Hybrid catalyst with monoclinic MoTe2 and platinum for efficient hydrogen evolution

Transition metal dichalcogenides (TMDs) are considered as promising catalysts for the hydrogen evolution reaction (HER) owing to their abundant active sites such as atomic vacancies and step edges. Moreover, TMDs have polymorphism, which has stimulated extensive studies on tuning of surface electronic structures for an active HER. The polymorphism in TMDs provides an opportunity for new hybrid catalysts with TMDs and other catalytic metals via surface engineering that can create a novel functional surface of the catalytic electrode for the active HER. Here, we report a hybrid catalyst with monoclinic MoTe2 and platinum (Pt) for the HER. Pt atoms were chemically bound to the surface of monoclinic MoTe2 that has an atomically distorted lattice structure, which produces a distinct Pt-Te alloy layer. The Pt/MoTe2 hybrid catalyst exhibits an active HER with a Tafel slope of 22 mV per decade and an exchange current density of 1.0 mA/cm, which are the best values among those reported for TMD-based catalysts. The use of minimum amount of Pt on atomically distorted metallic TMDs realizes rich catalytic active sites on large basal planes for efficient hydrogen production. © 2019 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). https://doi.org/10.1063/1.5094957., s Recent progress in two-dimensional energy materials has been highlighted due to their broad applications in photocatalysis, photovoltaics, batteries, supercapacitors, and electrochemical catalysis. Electrochemical catalysts based on transition metal dichalcogenides (TMDs) have been intensively studied in hydrodesulfurization, hydrogenation, and the hydrogen evolution reaction (HER). In particular, most semiconducting group 6 TMDs such as MoS2, MoSe2, and WS2 have shown excellent electrochemical catalytic performances for the HER at their highly conducting edges or atomic defect sites. Despite the active HER performance at certain atomic sites on the surface, the limited surface area of such atomically defined active sites remains an issue, and the industry of hydrogen production requires more efficient and stable electrochemical reactions with reactive sites on the entire surface area. However, the semiconducting basal plane of TMDs without edges or defects has been found to be electrochemically inactive for the HER. Extensive efforts have been made to exploit the whole basal plane of TMDs for the HER by fabricating hybrid catalysts. For example, various catalytic nanoparticles, such as Pt, Pd, and Cu, have been decorated on semiconducting TMD surfaces to make the entire basal planes more active. Moreover, in the HER, hybrid TMD catalysts with low dimensional electrodes such as graphene and carbon nanotubes have been suggested as a solution to resolve the high contact resistance between the semiconducting TMDs and metal electrodes. The TMD-based catalysts with external nanostructures (nanoparticles, graphene, or carbon nanotubes) or their hybrids have shown improved HER performances than pristine TMDs, but the functional role of the substrate in the simply combined catalyst geometry could not be studied rigorously by density APL Mater. 7, 071118 (2019); doi: 10.1063/1.5094957 7, 071118-1

Recent progress in two-dimensional energy materials has been highlighted due to their broad applications in photocatalysis, photovoltaics, batteries, supercapacitors, and electrochemical catalysis. [1][2][3][4][5][6] Electrochemical catalysts based on transition metal dichalcogenides (TMDs) have been intensively studied in hydrodesulfurization, hydrogenation, and the hydrogen evolution reaction (HER). [7][8][9] In particular, most semiconducting group 6 TMDs such as MoS 2 , MoSe 2 , and WS 2 have shown excellent electrochemical catalytic performances for the HER at their highly conducting edges or atomic defect sites. [10][11][12] Despite the active HER performance at certain atomic sites on the surface, the limited surface area of such atomically defined active sites remains an issue, and the industry of hydrogen production requires more efficient and stable electrochemical reactions with reactive sites on the entire surface area. However, the semiconducting basal plane of TMDs without edges or defects has been found to be electrochemically inactive for the HER. 13,14 Extensive efforts have been made to exploit the whole basal plane of TMDs for the HER by fabricating hybrid catalysts. For example, various catalytic nanoparticles, such as Pt, Pd, and Cu, have been decorated on semiconducting TMD surfaces to make the entire basal planes more active. 15,16 Moreover, in the HER, hybrid TMD catalysts with low dimensional electrodes such as graphene and carbon nanotubes have been suggested as a solution to resolve the high contact resistance between the semiconducting TMDs and metal electrodes. [17][18][19] The TMD-based catalysts with external nanostructures (nanoparticles, graphene, or carbon nanotubes) or their hybrids have shown improved HER performances than pristine TMDs, but the functional role of the substrate in the simply combined catalyst geometry could not be studied rigorously by density ARTICLE scitation.org/journal/apm functional theory calculations, and thus, further improvement of the HER has been limited. Motivated by the polymorphism of group 6 TMD single crystals, structural phase engineering has also been used to achieve higher HER performances. In the case of MoS 2 , HER performance is improved when their most stable (hexagonal) semiconducting phase is converted to the metastable (monoclinic or octahedral) metallic phase. [20][21][22] Experimental and theoretical studies have revealed that effective hydrogen evolution occurs at chalcogen atomic sites on the whole basal plane in the metallic phase, which enhances the total catalytic activity of the TMDs. Nevertheless, the HER performance with metallic TMDs is not high enough for real applications (compared to Pt or other hybrid catalysts), and the metallic phase is metastable, inevitably implying a stability issue in the HER. [23][24][25] In contrast to other group 6 TMDs mostly investigated for the HER, MoTe 2 possesses two stable structures under ambient conditions: semiconducting hexagonal (2H) and metallic monoclinic (1T ′ ) phases that can be selectively synthesized as high quality single crystals. 26 While 2H-and 1T-MoS 2 can be controlled by chemical lithium intercalation, the metallic 1T ′ -MoTe 2 can be converted from its semiconducting 2H phase via laser-illumination, and strain and Te vacancy creation, making the 1T ′ phase robust against thermal heating or aging. [27][28][29] Recently, an atomic-scale study on the HER with the metallic MoTe 2 single crystals has been reported by combining DFT calculations and atomic microscopy with the electrochemical measurement. 30,31 The stable metallic 1T ′ -MoTe 2 , allowing efficient charge transfer at the surface, would be promising for the HER, but the HER performance of 1T ′ -MoTe 2 was not on par with that of the conventional Pt catalyst.
In order to fully exploit the whole metallic basal planes as active sites for the HER performance, we propose a Pt/MoTe 2 hybrid catalyst where metallic basal planes of MoTe 2 , having an atomically distorted lattice structure, are coated by Pt atoms via electrochemical activation. The self-structuring of Pt nanoparticles on the carbonbased working electrodes has been observed during electrochemical activation when Pt is used as the counter electrode for the HER. [32][33][34] The dissolution of Pt atoms in the electrolyte leads to the growth of Pt on the MoTe 2 surface by a subsequent reduction process. This idea provides a breakthrough in two aspects: (1) a new chemical state of Pt can be formed on the distorted 1T (1T ′ ) MoTe 2 surface and (2) the stability can be improved by using a well-defined stable metallic group 6 TMD (1T ′ -MoTe 2 ), compared to that of metastable and inhomogeneous 1T-MoS 2 . Our Pt/MoTe 2 hybrid catalyst shows an excellent HER performance with a Tafel slope of 22 mV per decade and an exchange current density of 1.0 mA/cm 2 , which are the best values among those reported for TMD-based catalysts.
We prepared a Pt/MoTe 2 hybrid catalyst using the electrochemical method described in Fig. 1(a). The details of experiment are described in the supplementary material. The electrochemical method realized the deposition of a thin Pt layer on the

ARTICLE
scitation.org/journal/apm metallic MoTe 2 surface without changing the bulk lattice structures (the structural phase) of the substrate 1T ′ -MoTe 2 . Figure 1(b) supports the identical lattice vibration modes of the MoTe 2 by Raman spectra. We used an excitation laser with a power of 2.6 mW and a wavelength of 532 nm for the Raman study, which shows the same Raman modes of Ag and Bg before and after the electrochemical treatment (thin Pt layer deposition) on the monoclinic MoTe 2 single crystal. This demonstrates the stability of 1T ′ -MoTe 2 against additional electrochemical fabrication processes and the suitability as a new metallic TMD substrate for hybrid catalysts. Accordingly, we expected a new chemical state of the thin Pt film on the stable, atomically distorted 1T ′ -MoTe 2 substrate, which could crucially increase the reactive surface sites for the HER.
Although the bulk crystalline structure of monoclinic MoTe 2 was retained in the Pt/MoTe 2 hybrid catalyst [Raman spectra in Fig. 1(b)], the morphology of the pristine MoTe 2 was modified by the electrochemical deposition of Pt on the MoTe 2 . The surface change is exhibited in the scanning electron microscope (SEM) images in Fig. 1(c); a clean and flat surface of pristine monoclinic MoTe 2 [left picture in Fig. 1(c)] was changed to a corrugated surface with the Pt atoms (right picture) on the MoTe 2 . We note that, unlike former self-structuring of Pt nanoparticles on TMDs, 16,35 Pt nanoparticles were not observed in the SEM; our electrochemical deposition provided a uniform and well-wetted Pt film on the basal plane of the MoTe 2 . We explain the formation of the wetted film, rather than nanoparticle formation as previous studies reported, by a unique interaction (a relatively strong bonding) between Pt and distorted chalcogen atoms (Te) in the Pt/MoTe 2 . More evidence will be discussed with a chemical state study and corresponding theoretical calculations on the hybrid catalyst.
A progressive improvement of HER performance, as the coverage of the Pt/MoTe 2 hybrid increases on the catalyst surface, is described in Fig. 2(a). Starting from the lowest HER performance with pristine monoclinic MoTe 2 [orange curve in Fig. 2(a)], the Pt/MoTe 2 hybrid catalyst produced higher HER performances, as indicated by the magenta, blue, and red curves in Fig. 2(a). The Tafel slope [22 meV per decade, Fig. 2(b)] and the exchange current density (∼1.0 mA/cm 2 ) in the red curve [ Fig. 2(a)] are better than those of Pt or any other TMD-based catalysts. Other catalytic activities, such as the reversible hydrogen electrode (RHE) potential for a current density of −10 mA/cm 2 , the Tafel slope, and the exchange current density of MoTe 2 , Pt, and Pt/MoTe 2 hybrid catalysts are summarized in Table I. The exchange current density was derived from the Tafel curves [ Fig. 2(b)] using a linear extrapolation to an overpotential of 0 mV. Figure 2(c) shows that the Pt/MoTe 2 hybrid catalyst is more stable than a reference Pt film on SiO 2 under our HER condition over 8000 s. This indicates that the adhesion of Pt on MoTe 2 is more stable or stronger than that from the physisorption of Pt on SiO 2 ; Pt-MoTe 2 has chemisorption characteristics or an alloy formation, leading to different surface chemistry from bulk Pt.
To investigate the surface chemical state of the Pt/MoTe 2 hybrid catalyst, we conducted X-ray photoemission spectroscopy (XPS) with the pristine MoTe 2 [resulting in the orange curve in Fig. 2(a)], the as-prepared Pt/MoTe 2 hybrid catalyst [resulting in the red curve in Fig. 2(a)], and the Pt/MoTe 2 hybrid catalyst after the stability test [ Fig. 2(c)]. Figure 3 shows the core levels of Mo 3d, Te 3d, and Pt 4f electrons by the XPS with fitted lines by Doniac- Sunsic   FIG. 2. HER measurements. (a) Polarization curves of bulk Pt rod (green, as a reference), Pt film deposited on SiO 2 (black, as a reference), pristine MoTe 2 (orange), and Pt/MoTe 2 hybrid catalysts with various treatment times from 7 h to 21 h (magenta, blue, and red). (b) Tafel plots obtained from the polarization curves in Fig. 2(a) Fig. 3(a)]. Pt 4f electrons appear in the Pt/MoTe 2 [middle curve in Fig. 3(b)], but with chemical states slightly different from the Pt metal [top curve in Fig. 3(b)]. The broadened and asymmetric binding energy features of Pt 4f electrons in the Pt/MoTe 2 hybrid catalyst demonstrate the presence of Pt and Pt-Te alloy layers on the surface, which is a key to understand the hybridization of Pt and monoclinic MoTe 2 .
The deconvolution of the Pt 4f spectrum in Fig. 3(b) clearly shows double peak features (detailed fitting parameters are given in Table II). Considering the reference XPS of the Pt metal [top curve in Fig. 3(b)], the Pt 4f peak located at a binding energy (BE) of 70.74 eV [fitted by a green line in the middle curve in Fig. 3(b)] is assigned to the signal of the pristine Pt metal with a similar Lorentzian width (L.W.) and asymmetry factor (α) by the Doniac-Sunsic fitting (Table II). However, the Pt 4f peak located at 71.72 eV [fitted by a blue line in the middle curve in Fig. 3(b)] could be assigned to a Pt 2+δ signal. Moreover, the Te 3d peak located at 573 eV [fitted by a blue line in Fig. 3(a)] could be assigned to a Te 2−δ , demonstrating Pt-Te alloy formation (rather than forming Pt nanoparticles 36

ARTICLE
scitation.org/journal/apm  Fig. 3 the Pt-Te alloy rather than Pt nanoparticles as reflected in the XPS measurements (Fig. 3).
Optimizing HER catalysts, a moderate hydrogen adsorption Gibbs free energy change (ΔG H ), close to zero, is critical. In our DFT calculations, the thickness-dependent atomic structure of the Pt/MoTe 2 surface results in a significant variation in the ΔG H . While the ΔG H on the Pt (111) surface was calculated to be −0.14 eV, the ΔG H values of the Pt/MoTe 2 hybrid catalysts with different Pt thicknesses, 0.5 ML, 1 ML, 1.5 ML, and 2.0 ML, were found to be −0.25 eV, 0.08 eV (±0.22 meV), −0.11 eV (±1.75 meV), and −0.32 eV (±72 meV), respectively.
The calculations of ΔG H indicate that the buckled geometry of the metallic MoTe 2 surface provides a unique surface reconstruction of Pt atoms, which maximizes the catalytic activity and results in excellent HER performances (Fig. 2). Based on the performance of Pt/MoTe 2 [red curve in Fig. 2(a)], which is better than that of the reference Pt electrode, we estimate that the coverage (or the thickness) of Pt would be between 1 ML and 1.5 ML. We note that, in contrast to previous TMD-based HER studies that have used unclear active sites, our calculations are convincing given that singlecrystalline MoTe 2 (and its surface) was adopted for fabricating the hybrid catalyst.
We investigated the catalytic performance of a novel Pt/MoTe 2 hybrid catalyst fabricated with a single-crystalline metallic monoclinic MoTe 2 . The Pt/MoTe 2 hybrid catalyst exhibits excellent HER performance, with a Tafel slope of 22 mV per decade and an exchange current density of 1.0 mA/cm 2 , which exceeds the HER performance of previous TMD-based catalysts. The highly improved HER originates from rich active sites of the surface of the Pt/MoTe 2 hybrid catalyst via a unique reconstruction of Pt atoms on the buckled metallic basal plane. First-principles calculations indicate that H atoms can be adsorbed on the reconstructed Pt/MoTe 2 surface with an ideal ΔG H of 0.08 eV. The strategy of the metallic TMDbased hybridized catalyst has potential for superior catalysis with rich active sites as well as high electric conductivity, which are essential for an active HER with a minimum amount of precious catalytic metals.
See supplementary material for the experimental and calculation schemes.