Boron nitride on SiC(0001)

TL;DR

This study investigates boron nitride layers on SiC(0001), revealing they are low-quality B$_x$N$_y$ rather than high-quality hBN, and examines their transformation into graphene upon annealing, relevant for van der Waals heterostructures.

Contribution

It provides a comprehensive experimental and theoretical analysis showing boron nitride on SiC(0001) is low-quality B$_x$N$_y$, not high-quality hBN, and details its evolution into graphene.

Findings

Boron nitride forms as a low-quality B$_x$N$_y$ layer, not high-quality hBN.

The B$_x$N$_y$ layer is aligned with SiC and is replaced by graphene upon annealing.

The study combines experimental and theoretical techniques for analysis.

Abstract

In the field of van der Waals heterostructures, the twist angle between stacked two-dimensional (2D) layers has been identified to be of utmost importance for the properties of the heterostructures. In this context, we previously reported the growth of a single layer of unconventionally oriented epitaxial graphene that forms in a surfactant atmosphere [F. C. Bocquet, et al., Phys. Rev. Lett. 125, 106102 (2020)]. The resulting G-R0$^\circ$ layer is aligned with the SiC lattice, and hence represents an important milestone towards high quality twisted bilayer graphene (tBLG), a frequently investigated model system in this field. Here, we focus on the surface structures obtained in the same surfactant atmosphere, but at lower preparation temperatures at which a boron nitride template layer forms on SiC(0001). In a comprehensive study based on complementary experimental and theoretical techniques, we find -- in contrast to the literature -- that this template layer is a hexagonal B$_x$N$_y$ layer, but not high-quality hBN. It is aligned with the SiC lattice and gradually replaced by low-quality graphene in the 0$^\circ$ orientation of the B$_x$N$_y$ template layer upon annealing.