The Growth of Gold Within a Two-Dimensional Confined Space Created by Twisted Bilayer Molybdenum Disulfide

Two-dimensional (2D) materials hold potential for creating novel heterostructures and devices for sustainable devices, with metal contacts or intercalations playing a crucial role in their functionality. A comprehensive understanding of the metal-2D material interface, including its atomic structure, interfacial chemistry, and properties, is critical for optimizing performance and guiding the design of advanced devices. We choose gold-molybdenum disulfide (Au- MoS₂) as a model system owing to the well-established epitaxial growth of FCC gold on the basal plane of 2H-MoS₂ (P6₃/mmc). Our study aims to grow 2D gold layers within a MoS₂ - MoS₂ bilayer and explore how the bilayer’s confinement influences the atomic structure, growth mechanisms and physical properties of gold within this 2D confined space.

Scanning transmission electron microscopy (STEM) is a powerful tool for sub-nanometer and atomic-scale analysis of materials. Cross-sectional STEM, in particular, is suitable for examining the atomic structure and tracking the structural evolution of the Au- MoS₂ interfaces. Using cross-sectional STEM, we observed that confined Au evolves a nanodisc morphology with diameters ranging from 50 to 100 nm and a height of approximate 5 nm. This morphology is "thinner" than the non-confined epitaxial Au on MoS₂, which typically forms a triangular structrues with an edge length of ~25 nm and a height of ~8 nm. Moreover, we observed the presence of a horizontal twin planes within the confined Au nanodiscs, which to our knowledge, is not observed in the non-confined epitaxial Au-MoS₂. The formation of the horizontal twin boundary in the nanodiscs might arise as a mechanism to accommodate strain induced by the confinement.

In order to begin exploration of any unexpected electronic effects of the confined Au, we carried out electron energy loss spectroscopy (EELS) at varying locations on the gold nanodiscs. STEM-EELS analysis at the nanodisc perimeter shows a plasmon peak close to 1.0 eV, whereas that at the disc center is about 1.5 eV. The former is known to be the dipolar plasmon peak which is optically active, while the higher energy bulk "breathing mode" is optically inactive. Both these values have noticeably lower plasmon energies than those of non-confined epitaxial Au, which might be because they are sandwiched between higher refractive index MoS₂ substrates (refractive index = 4.77).

These new findings therefore open up several possibilities for studying the growth of crystal in a 2D confined space. For instance, can we get continuous thin films instead of discrete nanodiscs?
Whether the twin formation is driven by energetics, or a kinetic compromise during the growth process? Any unusual properties associated with the confinement? This paper will discuss progress along these lines and prospects for future avenues of research.