Since the advent of two dimensional (2D) materials, their applications for the renewable synthesis of solar fuels have attracted tremendous research efforts[52–55], owing to the sought-after advantages of these single-layer compounds including efficient charge transfer, abundant reaction sites, and highly-adjustable electronic structures. Compared to conventional three-dimensional (3D) bulk photocatalysts, 2D photocatalysts possess a series of unique properties which may potentially enhance photoelectrochemical device efficiency for solar fuel generation. These features include extremely high specific surface area for water redox reaction, ultra-small thickness facilitating the diffusion of electron and hole to the solid/water interface, and multiple choices of host-guest combinations based on van de Waals heterojunctions.
Prior experimental research has yielded more than 20 2D compounds for solar fuel generation with band-gap energies in the desirable range that strongly overlaps with the solar spectrum. Table 2 demonstrates the photocatalytic performance and material parameters of recently discovered 2D layered materials including metal oxides, metal chalcogenides, and metal-free nanosheets. These 2D photo catalysts showed distinct performance as compared to their 3D counterparts. For instance, Xie et al. found that SnS2 single-layers yield 70 times higher of photocurrent density than that of bulk SnS2 possibly due to multiple reasons including improved carrier density, a fully depleted pace charge layer, and fast interfacial charge transfer. Monolayer 1T-MoS2 exhibited a 26 000 μmol/(h·g) of H2 yield under the irradiation of 100 W halogen light, while bulk MoS2 is almost inert in catalyzing water due to the lack of active sites. Owing to the promise to address the long-lived solar energy conversion problem, the continued discovery of novel 2D photocatalysts is of great interest.
In the field of 2D photocatalysis, computational simulation has been a powerful tool to predict promising candidates and eliminate unlikely materials. For instance, Hennig group proposed a strategy to screening 2D materials for photo water-splitting and carried out a series of work in the field. Screening criteria applied include, but are not limited to, suitable band gap, band edge, low formation energy, and stability in water. As shown in Fig. 2, a couple of 2D materials were identified and predicted theoretically to be suitable for photo water-splitting, including a family of group IV monochalcogenides, MX (M = Ge, Sn, Pb; X = O, S, Se, Te), MoS2, WS2, PtS2, and PtSe2. Liu et al. predicted that single layer metal-phosphorus-trichalcogenides, MPX3 (A = MII, MI0.5MIII0.5; X = S, Se; MI, MII, and MIII represent Group-I, Group-II, and Group-III metals, respectively) exhibited low formation energy, suitable band gap, band edges, and outstanding photo absorption efficiency for photo water splitting. In recent years, several other 2D materials have been predicted for solar fuel generation[61–67].
Prior computational screening in the field of 2D photocatalysis have been limited to a relatively small compound space. A thorough search of promising candidates and, more importantly, a deeper understanding of why these 2D compounds host optimal material properties for photocatalysis, is critically missing. In recent years, extensive efforts have initiated the construction of several 2D material databases mostly based on the generation of single-layer structures through data-mining layered compounds in existing inorganic compound databases including the MP database and the Inorganic Crystal Structure Database (ICSD)[69–71]. Although only a limited number of materials’ properties are computed and included in these databases for now, these 2D compound repositories have become a fertile ground for research efforts in the field aiming to theoretically predict novel 2D compounds for photocatalysis.
In a recent work by our group (unpublished), 62 promising 2D compounds have been predicted or “re-identified” as photoanode or photocathode materials for solar fuel generation. The discovery process included a data-mining procedure for layered structure identification, a new electronic structure framework for 2D compounds including automatic first Brillouin zone identification and high-symmetry k-points definition, combined with a multiple-tier discovery pipeline incorporating multiple material screening criteria including type and size of band gaps, band edge energies, and exciton binding energies. The study also establishes the tunability of band edges of binary 2D compounds and the interplay between electronic structure, anion/cation electronegativity, and orbital hybridization. This work demonstrated the power of data-driven approach for accelerated discovery of 2D functional materials. These findings, together with other prior data-driven discovery work in the field, have provided a large set of potential candidates for future experimental investigation to make a real breakthrough.