Absorption $\textit{versus}$ Adsorption: High-Throughput Computation of Impurities in 2D Materials

TL;DR

This study systematically compares absorption and adsorption mechanisms for doping 2D materials using high-throughput DFT calculations, revealing preferences based on dopant size and valence electrons, and provides an open database for further research.

Contribution

It introduces a comprehensive high-throughput computational approach and database to distinguish doping mechanisms in 2D materials, utilizing a new Python tool for defect setup.

Findings

Interstitial positions favored for small, partially filled valence electron dopants in large lattice materials.

Adatoms preferred for dopants with fewer valence electrons due to lower coordination.

Generated data and structures are available in an open database for community use.

Abstract

Doping of a two-dimensional (2D) material by impurity atoms occurs \textit{via} two distinct mechanisms: absorption of the dopants by the 2D crystal or adsorption on its surface. To distinguish the relevant mechanism, we systematically dope 53 experimentally synthesized 2D monolayers by 65 different chemical elements in both absorption and adsorption sites. The resulting 17,598 doped monolayer structures were generated using the newly developed ASE \texttt{DefectBuilder} -- a Python tool to set up point defects in 2D and bulk materials -- and subsequently relaxed by an automated high-throughput density functional theory (DFT) workflow. We find that interstitial positions are preferred for small dopants with partially filled valence electrons in host materials with large lattice parameters. On the contrary, adatoms are favored for dopants with a low number of valence electrons due to lower coordination of adsorption sites compared to interstitials. The relaxed structures, characterization parameters, defect formation energies, and magnetic moments (spins) are available in an open database to help advance our understanding of defects in 2D materials.