Engineering grain boundaries at the 2D limit for the hydrogen evolution reaction

Atom-thin transition metal dichalcogenides (TMDs) have emerged as fascinating materials and key structures for electrocatalysis. So far, their edges, dopant heteroatoms and defects have been intensively explored as active sites for the hydrogen evolution reaction (HER) to split water. However, grain boundaries (GBs), a key type of defects in TMDs, have been overlooked due to their low density and large structural variations. Here, we demonstrate the synthesis of wafer-size atom-thin TMD films with an ultra-high-density of GBs, up to ~1012 cm−2. We propose a climb and drive 0D/2D interaction to explain the underlying growth mechanism. The electrocatalytic activity of the nanograin film is comprehensively examined by micro-electrochemical measurements, showing an excellent hydrogen-evolution performance (onset potential: −25 mV and Tafel slope: 54 mV dec−1), thus indicating an intrinsically high activation of the TMD GBs.

Au nanoparticles. a-b show the size of Au QDs on MoS2 is usually larger than on SiO2/Si. Some QDs are clearly seen to be located at the MoS2 edge. This is because the MoS2 layer will drive the Au QD droplets along its growth direction; QDs then coalesce into larger Au particles if brought in contact with one another. Therefore, Au QDs will increase in size and decrease in density as the growth of MoS2 proceeds. It is worth mentioning that we did not observe any obvious change in Au nanoparticle size and density when using different annealing temperatures (Supplementary Figure 3), suggesting that the growth of MoS2 is the dominant factor in mobilizing the Au QDs. On the other hand, no obvious change in the size or the density of initially larger Au nanoparticles was observed after growth for particles with an average of 15 nm in diameter, as shown in c-d. This indicates no movement of these larger Au nanoparticle during MoS2 growth, which is interpreted as being due to the solid state of Au nanoparticle at the growth temperature in our experiment. We also investigated the Faradaic efficiency of MoS2 nanograin film for hydrogen production using a previously-reported method 10-12 . The Faradaic efficiency was calculated by comparing the measured amount of H2 generated by cathodal electrolysis in GC setup with the calculated amount of H2 in the electrochemical measurements, as shown in Experiment Method. We observed a good correlation between the calculated and experimental amounts of H2 gas under varied current densities in our work.
The faradaic yields of H2 in our MoS2 nanograin film are 98.4±2.5%, 100.0±2.2%, and 102.0±3.6% at 20, 30 and 40 mA cm -2 , respectively. The results indicate a high-efficient hydrogen production (nearly 100% energy conversion) in our work.  According to this model, the Raman line intensity, i.e., I(ω), at the frequency ω can be written as 14 : where q is the wave vector expressed in units of 2π/a (a is the lattice constant), Γ0 is the half width of the Raman peak, L is the correlation length, and ω(q) is the function of Raman phonon dispersion.
Note that only in a perfect lattice is L equal to dg because there are no defects in the lattice to act as phonon scattering centers. However, in most cases, L is much smaller than dg. L is affected by many factors 13 such as grain size, defects/impurities interspacing, the size of polytypic domains or clusters in semiconducting alloys, etc.
Prior work reported 14 in thin-film materials suggest that: when the grain size is > 100 nm, Γ will not decrease with grain size, because nearly all of the phonon dispersion arises are from defects inside the grain; in contrast, when the grain size is < 100 nm, Γ decreases with grain size because the phonon dispersion is mainly caused by GBs. As for 2D ultra-thin films, Γ will be greatly decreased due to the in-plane mode while hardly affected by out-of-plane mode. This indicates that in-plane Γ of the nanograin films should be much smaller than that in CVD-grown and mechanically-exfoliated MoS2.
According to equation (5), the intensity of the E2g mode in the nanograin film is calculated to be much smaller compared to the two other types of samples. At the same time, the intensity of the A1g mode is comparatively unchanged. This results in a significant lower I(E2g)/I(A1g) ratio for the nanograin film, which is in consistent with our experimental observation. Additionally, the E2g peak full width at half maximum (FWHM) of the nanograin MoS2 film in this work is three times larger than that of CVDgrown and mechanically-exfoliated MoS2. According to the relationship between FWHM and grain size, 14 this suggests that the grains in this work are much smaller than those in CVD-grown and mechanically-exfoliated MoS2.

Supplementary Note 2
Fabrication procedure for the single-layer MoS2 microelectrode on a graphene supporting layer.
First, a 16 mm×16 mm SiO2 (285 nm)/Si chip with a pre-patterned set of 32 Au contact pads was fabricated using conventional photolithography (Supplementary Figure 21a). Second, a large-scale single-layer graphene film was grown on Cu foils by CVD 15 , and then transferred onto the prepatterned chip through the conventional PMMA-assisted transfer method (Supplementary Figure 21b).
Third, EBL and O2 plasma were employed to pattern the graphene film into isolated small strips This indicates that the graphene layer can facilitate electron injection into the MoS2 due to their lowbarrier band alignment [16][17][18] , which is a strategy that is widely adopted in TMD semiconductor devices [19][20][21] . . Assuming a value of 420 μC cm -2 for a saturated Cuupd monolayer formation on active metal sites 8,9 , the ECSA can be calibrated as: ECSA = QCu/(420 μC cm -2 ), and the values are shown in Supplementary Table 3.

Supplementary Note 6
Investigation of the HER activity of Au. In order to investigate the HER-activity contribution of the Au nanoparticles to our experiments, we carry out a control experiment whereby a 1 nm-thick Au layer was deposited Au onto carbon cloth, then followed exposed to the same growth condition of MoS2 but without Mo and S sources, and finally measure its HER performance. Therefore, we can make a clear