Hydrogen adsorption on doped MoS2 nanostructures

Electrochemical devices for efficient production of hydrogen as energy carrier rely still largely on rare platinum group metal catalysts. Chemically and structurally modified metal dichalcogenide MoS2 is a promising substitute for these critical raw materials at the cathode side where the hydrogen evolution reaction takes place. For precise understanding of structure and hydrogen adsorption characteristics in chemically modified MoS2 nanostructures, we perform comprehensive density functional theory calculations on transition metal (Fe, Co, Ni, Cu) doping at the experimentally relevant MoS2 surfaces at substitutional Mo-sites. Clear benefits of doping the basal plane are found, whereas at the Mo- and S-edges complex modifications at the whole edge are observed. New insight into doping-enhanced activity is obtained and guidance is given for further experiments. We study a machine learning model to facilitate the screening of suitable structures and find a promising level of prediction accuracy with minimal structural input.

. Co-doped basal plane (x-y plane shown, view from the top, along the z axis). The closest and farthest nearest-neighbor distances are indicated for Co as dopant and for Mo as host atom. For Co the shortest nearest-neighbor distance is d 1 = 2.18Å and the longest d 6 = 3.14Å (corresponding to the broken Co...S bond). The nearest-neighbor distances from Mo to its nearest neighbor sulfurs are d 1 = · · · = d 6 = 2.40Å. As a detail, notice that for cobalt the broken bond Co...S is toward the sulfur at the bottom side of the surface layer.

Excess charge in the simulation cell
The charge state effects on adsorption energies are reported in Table S2 for the basal plane.

Effect of doping on local geometry
The changes in the local geometry on doping of the edges is summarized in the main text (Table 2 of the main text). The effects on the structure are discussed here in more detail. To ease the discussion, we divide the six studied edges into three sets from the point of view of nominal coordination number, 4, 5 or 6, of the dopant (coordination number of the substitutional site at the edge). However, this nominal coordination number does not play a relevant role in determining the ∆G H values (see main text).
In the first set, comprising the Mo-50, Mo-100, S-100 edges (see Figure 1 of the main text), Mo is coordinated to six sulfurs similarly to the basal plane. Note, in addition, that in the case of Mo-100 the outermost two sulfurs form dimers. When Fe, Co and Ni are substitutionally doped, we find no essential deformation to the 6-fold coordinated positions. This is in contrast to the basal plane, where Co and Ni doping led to a 5-fold coordinated structure. Quite surprisingly, for Co, Ni, and Fe even the dimerization of the topmost sulfurs in the Mo-100 structure is not affected. In contrast, Cu on Mo-50, Mo-100, S-75 and S-100 edges leads to some deformation of the structure, what is likely connected to its large number of d electrons.
In the second set, comprising only the S-75 edge, Mo is coordinated to five sulfurs in the pristine structure and there is a practically a mirror plane which coincides with the MoS 2 sheet. Doping with Co and Ni preserves the coordination and the mirror symmetry, while doping with Fe preserves the coordination but leads to slightly reduced symmetry. Doping with Cu leads to some structural deformation. Finally in the third set, comprising Mo-0 and S-50, Mo is coordinated to four sulfurs. For the Mo-0 edge, Fe, Co Ni and Cu preserve practically fully the structure and coordination compared with the pristine structure. For the S-50 edge, Fe, Co and Cu doping preserve the coordination and the structure, whereas Ni leads to a slight symmetry change. Figure S2 gives an example of the computational system for single hydrogen adsorption on doped Mo-50 edge (for illustration the supercell is extended to 2 x 2 size). In this example hydrogen adsorbs on the second-nearest neighbor sulfur site with respect to the position of the dopant atom. Figure S2. Example of the computational system in the case of H adsorption on Ni doped Mo-50 edge at the second-nearest neighbor sulfur site. The actual supercell is 229 atoms, which is here repeated to 2 x 2 size in the xy plane for illustration. Table S3 shows the distribution of the cases in the dataset as a function of dopant and system type.