Interplay of Kinetic Limitations and Disintegration: Selective Growth of Hexagonal Boron Nitride and Borophene Monolayers on Metal Substrates

Abstract

The CVD growth of bielemental 2D-materials by using molecular precursors involves complex formation kinetics taking place at the surface and sometimes also subsurface regions of the substrate. Competing microscopic processes fundamentally limit the parameter space for optimal growth of the desired material. Kinetic limitations for diffusion and nucleation cause a high density of small domains and grain boundaries. These are usually overcome by increasing the growth temperature and decreasing the growth rate. In contrast, the nature of molecular precursors with limited thermal stability can result in dissociation and preferential desorption, leading to an undesired or ill-defined composition of the 2D-material.Here we demonstrate these constraints in a combined lowenergy electron diffraction and low-energy electron microscopy study by examining the selective formation of single-layer hexagonal boron nitride (hBN) and borophene on Ir(111) using a borazine precursor. We derive a temperature−pressure phase diagram and apply classical nucleation theory to describe our results. By considering the competing processes, we find an optimum growth temperature for hBN of 950 °C. At lower temperatures, the hBN island density is increased, while at higher temperatures the precursor disintegrates and borophene is formed. Our results introduce an additional aspect that must be considered in any high-temperature growth of bielemental 2D-materials from single molecular precursors.

Unidirectional Nano-modulated Binding and Electron Scattering in Epitaxial Borophene

“…In the follow-up control experiment, the sample was flashed to 1100 °C and immediately cooled down, in order to dissolve boron in Ir, desorb any contaminants from the sample surface, and produce a fresh Bo layer via B segregation to the Ir surface (see Supporting Information Figure S1). Such procedure is a straightforward route for obtaining pristine samples, relying on the temperature-dependent solubility of B in Ir that was already exploited in previous studies. , Indeed, the XPS spectrum of the annealed sample shown in Figure b (top image) looks essentially the same as the spectrum of the initial pristine Bo/Ir sample, exhibiting only B 1 and B 2 components with restored positions and intensities. In addition, LEED data exhibiting a sharp (6 × 2) pattern in three domains (not shown) confirmed the presence of a clean Bo layer.…”