The structure of interfacial water is governed by a delicate interplay between water−substrate and water−water interactions. In order to identify the structure-determining factors of ordered two-dimensional water, first calculations of free-standing water layers have been performed. We demonstrate that square bilayers, rhombic bilayers, truncated-square bilayers, and secondary-prism bilayers are energetically more favorable than the traditionally considered hexagonal bilayer. These two-dimensional water structures are stabilized by a combination of high coordination and optimum tetrahedral bonding geometry. The identified structuredetermining factors responsible for the polymorphism of water in two dimensions will be operative in any confined water structure. Graphene influence on the stability of water sheets is for the first time explicitly treated using first-principles electronic structure calculations, and changes in some ordered water structures due to non-negligible graphene−water interaction are described.
■ INTRODUCTIONInterfaces with water play important roles in many fields of natural sciences such as biology, (electro-)chemistry, materials science, and earth science. 1,2 While these solid−liquid interfaces are all around us, we still have not yet understood how interfacial water really behaves. What is clear is that the structure of water layers directly at the solid−liquid interface is governed by an interplay between the water−water and water− substrate interactions. 3,4 A basic understanding of the factors influencing the structure of the water layers is crucial for comprehending the function of water at interfaces, as interfacial water often exhibits properties that are distinctively different from those of bulk water. 5−7 Finding the most stable lowdimensional water polymorphs is an important step toward this fundamental understanding, since these systems provide benchmarks in which water−water interaction solely determines the structure and properties of interfacial water. It has long been assumed that water at flat surfaces tends to form hexagonal ice-like layers, 8 and only through the corrugation of the underlying substrate some other geometry might be imposed on the water layer. 4,9,10 Hence, the recent high-resolution electron microscopy observation of square ice confined between two graphene sheets 11 was rather surprising, as the water−graphene interaction possesses only a weak dependence on the position and orientation of the water molecule. 12,13 There is an ongoing discussion about whether or not the transmission electron microscopy images 11 should be attributed to a film of square water or to accumulated salt sandwiched between graphene sheets. 14,15 Still, this experimental study has raised interest in the fundamental question of whether under ambient conditions two-dimensional water layers can realize geometries that are not feasible for bulk ice. 16 All experimentally observed two-dimensional water structures are subject to water−substrate interactions that influence their shape. ...