a metal atom, for example, Mo, W, Co, Zn, Fe, Re, Nb, and/or Ni. In some aspects, a chalcogen atom of the metal chalcogenide films may be S and/or Se. In some embodiments, the metal chalcogenide is MoS₂, WS₂, MoSe₂, WSe₂, NiS, Ni₂S₃, NiS₂, CoS, Co₂S₃, CoS₂, ReS₂, NbS₂, or alloy thereof. The films may have a current density of about 15 mA/cm², about 18 mA/cm², about 20 mA/cm², about 22 mA/cm², about 25 mA/cm², or greater when measured at an overpotential of about 0.17 V, about 0.18 V, about 0.19V, about 0.20 V, or about 0.21 V.

The films can be polycrystalline, e.g., having an average grain size of about 2 nm to 2000 nm, about 20 nm to 120 nm, about 30 nm to 100 nm, about 40 nm to 100 nm, or about 50 nm to 120 nm. The films can be atomically smooth, e.g., having no or few edge sites. This is contrasted to layers of metal chalcogenide flakes which have a high density of edge sites. In some embodiments, the film may have a current density higher than the current density of a layer of metal chalcogenide flakes having about the same surface area when measured under the otherwise same conditions. In some embodiments, the current density is at least about 2 times, 5 times, or 10 times the current density of a layer of metal chalcogenide flakes having about the same surface area when measured under the otherwise same conditions.

In some embodiments, the metal chalcogenide films may have a plurality of chalcogen atom vacancies, wherein the chalcogen atom vacancies are present at a density of about 5% to 15%. The chalcogen atom vacancies can provide for improved catalytic activities, especially when at the density of about 5% to about 15%, more specifically, about 7% to about 10%. In some embodiments, the films may maintain high crystallinity even with the presence of the vacancies, e.g., as compare the vacancies generated by plasma treatment. In some embodiments, the film may have a current density higher than the current density of the otherwise same film prepared by the same methods and having about the same density of the chalcogen atom vacancies when tested under the otherwise same conditions, except where the film has been treated with Ar plasma to generate the chalcogen atom vacancies. In some embodiments, the current density may be at least about 2 times, 5 times, or 10 times the current density of the metal chalcogenide film where the film has been treated with Ar plasma to generate the chalcogen atom vacancies.

In some embodiments, the metal chalcogenide films may be on a substrate selected from the group consisting of a glassy carbon substrate, a gold substrate, a nickel substrate, a titanium substrate and a platinum substrate. In some embodiments, the use of a gold substrate or a nickel substrate may result in improved catalytic activities as compared to the otherwise same film but on a glassy carbon substrate. In some embodiments, the current density of the metal chalcogenide film on the nickel or gold substrate is at least about 2 times, 5 times, or 10 times the current density of the otherwise same metal chalcogenide film on the glassy carbon substrate.

In some embodiments, the metal chalcogenide films have a plurality of hydrogen atoms intercalated within the metal chalcogenide. The hydrogen atoms can result in improved catalytic activity as compared to the otherwise same film except without the hydrogen atoms.

Catalyst For Hydrogen Evolution Reaction (HER)

According to some embodiments of the present invention, catalysts for hydrogen evolution reaction (HER) are provided. FIGS. 36A, 36B, 37A, 37B and 38 show schematic diagrams of a catalyst according to some embodiments of the present invention.

Referring to FIG. 36A, the catalyst may include a substrate 10 and a metal chalcogenide film 100 discussed herein. The metal chalcogenide film 100 according to some embodiments of the present invention may be formed using methods discussed in U.S. Pat. No. 9,527,062, which is herein incorporated by reference in a matter consistent with the present application, and may continuously extend. The metal chalcogenide film 100 may continuously extend on the substrate 10. The metal chalcogenide film 100 may include chalcogen atom vacancies, and a density of the chalcogen atom vacancies may be about 5% to about 15%. In some embodiments, the density of the chalcogen atom vacancies may be about 7% to about 10%. Referring to FIG. 36B, the metal chalcogenide film 100 may include dopants 102. In some embodiments, the dopants 102 are nickel atoms and/or cobalt atoms. In some embodiments, the dopant atoms 102 may replace chalcogen atoms of the metal chalcogenide film 100.

In some embodiments, the metal chalcogenide film 100 may have a thickness of about 10 nm or less. In some embodiments, the metal chalcogenide film 100 may have a thickness of about 30 A or less. In some embodiments, the metal chalcogenide film 100 may be a monolayer film and may be a polycrystalline film having an average grain size of about 30 nm to 1000 nm. In some embodiments, the metal chalcogenide film 100 may be a film including less than 10 layers (e.g., 10 monolayers of metal chalcogenide). In some embodiments, an average grain size of the metal chalcogenide film 100 is about 2 nm to about 2000 nm. In some embodiments, the substrate 10 may include nickel, titanium, silver, cobalt, zinc, and/or platinum. In some embodiments, the substrate 10 may be a glassy carbon substrate, a gold substrate, a nickel substrate, a titanium substrate, a silver substrate, a cobalt substrate, and a platinum substrate.

Referring to FIG. 37A, the catalyst may include hydrogen ions 200 disposed between the substrate 10 and the metal chalcogenide film 100. Referring to FIG. 37B, the catalyst may include multiple of the metal chalcogenide films 100. The catalyst may also include hydrogen ions 200 disposed between two adjacent metal chalcogenide films 100.

Referring to FIG. 38, the substrate 10 may include a primary substrate 12 and a metal layer 14. The metal layer 14 may extend between the primary substrate 12 and the metal chalcogenide film 100. In some embodiments, the metal layer 14 may have a thickness of about 1 nm to about 10 nm. In some embodiments, the metal layer 14 may be a platinum layer. In some embodiments, the primary substrate 12 may be a nickel substrate or a metal substrate.

Making Metal Chalcogenide Films

A variety of methods are provided for making the metal chalcogenide films described herein. The methods may include performing a chemical vapor deposition (CVD) process using a metal precursor and a chalcogenide at a temperature, pressure, and flow rate to deposit the metal chalcogenide onto a receiving substrate to form the metal chalcogenide film. In some embodiments, the temperature may be about 200° C. to 900° C., about 700° C. to 900° C., about 750° C. to 900° C., about 800° C. to 900° C., or about 850° C. In some embodiments, the pressure may be about 0.1 Torr to about 500 Torr. In some embodiments, the pressure may be about 1.5 Torr to 2.5 Torr or about 2 Torr. In some embodiments, the flow rate may be about 25 sccm to 75 sccm, about 35 seem to 65 sccm, or about 50 sccm. The receiving substrate may be sapphire or other suitable substrate. In some embodiments, the metal precursor may be a metal chloride or a metal oxide. For example, the metal precursor may be MoCl₅, MoCl₃, MoO₂Cl₂, MoOCl₃, WCl₆, MoO₃, WO₃, Mo(CO)₆, W(CO)₆, a compound comprising Mo and/or a compound comprising W. In some embodiments, the chalcogen precursor may be sulfur powder, selenium powder, and/or hydrogen sulfide powder.

In some embodiments, the chemical vapor deposition (CVD) process may be performed additionally using a precursor including dopant atoms to form the metal chalcogenide film including dopants. The metal chalcogenide film may be a monolayer film, and the dopant atoms may be nickel atoms, cobalt atoms zinc atoms, iron atoms, Re atoms, and/or Nb atoms. In some embodiments, the metal chalcogenide film may include multiple layers and may include less than 10 layers (e.g., 10 monolayers of metal chalcogenide). The precursor including the dopant atoms may be nickel acetylacetonate (e.g., Ni(acac)₂), a compound comprising nickel, a compound comprising Co, a compound comprising Fe, a compound comprising Zn, a compound comprising Re, and/or a compounds comprising Nb, and the metal chalcogenide film may be a molybdenum disulfide monolayer including nickel atoms, cobalt atoms, zinc atoms, iron atoms, rhenium (Re) atoms, and/or Niobium (Nb) atoms. In some embodiments, the metal chalcogenide film may be a molybdenum disulfide film including less than 10 layers (e.g., 10 monolayers of molybdenum disulfide).The chemical vapor deposition (CVD) process may be performed at a temperature in a range of about 200° C. to about 900° C. In some embodiments, the chemical vapor deposition (CVD) process may be performed at a temperature in a range of about 800° C. to about 900° C. (e.g., about 850° C.). The chemical vapor deposition (CVD) process may be performed at a pressure in a range of about 0.1 Torr to about 500 Torr. In some embodiments, the pressure may be in a range of about 1.5 Torr to about 2.5 Torr (e.g., 2 Torr).

The methods may further include transferring the film to a different substrate. In some embodiments, the methods may include surface-energy assisted transfer of the metal chalcogenide film to a substrate selected from the group consisting of a glassy carbon substrate, a nickel substrate, a gold substrate, a titanium substrate, or a platinum substrate. The surface-energy assisted transfer may include spin-coating a polystyrene layer onto the metal chalcogenide film, applying water to the polystyrene layer to delaminate the metal chalcogenide film from the receiving substrate, transfer of the polystyrene layer and the metal chalcogenide film to the substrate, and rinsing with toluene to dissolve the polystyrene layer to produce the metal chalcogenide film on the substrate.

The methods may include intercalating hydrogen atoms into the metal chalcogenide. In some embodiments, the methods may include treating the metal chalcogenide film with acid to produce the metal chalcogenide film having a plurality of hydrogen atoms intercalated within the metal chalcogenide film.

According to some embodiments, methods of activating a catalyst, which is discussed herein, are provided. The method may include adding hydrogen ions on a surface of the metal chalcogenide film. In some embodiments, adding hydrogen ions may include electrochemically polarizing the metal chalcogenide film at negative potentials in an acidic media. In some embodiments, electrochemically polarizing the metal chalcogenide film at negative potentials may include performing a Cyclic Voltammetry (CV). In some embodiments, adding hydrogen ions may include immersing the metal chalcogenide film into an acidic solution, and acidic solution may be the bis(trifluoromethane)sulfonamide or hydrogen sulfuric acid.

Methods of Using Metal Chalcogenide Films

The metal chalcogenide films described herein can have high catalytic activity for hydrogen evolution from water. In some embodiments, methods of generating hydrogen from water are provided, the methods including contacting the water with metal chalcogenide films or catalysts described herein and metal chalcogenide films or catalysts formed using the methods described herein.

Calculation of the Turnover Frequencies at Single Active Sites

The TOF is estimated by using the current density at 0V vs. RHE. The turnover frequency is calculated as following

TOF=R
_H2
/N
_active

where R_H2is the number of hydrogen molecules produced per unit time and unit area at 0V vs. RHE, and N_activeis the number of catalytic active sites per unit area. The hydrogen molecules produced per unit time and unit area can be calculated from the current density at 0V vs. RHE J₀as R_H2=J₀/2e.

For sulfur vacancies, the number of active sites per unit area is calculated from the measured density of sulfur vacancy a % using the following equation.

$N_{sulfur} = \frac{a %}{(0.32 \times 0.32 \times \frac{\sqrt{3}}{2}) \times 10^{- 8}}$

For simplicity, it was assumed that the sulfur vacancies are evenly distributed in the top and bottom layers of sulfur atoms. Only the sulfur vacancies at the top layer of sulfur atoms was considered as catalytically active, because the sulfur vacancies in the bottom layer of sulfur atoms is embedded underneath and may not be accessible for reaction.

For the edge sites, it was considered that each edge site or grain boundary occupy a length of 0.32 nm, which is the lattice constant of MoS₂.

N
_edge=(measured length/unit area)/0.32 nm

The number of grain boundary can be calculated using the same method as the calculation for the edge sites.

EXAMPLES

Now having described embodiments of the present invention, in general, the following Examples describe some additional embodiments of the present invention. While embodiments of the present invention are described in connection with the following examples and the corresponding text and figures, there is no intent to limit embodiments of the present invention to this description. On the contrary, the intent is to cover all alternatives, modifications, and equivalents included within the spirit and scope of embodiments of the present invention.

Example 1: Catalytic Sites of MoS₂for Hydrogen Evolution

MoS₂presents a promising low-cost catalyst for the hydrogen evolution reaction (HER), but the understanding about its active sites has remained to be limited. The catalytic activities of all possible reaction sites of MoS₂, including edge sites, sulfur vacancies, and grain boundaries have been examined. The present inventors demonstrated that, in addition to the well-known catalytically active edge sites, sulfur vacancies provide another major active site for the HER while the catalytic activity of grain boundaries is much weaker. The intrinsic turnover frequencies (Tafel slopes) of the edge sites, sulfur vacancies, and grain boundaries are estimated to be 7.5 s⁻¹(65-75 mV/dec), 3.2 s⁻¹(65-85 mV/dec), and 0.1 s⁻¹(120-160 mV/dec), respectively. The present inventors also demonstrated that the catalytic activity of sulfur vacancies strongly depends on the density of the vacancies and the local crystalline structure at the proximity of the vacancies. Unlike edge sites, whose catalytic activity linearly depends on the length, sulfur vacancies show optimal catalytic activities when the vacancy density is in a range of about 5-15%, in some embodiments, in a range of about 7-10%. And the sulfur vacancies in the MoS₂whose crystalline quality is otherwise high tends to show better catalytic activities.

Synthesis and Transfer of Monolayer MoS₂Films and Flakes

MoS₂thin films were synthesized using a self-limiting chemical vapor deposition process that recently have been developed (Yu, et al., Sci Rep-Uk 2013, 3). Briefly molybdenum chloride (MoCl₅) powder (99.99%, Sigma-Aldrich) was placed at the center of the furnace and sulfur powder (Sigma-Aldrich) at the upstream entry of the furnace. Receiving substrates (sapphire) were placed in the downstream of the tube. Typical conditions include a temperature of 850° C., a flow rate of 50 sccm, and a pressure around 2 Torr. The film with different densities of sulfur vacancies were grown by varying the growth temperature in the range of 700-900° C. Monolayer MoS₂flakes were grown using a different chemical vapor deposition (CVD), in which MoO₃(99.99%, Sigma-Aldrich) instead of MoCl₅was used as the precursor (Yu, et al., Nano Letters 2014, 15, 486).

The transfer of the monolayers followed a surface-energy-assisted transfer approach that have been developed previously (Gurarslan, et al., ACS Nano 2014, 8, 11522). Briefly, a layer of polystyrene (PS) was spin-coated on the as-grown monolayers. A water droplet was then dropped on the top of the polymer. Due to the different surface energies of the monolayer and the substrate, water molecules could penetrate under the monolayer, resulting the delamination of the PS-monolayer assembly. The polymer/monolayer assembly was picked with a tweezers and then transferred to glassy carbon substrates. Finally, PS was removed by rinsing with toluene several times.

Repair of Sulfur Vacancies

Sulfur repair was conducted following a process reported previously (Qiu, et al., Nat Commun 2013, 4). Basically, monolayer MoS₂was dipped in a solution of 1/15 (volume ratio) MPS (Sigma-Aldrich)/dichloromethane (Sigma-Aldrich) for 48 hours in a dry glove box. The samples were then rinsed thoroughly with dichloromethane and isopropanol (Sigma-Aldrich), and blown dry with N₂(Cao, MRS Bulletin 2015, 40, 592). Finally the samples were annealed in a flow of Ar (with 5% H₂) at 350° C. for 20 minutes.

Structure and Composition Characterizations

Raman measurements were carried out by a Horiba xPlora system equipped with an excitation wavelength at 532 nm. AFM measurements were performed at a Veeco Dimension-3000 atomic force microscope. XPS measurements were carried out at X-ray photoelectron spectroscope (SPECS System with PHOIBOS 150 analyzer using an Mg Kα X-ray source). Magnetization measurements were performed at 350 K in a Quantum Design® MPMS SQUID VSM. The magnetic field was applied in the plane of the samples that were mounted on a diamagnetic quartz sample holder.

Electrochemical Characterizations

The electrochemical characterization was performed in 0.5 M H₂SO₄using a CH Instrument electrochemical analyzer (Model CHI604D) with a saturated calomel reference electrode (SCE). A Pt mesh (2.54 cm×2.54 cm) and a graphite rod were used as counter electrode, and the tested results did not show difference. Nitrogen gas was bubbled into the electrolyte throughout the experiment. The potential shift of the SCE is calibrated to be −0.262 V vs. RHE. Typical electrochemical characterizations of the monolayers were performed using linear sweeping from +0 V to −0.5V (vs. RHE) with a scan rate of 5 mV/s. The electrolyte resistance and capacitance of the electrocatalysts were characterized using electrochemical impedance spectroscopy (EIS). The AC impedance is measured within the frequency range of 10 to 1 Hz with perturbation voltage amplitude of 5 mV (typical EIS measurement results can be seen in FIG. 6) (Yu, et al., Nano Letters 2014, 15, 486). An equivalent Randles circuit model was fit to the data to determine the system resistance and capacitance (Yu, et al., Nano Letters 2014, 15, 486). The electrochemical deposition of Cu was performed in 1 M CuSO₄by linearly sweeping from +0 V to −0.08V (vs. RHE) with a scan rate of 5 mV/s.

Results and Discussion

The present inventors started with examining the catalytic activities of continuous monolayer MoS₂films and discrete monolayer MoS₂flakes as shown in FIGS. 1A-1B. The film and flakes are synthesized on sapphire substrates and then transferred onto glassy carbon substrates for catalytic characterizations using a surface-energy-assisted transfer process (FIGS. 1A-1B) (Gararslan, et al., ACS Nano 2014, 8, 11522). More specifically, the film was grown using a self-limiting CVD process previously developed, and the flakes are synthesized with another CVD process reported in the references (Yu, et al., Sci Rep-Uk 2013, 3; Lee, et al., Adv Mat 2012, 24, 2320). The present inventors have previously demonstrated that the crystalline and surface quality of the film and the flakes may be nicely preserved during the transfer process. The present inventors have also confirmed that the film is a continuous and highly uniform monolayer with no voids, cracks, and steps (FIGS. 7A-7D) (Li, et al., Nat Mater 2016, 15). The film is expected to bear little edge sites due to the structural continuity (FIG. 1A). In contrast, the discrete flakes are monolayers in size of 1 micrometer with well-defined edges (FIG. 1B and FIGS. 7A-7D). The presence of little edge sites at the film is supported by magnetic measurement. It is well known that the edge sites of MoS₂may provide ferromagnetic moments, and our magnetic measurement indeed shows substantial ferromagnetic moment at the flakes but not at the film (FIG. 1C) (Tongay, et al., Appl Phys Lett 2012, 101; Pan, et al., J Phys Chem C 2012, 116, 11752). According to the common theory, which believes only the edge site catalytically active, one would expect much worse catalytic performance at the film than at the flakes (Jaramillo, et al., Science 2007, 317, 100).

In stark contrast with the intuitive expectation, the edge-less monolayer film exhibits much better catalytic activities than the monolayer flakes. FIGS. 1D-1E show polarization curves and corresponding Tafel plots of the film and the flakes. The result for the flakes has already been normalized to the area coverage of the flakes on the substrate. The film can provide a current density of 20 mA/cm₂at overpotential of around 0.19 V, which is among the best of what previously reported with all kinds of MoS₂materials in references (Karunadasa, et al., Science 2012, 335, 698; Kibsgaard, et al., Nat Mater 2012, 11, 963; Kong, et al., Nano Letters 2013, 13, 1341; Shi, et al., ACS Nano 2014, 8, 10196; Xie, et al., Journal of the American Chemical Society 2013, 135, 17881; Ye, et al., Nano Letters 2016, 16, 1097; Gao, et al., Nat Commun 2015, 6; Voiry, et al., Nano Letters 2013, 13, 6222; Wang, et al., Natl. Acad Sci USA 2013, 110, 19701; Tan, et al., Adv Mater 2014, 26, 8023; Benck, et al., ACS Catalysts 2012, 2, 1916; Chen, et al., Nano Lett 2011, 11, 4168; Li, et al., Journal of the American Chemical Society 2011, 133, 7296; Li, et al., ACS Catalysts 2015, 5, 448; Merki, et al., Chem Sci 2011, 2, 1262). The present inventors could find the Tafel slope and current density by fitting the Tafel plots to the equation of η=ρlog(j)+log(j0), where η is the overpotential (vs. RHE), j0 is the exchange current density, and ρ is the Tafel slope (Bockris, Russ J Electrochem+ 1995, 31, 1211). Both the film and the flakes show similar Tafel slopes of 70 mV/dec, but the current density of the film is one order of magnitude higher, 40 μA/cm²vs. 3.5 μA/cm²at the flakes. Significantly, the current density of the monolayer film is more than one order of magnitude higher than what previously observed at MoS₂catalysts (Karunadasa, et al., Science 2012, 335, 698; Kibsgaard, et al., Nat Mater 2012, 11, 963; Kong, et al., Nano Letters 2013, 13, 1341; Shi, et al., ACS Nano 2014, 8, 10196; Xie, et al., Journal of the American Chemical Society 2013, 135, 17881; Ye, et al., Nano Letters 2016, 16, 1097; Gao, et al., Nat Commun 2015, 6; Voiry, et al., Nano Letters 2013, 13, 6222; Wang, et al., Natl. Acad Sci USA 2013, 110, 19701; Tan, et al., Adv Mater 2014, 26, 8023; Benck, et al., ACS Catalysts 2012, 2, 1916; Chen, et al., Nano Lett 2011, 11, 4168; Li, et al., Journal of the American Chemical Society 2011, 133, 7296; Li, et al., ACS Catalysts 2015, 5, 448; Merki, et al., Chem Sci 2011, 2, 1262). This is truly remarkable as the MoS2 materials studied in the previous studies often have surface roughness orders of magnitude greater than the atomically smooth film. The capacitance of our smooth film is around 2.6 μF/cm², while the capacitance of most of previous studies are in scale of several or tens of mF. Part of the reason for this extraordinarily high current density is rooted in the atomically thin dimension of the monolayer film. The present inventors have previously demonstrated that the current density is dependent on the charge transfer efficiency inside MoS₂, and monolayers can best facilitate the charge transfer in the vertical direction (Yu, et al., Nano Letters 2014, 15, 486). It is worthwhile to note that the monolayer shows remarkable stability with negligible decrease in the catalytic activity even after 10,000 cycles.

The present inventors found that the catalytic activity of the flakes can be mainly correlated to the edge sites, which is consistent with the common theory. This is evidenced by a linear dependence of the current density of the flakes on the length of edges as discussed later (see FIG. 4D). However, the excellent catalytic activity at the continuous MoS₂film is unexpected. Previous studies have demonstrated that the catalytic activity of MoS₂may be substantially improved when its structure is changed from 2 H to 1 T (Voiry, et al., Nano Letters 2013, 13, 6222; Lukowski, et al., J Am. Chem. Soc. 2013, 135, 10274; Wang, et al., P Natl. Acad Sci USA 2013, 110, 19701). But the present inventors could exclude out the possibility of structural changes to be the reason for the high catalytic activity of the film. The present inventors examined the Raman (FIG. 2A) and XPS (FIGS. 9A-9B) of the monolayers before and after the catalytic reaction, and find no change in the composition and crystalline structure of both the film and flakes. Therefore, the unexpected excellent catalytic performance of the edge-less monolayer film strongly suggests the presence of catalytic active sites other than the edge.

To understand the catalytic activity of the film, the present inventors examined the possible difference in composition and structure between the film and the flakes. The film is distinguished by the presence of grain boundaries and sulfur vacancies. Unlike the flake, which is well known to be single crystalline, the film is polycrystalline with grain size in the range of 30-100 nm. Additionally, XPS measurements show that the stoichiometric ratio of S:Mo in the film is smaller than that of the flakes (FIG. 2B), indicating the presence of sulfur vacancies in the film. However, despite the presence of these crystalline defects, the Raman spectrum of the film is similar to that of the flakes with comparable intensities and line shapes in the two main characteristic Raman peaks A_1gand E¹_2g(FIG. 10). This indicates that the film has an overall high crystalline quality but bears some defects including grain boundaries and sulfur vacancies.

The experimental result indicates that the catalytic activity of the film can be correlated to the sulfur vacancies. The present inventors treated the film and the flakes with a process well established to repair sulfur vacancies, and monitor the catalytic activities before and after the repair (Qiu, et al., Nat Commun 2013, 4; Makarova, et al., J Phys Chem C 2012, 116, 22411; Cho, et al., ACS Nano, 2015, 9, 8044). Briefly, the film and the flake are immersed in (3-mercaptopropyl) trimethoxysilane (MPS) followed by annealing at 300° C. MPS molecules may be adsorbed at sulfur vacancies and transfer sulfur atoms to the vacancies through the dissociation of S—C bonds under elevated temperature as illustrated in FIG. 3A. The film shows a decrease in photoluminescence (PL) intensity and an increase in the S:Mo stoichiometric ratio after the treatment (FIGS. 3B-3C), which is consistent with what reported previously and indicates the successful repair of the sulfur vacancies (Cho, et al., ACS Nano, 2015, 9, 8044). In contrast, the flakes show negligible change in PL (FIG. 3D) and S/Mo ratio (FIG. 11) after the treatment, suggesting the presence of little sulfur vacancies in the flakes as expected. Accordingly, little change can be found in the catalytic performance of the flakes after the repair (FIG. 3D). But the catalytic activity of the film dramatically decreased afterward. The current density decreased by more than one order of magnitude from 30.1 μA/cm²to 1 μA/cm²and the Tafel slope from 70 mV/dec to 125 mV/dec (see the Tafel plots in FIG. 12). The present inventors could exclude out the possible physical covering of residual MPS molecules to be the reason for the decrease in catalytic activity, as it would otherwise give rise to similar decrease in both film and flakes. The negligible physical coating is also supported by AFM measurements that indicate no substantial change in the thickness of the film after the repair treatment. Therefore, the dramatic decrease in catalytic activity can be ascribed to the elimination of sulfur vacancies by the repair. It indicates that the catalytic activity of the film is mainly contributed by the sulfur vacancies.

Experimental results provided herein also indicate that the catalytic activity of the grain boundaries is weak. This is evidenced by the weak catalytic performance of the film after the repair of sulfur vacancies. With a grain size in the range of 30-100 nm, the film has a considerable amount of grain boundaries. The weak catalytic activity of the grain boundaries is also supported by a poor catalytic performance of the monolayer MoS₂film with little sulfur vacancies. This vacancy-less film is synthesized with the same process used to grow the flakes and formed by the merge of neighboring flakes in the case of high nucleation densities. The present inventors have confirmed that the flake-merged film have identical composition as the individual flakes with little sulfur vacancies by Raman, PL, and XPS measurements. The flake-merged film also has little edge sites due to structural continuity, and its grain size is estimated to be around 1 μm. It shows very poor catalytic performance as indicated by the polarization curve given in FIG. 3C (dashed curve). The current density is as low as 0.04 μA/cm²and the Tafel slope around 160 mV/dec. The present inventors found in experiments that the current density of the repaired and flake-merged films shows a roughly linear dependence on the length of the grain boundaries (FIG. 13). This further supports that the weak catalytic activities of the repaired and the flake-merged films are mainly contributed by the grain boundaries.

The present inventors could quantitatively evaluate the catalytic activity for each of the different sites, including sulfur vacancies, edge sites, and grain boundaries. In order to evaluate the catalytic activity of single sulfur vacancies, the present inventors examined the catalytic performance of monolayer MoS₂films with different densities of sulfur vacancies. The density of sulfur vacancies is controlled by controlling the growth conditions such as temperature (in the range of 700-900° C.), and may be quantitatively estimated from XPS measurement (FIG. 4A and FIG. 14). The vacancy density estimated from the XPS measurement is also confirmed by STEM measurements, in which the present inventors could directly visualize the sulfur vacancies. The present inventors have confirmed that the overall crystalline qualities of all the films are reasonably comparable as indicated by comparable intensities of the characteristic Raman peaks (FIG. 15). FIG. 4B shows the polarization curves collected from these films. The catalytic performance is strongly dependent on the vacancy density. To better illustrate the dependence, the present inventors extracted the current density and Tafel slope from the measured polarization curves and plot them as a function of the vacancy density in FIG. 4C. The catalytic activity of the film is high when the density of sulfur vacancies is in a range of 7-10%, with exchange current densities in the range of 30-60 μA/cm²and Tafel slopes in the range of 65-75 meV/dec. The present inventors could estimate the turnover frequency (TOF), which indicate the catalytic reaction rate at single active sites, at each sulfur vacancy to be around 3.2±0.4 s⁻¹(the turnover frequencies at single active sites were calculated as described above). As a reference, the present inventors also studied the catalytic performance of the flakes as a function of the density of edge sites. The result indicates a linear dependence of the current density on the edge length per unit area (FIG. 4D), which is consistent with what reported previously (Shi, et al., ACS Nano 2014, 8, 10196; Jaramillo, Science 2007, 317, 100). The present inventors could estimate the turnover frequency of each edge site to be around 7.5±1.5 s⁻¹(the turnover frequencies at single active sites were calculated as described above). Additionally, the present inventors could roughly estimate that the turnover frequency of each grain boundary site is 0.1 s⁻¹, more than one or two orders of magnitude smaller than that of the sulfur vacancies and edge sites. Unlike the exchange current density, the Tafel slope shows much less dependence on the density of catalytic active sites. The Tafel slope usually remains to be in the range of 65-75 mV/dec and 65-85 mV/dec for the edge sites and sulfur vacancies, but may turn to be much larger when the density of the active sites gets to be very small (FIGS. 4C-4D). The Tafel slope of the grain boundaries is much higher, in the range of 120-160 mV/dec. It should be noted that the result for the grain boundaries is actually an averaged result over many different types of ground boundaries. In reality, the activity of some grain boundaries could be stronger than others.

The present inventors could directly visualize the active sites by performing electrochemical deposition of Cu metal (Cu²⁺→Cu⁰) at the MoS₂film and flakes. The treatment of sulfur vacancy repair may substantially suppress the Cu deposition at the film but not at the flakes, similar to the effect of the repair on the HER (FIGS. 3C-3D). This indicates that the electrochemical deposition of Cu shares the same catalytic active sites with the HER and can be used to directly visualize the catalytic active sites. FIGS. 5A-5D show the Cu particles deposited at the film and the flakes. The Cu particles distribute uniformly everywhere on the MoS₂film while predominantly at the edge sites for the flakes, which is consistent with expectation based on the studies of hydrogen evolution. The Cu deposition may also occur at mirror grain boundaries (FIG. 5C) but the deposited particles are smaller and less dense than those deposited at the neighboring edge sites. No Cu deposition can be found at tilt grain boundaries (FIG. 5D). This suggests that the mirror grain boundary may have some catalytic activity that is weaker than the edge sites, while the tilt grain boundary is not catalytically active.

The catalytic activity of the sulfur vacancies also shows dependence on the local crystalline structure at the proximity of the vacancy. Generally, the monolayer MoS₂film with low crystalline quality, as indicated by low Raman and PL intensities, exhibits much worse catalytic activity than the counterpart with comparable sulfur vacancies but higher crystalline quality (FIGS. 16A-16D). With the densities of sulfur vacancies being comparable, the film with lower crystalline quality involves more grain boundaries. The worse catalytic performance suggests that the sulfur vacancies located in or close to grain boundaries are not very active. This observation is consistent with the result of our previous studies. The present inventors have previously demonstrated that the monolayer film directly grown on glass carbon substrates, which is in low crystalline quality as indicated by Raman spectra, shows very poor catalytic activity with current density being 1 μA/cm²and Tafel slope of 140 mV/dec (Yu, et al., Nano Letters 2014, 15, 486). The present inventors also found that the sulfur vacancies created by Ar plasma treatment are not catalytic active (see FIGS. 17A-19C). This likely due to unfavorable crystalline structures at the proximity of the sulfur vacancies as the plasma treatment may likely also introduce other crystalline imperfection like Mo vacancies. The local environment could affect the electronic structure at the sulfur vacancy and eventually affect the catalytic activity.

Conclusion

The experimental result indicates that both sulfur vacancies and edge sites may be exploited to improve the catalytic performance of MoS₂, while the grain boundaries may only provide minor benefit. It explicitly suggests that engineering sulfur vacancies provides a better strategy than increasing the number of edge sites, at least from the perspective of viability. According to the TOF of the sulfur vacancies and the edge sites, achieving catalytic performance comparable to the films with the optimal range of sulfur vacancies would require a high coverage of well separated monolayer MoS₂flakes in size of less than 100 nm, the latter of which is very difficult to obtain in experiments. The result also points out that the desired structure would be MoS₂films with an overall high crystalline quality but involving an optimal density of sulfur vacancies, as the local crystalline structure at the proximity of the vacancy may strongly affect the activity of the vacancy.

Example 2: Catalytic Inactivity of Sulfur Vacancies Created by Ar Plasma Treatment

Our experimental results indicate that the sulfur vacancies created by Ar plasma treatment are not catalytic active. This is in stark contrast with what reported previously by a recent work. The previous study reports an increase in the catalytic activity of MoS₂films after being treated by Ar plasma and ascribes the increase to the creation of sulfur vacancies. However, the present inventors found that the increase in catalytic activity resulting from Ar plasma treatment is mainly due to other two effects of the treatment: cleaning the surface and creation of cracks that may have edge sites. The sulfur vacancies created by Ar plasma treatment actually contribute, if any, little to the catalytic activity.

The present inventors have investigated the catalytic activity of MoS₂films treated by Ar plasma in a way similar to what reported by the previous work. The film is grown using the same process for the synthesis of MoS₂flakes, similar to what used in the previous study. The film is then transferred onto glassy carbon substrates for catalytic characterization using the surface-energy-assisted transfer technique developed by the inventors. Ar plasma treatment was performed at the film using two different ways. In one way, cyclic voltammetry (CV) was first performed at the film in the range of 0 - −0.5V (vs. RHE) for thousands times till the catalytic activity appears to be stable, and then treat the film with Ar plasma. In the other way, the transferred film with no pretreatment cycling was treated using Ar plasma. The treatment conditions at both ways are kept to be comparable.

During the Ar plasma treatment, MoS₂samples were bombarded with argon ions using a radiofrequency, inductively-coupled plasma source in a cylindrical chamber that is 4 in in diameter and 6 in in length with a quartz housing. Using a pressure control system coupled to the chamber, all experiments were conducted at 20 mTorr. The sample was suspended in the middle of the chamber attached to a ceramic rod. The plasma is produced inside the chamber using a 3-turn copper coil wrapped around the quartz housing, driven by a pulsing generator at 13.56 MHz. The generator was pulsed at 50 W for 20% of a 1 kHz duty cycle. This power delivery scheme allows for a lower particle flux to the substrate compared to constant power source operation. Ions produced in this discharge are accelerated to approximately 15 V before impacting the samples, with the ion flux up to 11.0 A/m2 (6.9E19 particles/m2/s) for each second of sample exposure time.

The Ar treatment may indeed induce defects, as indicated by the defect peaks in Raman spectra. This defect peaks are similar to what reported in the previous work, suggesting the formation of comparable defects. XPS measurement also shows an decrease in the stoichiometric ratio of S:Mo after the treatment, indicating the formation of sulfur vacancies as reported by the previous study. From the XPS measurement, the present inventors could quantitatively estimate that around 10% sulfur vacancies are created by the treatment. Additionally, the Raman and XPS measurements have confirmed that the pre-treatment cycling may not change the composition and structures of the film at all.

The present inventors monitored the catalytic performance of the films through the process of the treatment, and plot the results in FIGS. 18A-18B. The catalytic performance of the film shows dramatic improvement during the pre-treatment cycling process and usually tends to be stable after thousands of cycles. This cycling process just serves to clean the surface to expose catalytic active sites. A couple of recent works and the inventors' own studies have demonstrated that air-borne carbonaceous contaminants may adsorb to the surface of the film in ambient environment. However, the Ar treatment shows negligible effects on the catalytic performance of these pre-cycled films. This indicates that the sulfur vacancies created by the Ar plasma are not catalytically active. To further confirm the catalytic inactivity of the plasma-generated sulfur vacancies, the present inventors have repaired the sulfur vacancies at the Ar-treated films using the process reported in the main text, and find negligible changes in the catalytic performance after the repair. On the other hand, the present inventors observed an increase in catalytic activity at the non-pre-cycled film after the Ar treatment. But this increase may be just due to the cleaning effect of Ar plasma treatment, which may remove the carbonaceous contaminants at the surface of the film as the cycling process. The catalytic performance of the treated non-pre-cycled film may be further improved by post-treatment cycling to be similar to that of the thoroughly pre-cycled film with no Ar treatment.

The present inventors have also examined the catalytic performance of the films treated by Ar plasma for different durations. The films are cycled till stable catalytic performance prior to the Ar plasma treatment. FIG. 19A shows Raman spectra of the films treated by Ar plasma for 2, 10, and 15 seconds. The Raman spectra indeed shows more defects at the films treated longer time, and the defect-associated peaks are again similar to what reported in the previous work, indicating the formation of comparable defects. The catalytic characterization results for these treated films are plotted in FIGS. 19B. The results of the films prior to the treatment are also given as a reference. There are negligible changes in catalytic performance at the films treated for 2 s and 10 s, while improvement in catalytic activity at the film treated for 15 s. However, this increase may result from the formation of many cracks in the films due to the treatment, instead of sulfur vacancies. SEM images of FIG. 19C clearly indicate the presence of holes (dark areas) with missing materials in the film whose sizes and densities increases with the treatment duration. The cracks may provide edge sites to be catalytic active sites for the HER. To further confirm that the increase of catalytic activity not resulting from the formation of sulfur vacancies, the present inventors perform the repair of sulfur vacancies at the films treated for 15 s, and found negligible change afterward.

Example 3: Mechanism of Activating MoS₂for PH-Universal Hydrogen Evolution

The present inventors demonstrated that the intercalation of hydrogen could enhance the catalytic performance of MoS₂with high crystalline quality. This could be due to the electronic structure change of active sites caused by the intercalation of hydrogen with MoS₂. After hydrogen intercalation, the Tafel slope decreases but the current density increase order of magnitude. The activated monolayer MoS₂in acid possess reasonable high catalytic activity in neutral and mild base electrolytes. This understanding could help the design of high-performance MoS₂HER catalyst for medias with a wide range of pH values.

The hydrogen evolution from water represents a key step towards the utilization of clean energy, but its implementation has been delayed by the lack of low-cost high-performance catalysts. While noble metals such as Pt may provide the best catalytic activity for the hydrogen evolution reaction (HER), they are too expensive and scarce to be useful for mass production of hydrogen. Transition metal chalcogenide materials, such as molybdenum disulfide (MoS₂), are widely considered to be a promising low-cost alternative to Pt. These materials are earth abundant and able to provide good catalytic performance for the HER. However, despite considerable amount of effort, the catalytic efficiency of MoS₂is still way inferior to that of Pt. The major reason for that lies in the limited fundamental understanding for the catalytic hydrogen evolution at MoS₂.

One particularly puzzling issue is the activation process of MoS₂for catalytic hydrogen evolution. It has been observed that the catalytic performance of MoS₂may gradually increases with electrochemical scanning such as cyclic voltammetry in a range of mildly negative (vs. RHE) voltages and eventually turns to be stable. The process could take thousands even tens of thousands of cycles and the final performance may be orders of magnitude better than the original one. It is more prominent for MoS₂with high crystallinity than amorphous MoS₂.

The present inventors have demonstrated that the activation process of MoS₂is essentially a process of intercalation of hydrogen ion. Although hydrogen intercalation was found during the electrolysis process of some layered materials (TaS₂) in 2 H phase, the effect of such hydrogen intercalation on the catalytic performance of hydrogen evolution has not been discussed. The intercalated hydrogen ion improves the electrical conductivity between the monolayer and the substrate. The present inventors also demonstrated that the MoS₂intercalated hydrogen ions may show substantially improved catalyst activities in neutral and base solutions. Although the catalytic activity in acid is higher than that in neutral and base media

FIG. 20A shows a typical activation process for monolayer MoS₂films. The catalytic performance of the film shows dramatic increase with the cycling and eventually turns to be stable after 8000 cycles. More specifically, the Tafel slope decreases during the cycling, while the current density substantially change by order of magnitude from 10⁻⁴to 10⁻³mA/cm². The present inventors could confirm that the increase in catalytic performance does not result from change in the composition or crystalline structure (like phase change) of MoS₂, as XPS (FIG. 20D) and Raman spectrum (FIG. 20E) remain pretty much unchanged after the cycling. However, the PL (FIG. 20C) of the MoS₂shows an increase in efficiency and blueshifts in wavelength after the cycling. This indicates that the MoS₂is p-doped during the cycling process. The p-doping is also supported by Raman measurement, which shows blueshifts in the A_1gpeak. It has been well known that the positions of the PL and the A_1gpeak may blueshifts in the case of p-doping. Similar increase of catalytic performance can be also found at single crystalline MoS₂flakes, although the number of cycles required to get the stable performance may be smaller than that for the films.

The p-doping of MoS₂during the electrochemical process strongly suggests the intercalation of hydrogen ions underneath the monolayer films. The present inventors have previously demonstrated that hydrogen ion intercalated underneath the MoS₂may provide p-doping and enable increase and blueshift in the photoluminescence. To further confirm the correlation between the intercalation of hydrogen ion and the improvement of catalytic performance, a separate MoS2 film was treated using TFSI solutions. The inventors have previously demonstrated that the treatment of TFSI solution (immersing the film into TFSI solution followed with mild annealing) may lead to intercalation of hydrogen ions underneath of the MoS₂. For the convenience of comparison, the film treated by the TFSI is grown at the same condition as the one electrochemically cycled. As-transferred film with no electrochemical processing was treated multiple times using TFSI, and the catalytic performance, PL, and Raman of the film was monitored after every treatment, as shown in FIGS. 21A-21E. The present inventors could find that the film shows an increase in catalytic performance with the number of TSFI treatment and eventually turns to be stable. Similar to the electrochemical cycling, the Tafel slope decreases while the current density may increase by orders of magnitude. At the same time, the PL intensity continuously increases and blueshifts with every treatment. The present inventors have also confirmed that the composition and structure of the film do not change during the treatment as indicated by XPS and Raman measurement. The similarity of the effects of electrochemical processing and TFSI treatment confirm that the electrochemical processing may induce the intercalation of hydrogen ions as the TFSI treatment.

The hydrogen ion may not just intercalate between the substrate and the monolayer MoS₂. It can also into the interlayer spacing of MoS₂. The present inventors have performed electrochemical cycling and TFSI treatment at bilayer MoS₂films. The present inventors first treated the bilayer films multiple times using TFSI till the catalytic performance gets to be stable, and then perform cyclic voltammetry. The present inventors have previously demonstrated that it is very difficult for the hydrogen ion to intercalate between two 2D materials layers with the TFSI treatment. With this said, the improvement in the catalytic performance by the treatment of TFSI is likely caused by the intercalation of hydrogen ions between the substrate and the bottom layer, while the improvement caused by the CV is due to the intercalation of hydrogen ions between the top and the bottom layers. As a further evidence to confirm the interlayer intercalation, The PL and Raman of the entire process was monitored. The present inventors did find further increase in the PL and blueshift of the PL and Raman peaks after the CV processing. In experiments, it was found that the CV process may better facilitate the hydrogen ion intercalation, as negative potentials could better attract the interaction of positively charged hydrogen ions. In contrast, in the TFSI treatment, the interaction is simply driven by diffusion of hydrogen ions caused by concentration gradient. As a further evidence for the better capabilities of the CV process to facilitate the intercalation of hydrogen ions, the present inventors found less cycling number is required for the film to get to the stable performance when more negative potentials were applied.

The intercalation of hydrogen ion may account for the improved catalytic performance of MoS₂with thicker dimension. Crystalline MoS₂nanosheets in thickness of tens of nanometers and in lateral size of hundreds of nanometers were synthesized following a process developed by the inventors, and then its catalytic performance as a function of electrochemical cycling were examined. Similar to monolayer and bilayer MoS₂, the catalytic activity of the MoS₂nanosheets substantially increases with the electrochemical cycling in acidic electrolytes, and eventually tends to be stable as shown in FIG. 25A. And smaller nanosheets usually require less cycles to reach stable performance. The present inventors have confirmed no observerable change in the crystalline structures and composition of the nanosheets as shown in FIG. 25B. However, XRD measurement shows that the (002) peak slightly shifts toward smaller angle and shows decrease in the intensity as shown in FIG. 25C. Quantitative analysis indicates that the interlayer spacing increases by around 0.01 nm on average. From the measurement, the average interlayer spacing change from 0.6 nm to 0.9 nm was estimated. This again indicates the interaction of hydrogen ions.

In conclusion, the intercalated hydrogen ion could enhance the catalytic performance of 2 H MoS₂for hydrogen evolution. The present inventors have demonstrated sulfur vacancy is another major catalytic active site. The electronic structure of sulfur vacancies could be changed dramatically by hydrogen intercalation. Unlike the lithium interaction, the phase change contributes the increase of catalytic performance. The hydrogen intercalation would not convert 2 H phase to 1 T phase. The MoS₂intercalated by hydrogen ion also shows reasonable performance in neutral and base solutions.

Methods

Synthesis and transfer of monolayer MoS₂films and flakes: MoS₂thin films were synthesized using a self-limiting chemical vapor deposition process (Yu, Sci Rep-Uk 3, 2013). Briefly molybdenum chloride (MoCl₅) powder (99.99%, Sigma-Aldrich) was placed at the center of the furnace and sulfur powder (Sigma-Aldrich) were placed at the upstream entry of the furnace. Receiving substrates (sapphire) were placed in the downstream of the tube. Typical conditions include a temperature of 850° C., a flow rate of 50 sccm, and a pressure around 2 Torr. Monolayer MoS₂flakes were grown using a different chemical vapor deposition (CVD), in which MoO₃(99.99%, Sigma-Aldrich) instead of MoCl₅was used as the precursor. Typical growth was performed at 750° C. under a flow of Ar gas in rate of 100 sccm and ambient pressure (Yu, Nano Letters 15, 486-491, 2014).

The transfer of the monolayers followed a surface-energy-assisted transfer approach (Guararslan, et al., ACS Nano 8, 11522-11528, 2014). Briefly, a layer of polystyrene (PS) was spin-coated on the as-grown monolayers. A water droplet was then dropped on the top of the polymer. Due to the different surface energies of the monolayer and the substrate, water molecules could penetrate under the monolayer, resulting the delamination of the PS-monolayer assembly. The polymer/monolayer assembly was picked up with a tweezers and was transferred to glassy carbon substrates. Finally, PS was removed by rinsing with toluene several times.

Structure and Composition Characterizations: Raman measurements were carried out by a Horiba xPlora system equipped with an excitation wavelength at 532 nm. AFM measurements were performed at a Veeco Dimension-3000 atomic force microscope. XPS measurements were carried out at X-ray photoelectron spectroscope (SPECS System with PHOIBOS 150 analyzer using an Mg α X-ray source). Magnetization measurements were performed at 350 K in a Quantum Design® MPMS SQUID VSM. The magnetic field was applied in the plane of the samples that were mounted on a diamagnetic quartz sample holder.

Electrochemical Characterizations: The electrochemical characterization was performed in 0.5 M H₂SO₄using a CH Instrument electrochemical analyzer (Model CHI604D) with a Pt-wire counter electrode and a saturated calomel reference electrode (SCE). Nitrogen gas was bubbled into the electrolyte throughout the experiment. Calibration of the reference electrode for the reversible hydrogen potential was performed using a platinum (Pt) disk as working electrode and a Pt wire as counter electrode in 0.5 M H₂SO₄. The electrolyte was purged with ultrahigh purity hydrogen (Airgas) during the measurement. The potential shift of the SCE is calibrated to be −0.262 V vs. RHE. Typical electrochemical characterizations of the monolayers was performed using linear sweeping from +0 V to −0.5V (vs. RHE) with a scan rate of 5 mV/s. The electrolyte resistance and capacitance of the electrocatalysts were characterized using electrochemical impedance spectroscopy (EIS). The AC impedance is measured within the frequency range of 106 to 1 Hz with perturbation voltage amplitude of 5 mV. An equivalent Randles circuit model was fit to the data with ZSimpWin software to determine the system resistance and capacitance.

The electrochemical characterization in neutral solution was performance in 1M PBS solution. The mild base solution (pH=12) was prepared by mixing 327.225 mL 0.05M Na₂HPO₄and 172.775 mL 0.1M sodium hydroxide (NaOH). Before testing in base or neutral solution, monolayer MoS2 was cycled to stable in acid. Typical electrochemical characterization was performance using linear sweeping from 0V to −1.0V (vs. RHE) with a scan rate of 5 mV/s.

Example 4: Tuning the Catalytic Performance of MoS₂for Hydrogen Evolution Through Supporting Substrates

Due to the small thickness (0.7 nm) of monolayer MoS₂, its electronic and optical properties are substantially affected by supporting substrates. The choice of various substrate could change the catalytic activity of MoS₂for hydrogen evolution dramatically through the influence on the hydrogen adsorption energy of active sites. FIG. 26A shows polarization curves of monolayer MoS₂on different substrates including nickel, gold, glassy carbon, and copper for hydrogen evolution, and FIG. 26B shows Tafel plots of monolayer MoS₂on different substrates including nickel, gold, glassy carbon, and copper for hydrogen evolution. A piece of monolayer MoS₂was synthesized utilizing the self-limiting method. It is uniform across the whole surface. The film was cut into several pieces and perfectly transferred onto different substrate in the same way using a surface-energy-assisted transfer method. The nickel substrate shows better hydrogen evolution performance of MoS₂than other substrates. Gold supporting MoS₂also shows reasonable high performance. Copper and glassy carbon supporting MoS₂show relatively poor catalytic performance.

Example 5: Catalytic Performance of MoS₂for Hydrogen Evolution on a Pt substrate

The inventors demonstrated that monolayer MoS₂films may be improved relative to Pt for HER catalysis when the MoS₂films are disposed on Pt substrates. Pt substrates may be used to provide proper interaction. The Pt substrates may not participate catalytic reactions, but may boost the activity of the MoS₂films by forming a lower interface tunneling barrier and affecting the electronic structure of the MoS₂film, such as through charge transfer. A relatively minimal amount of Pt, for example, a thin Pt layer having about 1 nm thickness, may be enough to improve performances of the MoS₂films.

FIG. 27 shows polarization curves collected from the monolayer MoS₂film on Pt substrates under pH values of 1, 7 and 14. For the convenience of comparison, the results collected from bare Pt substrates and the monolayer MoS₂film on glassy carbon (GC) substrates are also plotted. The films were grown on sapphire substrates using a self-limiting chemical vapor deposition process (Yu, Y. F. et al. Controlled Scalable Synthesis of Uniform, High-Quality Monolayer and Few-layer MoS2 Films Sci Rep-Uk 3, doi:Artn 1866), and then were cut in half for separate transfer onto Pt and glassy carbon substrates using a surface-energy-assisted transfer process. It was confirmed that the transferred films are smooth and continuous and are substantially free of wrinkles and voids. This means that the film and the underlying substrate, which is reasonably smooth, have comparable surface area. It was also confirmed that no change in the composition, crystalline structure, and morphology of the film occurred during the catalytic reaction. Significantly, the MoS₂film on Pt substrates shows better catalytic performance than the bare Pt substrate in all the kinds of media. FIG. 28 shows Tafel plots corresponding to the polarization curves of FIG. 27. Referring to FIG. 28, the monolayer MoS₂film on Pt substrates always have lower Tafel slopes and higher exchange current densities than the bare Pt substrate.

FIG. 29 shows polarization curves of the MoS₂films with different sulfur vacancy densities on Pt substrates. The catalytic performance of the MoS₂films on Pt substrates shows strong dependence on the sulfur vacancy density: the film with ˜10% sulfur vacancies showing better performance than those with either much lower (1% or 3%) or higher (14% or 20%) densities. Accordingly, it will be understood that except the sulfur vacancies, the other parts of the MoS₂films, such as grain boundaries and fully coordinated atoms in the basal plane, may neither be enough catalytically active nor be affected by the sulfur vacancy repair process. Therefore, the strong dependence of the catalytic performance on sulfur vacancies may indicate that the other parts of the MoS₂films, such as grain boundaries and the fully coordinated atoms at the basal plan, have negligible contribution to the observed improved catalytic performance.

While not directly acting as the catalyst, the Pt substrate may substantially boost the catalytic activity of the film, as evidenced by the better catalytic performance of the monolayer film on Pt than on glassy carbon substrates as illustrated in FIG. 27. To better understand the boosting effect, the inventors examined the catalysis of MoS₂films with different numbers of layers on both Pt and glassy carbon substrates. The different numbers of layers provide a way to precisely tune the effect of the substrates. The inventors have confirmed the numbers of the layers of the films with Raman and AFM measurements and also confirmed that all these films bear similar sulfur vacancy densities. FIG. 30A shows polarization curves of the MoS₂films having different numbers of layers on Pt substrates. FIGS. 31A and 31B show polarization curves of the MoS₂films having different numbers of layers on Pt substrates and on glassy carbon (GC) substrates, respectively. The MoS₂films always show better catalytic performance on Pt than on glassy carbon substrates, but the difference may become smaller as the number of layers increases.

FIG. 30B shows exchange current density of the MoS₂films having different numbers of layers on Pt substrates and on glassy carbon (GC) substrates, and FIG. 30C shows Tafel slope of the MoS₂films having different numbers of layers on Pt substrates and on glassy carbon (GC) substrates. The inventors obtained more insight into the substrate effect by separately examining the Tafel slope and exchange change density. The Tafel slope of the film on Pt substrates increases with the number of layers, turning from much lower (at monolayer and bilayer) to comparable (at quadra-layer) when compared to the film on glassy carbon substrates. In contrast, the exchange current densities of the films on both substrates show similar exponential dependence on the number of layers, but their absolute values are differentiated by a reasonably constant ratio in the range of 10-17 regardless the number of layers. The inventors found that the film on Pt substrates also show exchange current densities larger than the counterpart on glassy carbon substrates by a ratio of 10-17 regardless the environmental pH value and the density of sulfur vacancies. The different dependences of the Tafel slope and exchange current density on the number of layers strongly suggest that the two parameters may be affected by the substrates through different mechanisms.

The enhancement effect of Pt substrates for the exchange current density may be explained based on the effect of the substrate on the electron transport at the substrate-MoS₂interface. It has been known that placing MoS₂onto conductive substrates may form a tunneling barrier at the interface. Intuitively, in order to drive the catalytic reaction at the film, electrons must tunnel from the conductive substrate across the interface to the outmost layer of the film. According to previous studies, the exchange current density of MoS₂films is related with the efficiency of the electron tunneling through the interlayer barriers and the barrier at the substrate-MoS₂interface. It has been indicated that the tunneling through the interlayer barriers gives rise to the observed layer dependence of the exchange current density, exponentially decreasing with increase of the layer number (See FIG. 30B). By the same token, the tunneling through the interface barrier is expected to give rise to substrate dependence of the exchange current density. This is supported by the reasonable consistence between the interfacial tunneling efficiency and the observed substrate dependence of the exchange current density. The tunneling efficiency is related with the tunneling barrier V_oand the tunneling distance L as T=e^−2kLand k=(2m₀V_o)^1/2/ℏ, where m₀is the mass of electrons, and ℏ is the Planck's constant. By using reasonable approximation (L=0.31 nm) and the interface tunneling barriers calculated with first principle techniques (Zhong, H. X. et al., Sci Rep. 2016; 6: 21786), the tunneling efficiency at the Pt-MoS₂interface is found higher than that of the GC-MoS₂interface by 13.7 times. This is reasonably consistent with the observed enhancement in exchange current density (10-17 folds) by Pt substrates compared to glassy carbon substrates. Additionally, the interface tunneling barrier is reported independent of the number of layers and expected to be insensitive to the sulfur vacancy density as well as environmental pH values. This matches the observed independence of the enhancement effect in exchange current density from pH values, numbers of layers, and sulfur vacancies, which further confirms the correlation.

The effect of substrates on the Tafel slope may be correlated to the change in the electronic structure of MoS₂induced by the substrate. It has been well known that substrates may affect the electronic structure of MoS₂and its active sites, for instance, through charge transfer, which may subsequently affect the catalytic activity. This substrate effect mainly affects the Tafel slope but not the exchange current density. From both intuitive perspective and theoretical simulation, the substrate effect on the electronic structure may rapidly decrease as the number of layers increases. It is in stark contrast with the layer independence observed at the substrate-induced enhancement for the exchange current density, but matches the strong substrate dependence of the Tafel slope, which also shows rapidly changing with the number of layers (See FIG. 30B). This substrate effect may usually be elucidated from change in the adsorption free energy of hydrogen atoms at the active sites.

FIG. 32A shows Raman spectra of monolayer MoS₂on a Pt substrate and on a glassy carbon substrate. FIG. 32A shows a smaller intensity ratio of A_1g/E¹_2gand a red shift in A_1gwith respect to the spectrum from the film on glassy carbon substrates. This suggests that Pt may provide more n-doping to MoS₂monolayer as previous studies have demonstrated that n-doping may cause a red shift in A_1gand reduced A_1g/E¹_2gratio. Pt substrates may transfer more electrons (n-doping) to the MoS₂film than glassy carbon substrates, which is expected to give rise to better catalytic activity as n-doping is known able to increase the catalytic activity of MoS₂. FIG. 32B shows XPS spectra of the monolayer MoS₂on a Pt substrate and on a glassy carbon substrate. The inventors calculated the adsorption free energies of hydrogen atoms at the sulfur vacancies in the MoS₂film on different substrates. The calculation indeed indicates higher catalytic activities at the MoS₂film on Pt substrates as its adsorption free energy of hydrogen atoms at sulfur vacancies is closer to zero than that of the film on glassy carbon substrates (−0.069 eV vs. −0.087 eV).

The inventors have demonstrated better HER catalysis at monolayer MoS₂films than Pt. This is achieved by optimizing the density of sulfur vacancies in the film and leveraging the boosting effect of proper substrate interactions. The substrate does not participate the reaction, but it could significantly boost the activity of the film via forming a low interface tunneling barrier and affecting the electronic structure of the film, for instance, through charge transfer. Pt substrates were used to illustrate this notion. FIG. 33A shows polarization curves collected from monolayer MoS₂films on thin Pt layers deposited on nickel substrates in acidic media (pH=1). Two different thickness of the Pt layers were tested, 1 nm and 10 nm. Inset shows a schematic illustration for the film on thin Pt substrates. FIG. 33B shows stable catalytic performance of monolayer MoS₂films on Pt substrates with continuous reaction under with a density of >20 mA/cm²for more than two months in acidic media (pH=1). The potential is maintained at −0.05 V (vs. RHE) during the reaction. Referring to FIG. 33A, even a minimal amount of Pt, such as a thin Pt layer having about 1 nm thickness, may be enough to enable the superior catalytic performance in the film. FIG. 33B, the superior catalytic performance shows remarkable stability with no decrease after continuous reaction for more than two months. Although only Pt substrates have been described herein, other substrates with capabilities of forming low interface tunneling barriers and properly affecting the electronic structure of the film may also be used to enable improved catalysis at MoS₂.

Material and Methods

Synthesis and transfer of MoS₂films: MoS₂thin films were synthesized using a self-limiting chemical vapor deposition process (Yu, et al., Sci Rep-Uk 2013, 3). Briefly molybdenum chloride (MoCl₅) powder (99.99%, Sigma-Aldrich) was placed at the center of the furnace and sulfur powder (Sigma-Aldrich) at the upstream entry of the furnace. Receiving substrates (sapphire) were placed in the downstream of the tube. Typical conditions include a temperature of 850° C., a flow rate of 50 sccm, and a pressure around 2 Torr. The film with different densities of sulfur vacancies were grown by varying the growth temperature in the range of 700-900° C. The number of layers of the films were controlled by controlling the amount of precursor. The transfer of the monolayers followed a surface-energy-assisted transfer approach (Gurarslan, A. et al. Surface-Energy-Assisted Perfect Transfer of Centimeter-Scale Monolayer and Few-Layer MoS2 Films onto Arbitrary Substrates. ACS nano 8, 11522-11528 (2014)). Briefly, a layer of polystyrene (PS) was spin-coated on the as-grown monolayers. A water droplet was then dropped on the top of the polymer. Due to the different surface energies of the monolayer and the substrate, water molecules could penetrate under the monolayer, resulting the delamination of the PS-monolayer assembly. The polymer/monolayer assembly was picked up with a tweezers and was transferred it to either Pt or glassy carbon substrates. Finally, PS was removed by rinsing with toluene several times.

Repair of sulfur vacancies: Sulfur repair was conducted following a process reported previously (Qiu, H. et al.Nat Commun 2013, 4). Basically, monolayer MoS₂was dipped in a solution of 1/15 (volume ratio) MPS (Sigma-Aldrich)/dichloromethane (Sigma-Aldrich) for 48 hours in a dry glove box. The samples were then rinsed thoroughly with dichloromethane and isopropanol (Sigma-Aldrich), and blown dry with N₂. Finally the samples were annealed in a flow of Ar (with 5% H₂) at 350° C. for 20 minutes.

Structure and Composition Characterizations: Raman and PL measurements were carried out by a Horiba xPlora system equipped with an excitation wavelength at 532 nm. AFM measurements were performed at a Veeco Dimension-3000 atomic force microscope. XPS measurements were carried out at X-ray photoelectron spectroscope (SPECS System with PHOIBOS 150 analyzer using an Mg Kα X-ray source).

Electrochemical Characterizations: The electrochemical characterization was performed in 0.5 M H₂SO₄using a CH Instrument electrochemical analyzer (Model CHI604D) with a saturated calomel reference electrode (SCE). A Pt mesh (2.54 cm×2.54 cm) or a graphite rod was used as counter electrode, and the results did not show difference for each of the counter electrodes. Nitrogen gas was bubbled into the electrolyte throughout the experiment. The potential shift of the SCE is calibrated to be −0.262 V vs. RHE. Typical electrochemical characterizations of the monolayers were performed using linear sweeping from +0 V to −0.6V (vs. RHE) with a scan rate of 5 mV/s. The electrolyte resistance and capacitance of the electrocatalysts were characterized using electrochemical impedance spectroscopy (EIS). The AC impedance was measured within the frequency range of 10⁶to 1 Hz with perturbation voltage amplitude of 5 mV (typical EIS measurement results can be seen in FIG. 6). An equivalent Randles circuit model was fit to the data to determine the system resistance and capacitance.

DFT Computation: The geometry optimization and following electronic property calculations are all performed with the projected augmented wave (PAW) method implemented in the Vienna ab initio simulation package (VASP). The Perdew-Becke-Ernzerhof (PBE) exchange-correlation functional is used in all calculations, along with the DFT-D2 correction for molecular interactions. The plane wave cut off energy is set to 500 eV. In order to find structures with small lattice mismatch, the present inventors calculated the improved or optimized structures of the MoS2 monolayer and bulk Pt. The in-plane lattice constant of MoS2 monolayer is 3.182 Å (21×21×1 gamma-centered k-points). The optimized Pt fcc cell has a lattice constant of 3.977 Å (21×21×21 Monkhorst-Pack k-points). Then a 4-layer 4×3 rectangular Pt (111) surface is created, with the bottom layer fixed. The MoS2 monolayer is stretched by 2.0% to fit the Pt surface. A vacuum buffer of 15 Å thickness is added to the model to prevent interactions between neighbor slabs. The present inventors chose a similar shape as what has been used in the situation for Pt to create the cell for glassy carbon, but in this case, there can be zero lattice mismatch. For the calculations, the criteria for energy convergence is set to be smaller than 10⁻⁵eV. In geometry optimization, the cell shape is fixed and the force is required to be less than 0.01 eV/A. 2×2×1 Monkhorst-Pack k-points are used for atom coordinate relaxation and 6×6×1 for property calculations. Monopole corrections are used in all calculations.

Example 6: Catalytic Performance of MoS₂Including Dopants

The inventors developed a chemical vapor deposition (CVD) process to form a single layer MoS₂including dopants atoms (e.g., single atom Ni dopants) and demonstrated that the single layer MoS₂including dopants atoms show better performance than the single layer MoS₂without dopants atoms.

Material and Methods

MoS₂thin films with Ni dopants were synthesized using a modified self-limiting chemical vapor deposition process (Yu, Sci Rep-Uk 3, 2013). In this process, molybdenum chloride (MoCl₅) was used as the Mo precursor, and Nickel acetylacetonate (e.g., Ni(acac)₂) was used as the Ni precursor. Briefly, molybdenum chloride powder (99.99%, Sigma-Aldrich) and Ni(acac)₂were placed at the center of the furnace, and sulfur powder (99%, Sigma-Aldrich) was placed at the upstream entry of the furnace. Receiving substrates (sapphire) were placed in the downstream of the tube. Ni dopant concentration can be controlled by tuning the amount of Ni(acac)₂. Typical conditions include a temperature of 850° C., a flow rate of 50 sccm, and a pressure around 2 torr.

The transfer of the monolayer followed a surface-energy-assisted transfer approach discussed in Gurarslan, et al., ACS Nano 2014, 8, 11522. Briefly, a layer of polystyrene (PS) was spin-coated on the as-grown monolayer. A water droplet was then dropped on the top of the polymer. Due to the different surface energies of the monolayer and the substrate, water molecules could penetrate under the monolayer, resulting the delamination of the PS-monolayer assembly. The polymer/monolayer assembly was picked up with a tweezers and was transferred to either nickel or glassy carbon substrates. Finally, PS was removed by rinsing with toluene several times.

Electrochemical Characterizations

The electrochemical characterization was performed in 0.5 M H₂SO₄using a CH Instrument electrochemical analyzer (Model CHI604D) with a saturated calomel reference electrode (SCE). A Pt mesh (2.54 cm×2.54 cm) or a graphite rod was used as counter electrode, and the results did not show difference for each of the counter electrodes. Nitrogen gas was bubbled into the electrolyte throughout the experiment. The potential shift of the SCE is calibrated to be −0.262 V vs. RHE. Typical electrochemical characterizations of the monolayers were performed using linear sweeping from +0 V to −0.6V (vs. RHE) with a scan rate of 5 mV/s. The electrolyte resistance and capacitance of the electrocatalysts were characterized using electrochemical impedance spectroscopy (EIS). The AC impedance was measured within the frequency range of 106 to 1 Hz with perturbation voltage amplitude of 5 mV. An equivalent Randles circuit model was fit to the data to determine the system resistance and capacitance.

Results and Discussion

FIG. 34A shows Mo peaks of XPS for pure MoS₂and Ni doped MoS₂. FIG. 34B shows Ni peaks of XPS for Ni doped MoS₂. FIG. 34C shows Raman spectra for pure MoS₂and Ni doped MoS₂. FIG. 34D shows polarization curves of MoS₂with and without Ni dopants. Bare glassy carbon shows no catalytic activity toward hydrogen evolution reaction. MoS₂with Ni dopants shows better performance than pure MoS₂.

To prove the doping was successful, XPS and Raman measurements were performed on the as-synthesized monolayer MoS₂with Ni dopants. Mo peaks of XPS for MoS₂with and without Ni dopants are shown in FIG. 34A. Comparing with pure MoS₂, the Mo4⁺peaks shift to lower binding energy for MoS₂doped with Ni atoms and thus successful doping of Ni was confirmed. Referring to FIG. 34B, Ni atoms were detected on the samples, and the Ni peaks (855.0 eV) of MoS2 with Ni dopants indicates the formation of NiMoS. Referring to FIG. 34C, Raman spectra of MoS₂with and without Ni dopants are very similar. However, Ni doped MoS₂shows more distorted structure demonstrated by the more peaks between 100 cm⁻¹to 300 cm⁻¹. FIG. 34D shows polarization curves from the monolayer MoS₂film with and without single Ni atom dopants on glassy carbon. For the convenience of comparison, the results collected from bare glassy carbon and the platinum are also plotted. Glassy carbon shows negligible catalytic activity for hydrogen evolution reaction. Comparing with pure MoS₂(i.e., MoS₂film without single Ni atom dopants), MoS₂with Ni dopants shows significant improvement of the catalytic activity.

FIG. 35A shows catalytic activity according to Ni precursor amount used in synthesis. FIG. 35B catalytic performance of a bare Pt plate and Ni doped MoS₂on a Ni substrate.

The catalytic activity of Ni doped monolayer MoS₂can be controlled by tuning the Ni precursor amount used in the synthesis process. If the Ni dopant amount is small, the active sites could be less. While if the Ni dopant amount is too large, the crystalline quality of monolayer MoS₂could become worse, which can affect the catalytic activity of the films. The catalytic dependence on Ni precursor amount used in the synthesis process is shown in FIG. 35A. The inventors demonstrated that supporting substrate may affect catalytic activities of two dimensional films, such as MoS₂and WS₂. For Ni doped MoS₂, the same phenomena was observed. It was found that a Ni substrate could promote the catalytic activity of MoS₂. When the Ni doped MoS₂film with the best performance on glassy carbon was transferred to a Ni substrate, the catalytic activity became much better, even better than bare Pt as shown in FIG. 35B.

The above-disclosed subject matter is to be considered illustrative, and not restrictive, and the appended claims are intended to cover all such modifications, enhancements, and other embodiments, which fall within the true spirit and scope of the inventive concept. Thus, to the maximum extent allowed by law, the scope is to be determined by the broadest permissible interpretation of the following claims and their equivalents, and shall not be restricted or limited by the foregoing detailed description.

CATALYSTS FOR HYDROGEN EVOLUTION REACTION INCLUDING TRANSITION METAL CHALCOGENIDE FILMS AND METHODS OF FORMING THE SAME

Information

Publication Number

Date Filed

Date Published

Inventors

Original Assignees

CPC

International Classifications

Abstract

Description

Claims

RELATED APPLICATION

STATEMENT OF GOVERNMENT SUPPORT

PCT Information

Provisional Applications (1)