The formation of lattice defects and their effect on the electronic properties of single-layer ${\mathrm{FeCl}}_2$ are investigated by means of first-principles calculations. Among the vacancy defects, namely mono-, di-, and three-Cl vacancies and mono-Fe vacancy, the formation of mono-Cl vacancy is the most preferable. Comparison of two different antisite defects reveals that the formation of the Fe-antisite defect is energetically preferable to the Cl-antisite defect. While a single Cl vacancy leads to a $1{\mu }_B$ decrease in the total magnetic moment of the host lattice, each Fe vacant site reduces the magnetic moment by $4{\mu }_B$. However, adsorption of an excess Cl atom on the surface changes the electronic structure to a ferromagnetic metal or to a ferromagnetic semiconductor depending on the adsorption site without changing the ferromagnetic state of the host lattice. Both Cl-antisite and Fe-antisite defected domains change the magnetic moment of the host lattice by $-1{\mu }_B$ and $+3{\mu }_B$, respectively. The electronic ground state of defected structures reveals that (i) single-layer ${\mathrm{FeCl}}_2$ exhibits half-metallicity under the formation of vacancy and Cl-antisite defects; (ii) ferromagnetic metallicity is obtained when a single Cl atom is adsorbed on upper-Cl and Fe sites, respectively; and (iii) ferromagnetic semiconducting behavior is found when a Cl atom is adsorbed on a lower-Cl site or a Fe-antisite defect is formed. Simulated scanning electron microscope images show that atomic-scale identification of defect types is possible from their electronic charge density. Further investigation of the periodically Fe-defected structures reveals that the formation of the single-layer ${\mathrm{FeCl}}_3$ phase, which is a dynamically stable antiferromagnetic semiconductor, is possible. Our comprehensive analysis on defects in single-layer ${\mathrm{FeCl}}_2$ will complement forthcoming experimental observations.

• Received 20 July 2020
• Revised 13 October 2020
• Accepted 22 December 2020

DOI:https://doi.org/10.1103/PhysRevB.103.014106