Role of H Transfer in the Gas-Phase Sulfidation Process of MoO3: A Quantum Molecular Dynamics Study.
Academic Article
Overview
Identity
Additional Document Info
Other
View All
Overview
abstract
Layered transition metal dichalcogenide (TMDC) materials have received great attention because of their remarkable electronic, optical, and chemical properties. Among typical TMDC family members, monolayer MoS2 has been considered a next-generation semiconducting material, primarily due to a higher carrier mobility and larger band gap. The key enabler to bring such a promising MoS2 layer into mass production is chemical vapor deposition (CVD). During CVD synthesis, gas-phase sulfidation of MoO3 is a key elementary reaction, forming MoS2 layers on a target substrate. Recent studies have proposed the use of gas-phase H2S precursors instead of condensed-phase sulfur for the synthesis of higher-quality MoS2 crystals. However, reaction mechanisms, including atomic-level reaction pathways, are unknown for MoO3 sulfidation by H2S. Here, we report first-principles quantum molecular dynamics (QMD) simulations to investigate gas-phase sulfidation of MoO3 flake using a H2S precursor. Our QMD results reveal that gas-phase H2S molecules efficiently reduce and sulfidize MoO3 through the following reaction steps: Initially, H transfer occurs from the H2S molecule to low molecular weight Mo xO y clusters, sublimated from the MoO3 flake, leading to the formation of molybdenum oxyhydride clusters as reaction intermediates. Next, two neighboring hydroxyl groups on the oxyhydride cluster preferentially react with each other, forming water molecules. The oxygen vacancy formed on the Mo-O-H cluster as a result of this dehydration reaction becomes the reaction site for subsequent sulfidation by H2S that results in the formation of stable Mo-S bonds. The identification of this reaction pathway and Mo-O and Mo-O-H reaction intermediates from unbiased QMD simulations may be utilized to construct reactive force fields (ReaxFF) for multimillion-atom reactive MD simulations.
