▎ 摘　　要

We perform ab initio calculations that indicate that the relative stability of antiphase boundaries (APBs) with armchair and zigzag chiralities in monolayer boron nitride (BN) is strongly dependent on the chemical potentials of boron (B) and nitrogen (N) in the synthesis process. For a realistic interval of chemical potentials, ranging from gas-phase to solid-state sources, a stoichiometric armchair APB with tetragons and octagons in its core and no homopolar bonds (N Nor B B) is found to be the most stable in the intrinsic region of the chemical-potential interval, while for N-rich (B-rich) environments a zigzag-oriented boundary with a core consisting of hexagons only, containing N N (B B) bonds is found to be the most stable. Zigzag boundaries with core structures consisting of pentagons and octagons, containing homopolar bonds, are also considered in both undoped and carbon-doped forms. We find that at the typical temperatures of BN-nanostructure synthesis several of these zigzag boundaries should coexist with the armchair one, under N-rich or B-rich conditions. This stability transition arises from a competition between homopolar-bond and elastic-energy costs in the core of the APBs, in contrast with the case of graphene, where the zigzag geometry with pentagons and octagons is the most stable. We also address the stability of N- and B-terminated edges of BN triangular islands and find a robust stability for N-terminated edges, across a wide range of chemical potential values, in agreement with the experimentally observed tendency of BN islands to display N-terminated zigzag edges. We predict that fine-tuning the chemical potentials with a B-rich source where B atoms are in a bonding environment of low stability may lead to the formation of B-terminated zigzag edges. The electronic structure of the two types of APBs in BN is shown to be particularly distinct, with the zigzag APB depicting defect-like deep electronic bands in the band gap, while the armchair APB shows bulk-like shallow electronic bands.