Identifying substitutional oxygen as a prolific point defect in monolayer transition metal dichalcogenides with experiment and theory

TL;DR

This study combines experimental microscopy and advanced theoretical calculations to identify substitutional oxygen as the prevalent point defect in 2D transition metal dichalcogenides, challenging the common assumption of chalcogen vacancies.

Contribution

It provides the first direct experimental and theoretical evidence that substitutional oxygen, not vacancies, dominates in 2D TMDs under standard conditions.

Findings

No in-gap states observed for chalcogen defects

Substitutional oxygen identified as the main defect

Chalcogen vacancies are less common than previously thought

Abstract

Chalcogen vacancies are considered to be the most abundant point defects in two-dimensional (2D) transition-metal dichalcogenide (TMD) semiconductors, and predicted to result in deep in-gap states (IGS). As a result, important features in the optical response of 2D-TMDs have typically been attributed to chalcogen vacancies, with indirect support from Transmission Electron Microscopy (TEM) and Scanning Tunneling Microscopy (STM) images. However, TEM imaging measurements do not provide direct access to the electronic structure of individual defects; and while Scanning Tunneling Spectroscopy (STS) is a direct probe of local electronic structure, the interpretation of the chemical nature of atomically-resolved STM images of point defects in 2D-TMDs can be ambiguous. As a result, the assignment of point defects as vacancies or substitutional atoms of different kinds in 2D-TMDs, and their influence on their electronic properties, has been inconsistent and lacks consensus. Here, we combine low-temperature non-contact atomic force microscopy (nc-AFM), STS, and state-of-the-art ab initio density functional theory (DFT) and GW calculations to determine both the structure and electronic properties of the most abundant individual chalcogen-site defects common to 2D-TMDs. Surprisingly, we observe no IGS for any of the chalcogen defects probed. Our results and analysis strongly suggest that the common chalcogen defects in our 2D-TMDs, prepared and measured in standard environments, are substitutional oxygen rather than vacancies.