Understanding 2D Semiconductor Edges by Combining Local and Nonlocal Effects: The Case of MoSi2N4

Similar to surfaces of three-dimensional (3D) bulk materials, edges are inevitable in 2D materials and have been studied a lot (e.g., for MoS2). In the current work, taking the ambient-stable MoSi2N4 as an example, nonpolar and polar edges as well as polar-edge reconstructions are studied based on first-principles calculations. We demonstrate that a combination of the "local" electron counting model (ECM) at edges and "nonlocal" charge polarity analysis (CPA) across the ribbon is essential for a unified understanding of the "local" edge properties and edge reconstructions in the following aspects. For pristine edges, the semiconducting (metallic) property of nonpolar armchair (polar zigzag) edges is related to CPA, and the spin-paired (spin-polarized) electronic structure of nonpolar (polar) edges is related to the ECM. For polar-edge reconstructions: (1) the polar edges become semiconducting when the reversed dipole from edge-reconstruction partially cancels the accumulated electric dipole within the ribbon; (2) the polar edges can further be spin-paired when edge-reconstruction fulfills the ECM for both the double cation (Mo, Si)-edge and the anion N-edge; and (3) ECM and CPA give the same conclusion for edge-reconstruction. Our analysis of combining ECM and CPA not only gives the general guidance for obtaining spin-paired and semiconducting polar edges but also potentially helps deepen the understanding of edges of other 2D layered materials.