Single-layer transition-metal dichalcogenides (TMDs) receive significant attention due to their intriguing physical properties for both fundamental research and potential applications in electronics, optoelectronics, spintronics, catalysis, and so on. Here, we demonstrate the epitaxial growth of high-quality single-crystal, monolayer platinum diselenide (PtSe2), a new member of the layered TMDs family, by a single step of direct selenization of a Pt(111) substrate. A combination of atomic-resolution experimental characterizations and first-principle theoretic calculations reveals the atomic structure of the monolayer PtSe2/Pt(111). Angle-resolved photoemission spectroscopy measurements confirm for the first time the semiconducting electronic structure of monolayer PtSe2 (in contrast to its semimetallic bulk counterpart). The photocatalytic activity of monolayer PtSe2 film is evaluated by a methylene-blue photodegradation experiment, demonstrating its practical application as a promising photocatalyst. Moreover, circular polarization calculations predict that monolayer PtSe2 has also potential applications in valleytronics.

Germanene, a 2D honeycomb lattice analogous to graphene, is fabricated on a Pt(111) surface. It exhibits a buckled configuration with a (3 × 3) superlattice coinciding with the substrate's (√19 × √19) superstructure. Covalent bonds exist throughout the germanene layer. The resulting high-quality germanene enables researchers to explore the fundamentals of germanene and its potential applications.

SignificanceMotivated by the recent achievements of artificial intelligence (AI) in linguistics, we design AI to learn properties of atoms from materials data on its own. Our work realizes knowledge representation of atoms via computers and could serve as a foundational step toward materials discovery and design fully based on machine learning.

Silicon-based two-dimensional (2D) materials are uniquely suited for integration in Si-based electronics. Silicene, an analogue of graphene, was recently fabricated on several substrates and was used to make a field-effect transistor. Here, we report that when Ru(0001) is used as a substrate, a range of distinct monolayer silicon structures forms, evolving toward silicene with increasing Si coverage. Low Si coverage produces a herringbone structure, a hitherto undiscovered 2D phase of silicon. With increasing Si coverage, herringbone elbows evolve into silicene-like honeycomb stripes under tension, resulting in a herringbone-honeycomb 2D superlattice. At even higher coverage, the honeycomb stripes widen and merge coherently to form silicene in registry with the substrate. Scanning tunneling microscopy (STM) was used to image the structures. The structural stability and electronic properties of the Si 2D structures, the interaction between the Si 2D structures and the Ru substrate, and the evolution of the distinct monolayer Si structures were elucidated by density functional theory (DFT) calculations. This work paves the way for further investigations of monolayer Si structures, the corresponding growth mechanisms, and possible functionalization by impurities.