MoS2 Photoelectrodes for Hydrogen Production: Tuning the S-Vacancy Content in Highly Homogeneous Ultrathin Nanocrystals

Abstract

Tuning the electrocatalytic properties of MoS₂ layers can be achieved through different paths, such as reducing their thickness, creating edges in the MoS₂ flakes, and introducing S-vacancies. We combine these three approaches by growing MoS₂ electrodes by using a special salt-assisted chemical vapor deposition (CVD) method. This procedure allows the growth of ultrathin MoS₂ nanocrystals (1-3 layers thick and a few nanometers wide), as evidenced by atomic force microscopy and scanning tunneling microscopy. This morphology of the MoS₂ layers at the nanoscale induces some specific features in the Raman and photoluminescence spectra compared to exfoliated or microcrystalline MoS₂ layers. Moreover, the S-vacancy content in the layers can be tuned during CVD growth by using Ar/H₂ mixtures as a carrier gas. Detailed optical microtransmittance and microreflectance spectroscopies, micro-Raman, and X-ray photoelectron spectroscopy measurements with sub-millimeter spatial resolution show that the obtained samples present an excellent homogeneity over areas in the cm² range. The electrochemical and photoelectrochemical properties of these MoS₂ layers were investigated using electrodes with relatively large areas (0.8 cm²). The prepared MoS₂ cathodes show outstanding Faradaic efficiencies as well as long-term stability in acidic solutions. In addition, we demonstrate that there is an optimal number of S-vacancies to improve the electrochemical and photoelectrochemical performances of MoS₂.

Keywords: Defect Engineering; Electrocatalysis; Molybdenum Disulfide; Salt-Assisted Chemical Vapor Deposition; Sulfur Vacancies; Water Splitting.

Figure 1

(a) AFM image of a MoS₂ layer grown on Si/SiO₂ substrate. (b) Topography distribution histograms obtained from the regions A and B. Values included in the figures indicate the mean values and standard deviations of each peak of the bimodal distribution. (c) STM image (50 × 50 nm²) of a MoS₂ layer grown on HOPG and enlargement of a 20 × 20 nm² region in the same zone. Tunneling parameters: V_s = −2.5 V and I_T = 20 pA.

Figure 2

Optical characterizations of a MoS₂ layer grown on different substrates: Si/SiO₂ (a, b) and fused silica (c, d). (a) Raman spectrum and (b) PL spectrum. (c) Optical density measurements done with a macroscopic setup (millimeter-scale spot size) and with a microtransmittance setup (micrometer-scale spot size) recorded in two different zones of the sample. (d) Differential reflectance measurements acquired with a microreflectance setup (micrometer-scale spot size) in two different zones of the sample, to show the homogeneity in the optical properties of the films.

Figure 3

(a) Picture of the MoS₂ sample grown on Si and mounted on a tantalum sample holder used for XPS measurements. Positions of the zones where Raman mappings were acquired are labeled. (b) Optical microscopy image. The red square shows the region C in (a) in which the Raman mapping was performed. (c) Histogram distributions of Δk values in the Raman spectra for the five regions indicated in (a). (d) Micro-XPS spectra recorded at different zones of a line scan of 7 mm in length along the X axis. Mo 3d and S 2s BE peaks are shown.

Figure 4

(a) XPS spectrum of a MoS₂ sample (MS-1.7) in the region of Mo 3d, obtained at a pass energy of 5 eV: experimental data (dots), single-component fitting curves (colored and filled lines), complete fitting curve (red continuous line), and Shirley-shape background (gray line). (b) Raman spectra for samples grown with and without the use of a hydrogen flow during the salt-assisted CVD. (c) Raman features (Δk values) against stoichiometry (S/Mo ratio) for three different samples grown under different conditions.

Figure 5

(a) Polarization LSV curves of a bare GC and of MoS₂ samples with different S/Mo ratios (2.2, 1.8, and 1.6). The vertical dotted line indicates the equilibrium onset potential of the HER. (b) Tafel plots for the same samples and their corresponding linear fits. (c) Time evolution of the i₂ mass spectrometric signal recorded at two applied electrolytic currents for sample MS-2.2. (d) Time integrals of i₂ signals as a function of the theoretical amount of H₂ generated at the electrodes (obtained by using the Faraday law). (e) Chronoamperometry test recorded at a bias potential of −0.484 V vs RHE for sample MS-1.6. (f) Photocurrents of the MoS₂ samples as a function of the applied electrode potential.