Hexagonal boron nitride (hBN) is a two-dimensional material with a structure analogous to graphene but with a wide bandgap, excellent thermal stability, and chemical inertness. Its applications span from dielectric layers in 2D electronics to protective coatings and quantum emitters. The synthesis of large-area hBN films is critical for industrial scalability, and several techniques have been developed to achieve high-quality growth, including chemical vapor deposition (CVD), physical vapor deposition (PVD), and epitaxial growth. Each method has distinct advantages and challenges in terms of substrate compatibility, nucleation control, and film uniformity.
Chemical vapor deposition is the most widely used technique for synthesizing large-area hBN films due to its scalability and ability to produce continuous layers. The process typically involves the decomposition of boron and nitrogen precursors, such as ammonia borane (NH3-BH3) or borazine (B3N3H6), at high temperatures (900–1200°C) on catalytic metal substrates like copper or nickel. Copper is preferred for its low carbon solubility, which promotes surface-mediated growth, resulting in monolayer or few-layer hBN. Nickel, with higher boron solubility, often leads to thicker films due to segregation during cooling. The choice of precursor also affects film quality; borazine yields higher crystallinity but requires careful handling due to its instability. Nucleation density is controlled by optimizing temperature, gas flow rates, and substrate pretreatment. Excessive nucleation sites lead to polycrystalline films with grain boundaries, while sparse nucleation can result in incomplete coverage. Post-growth annealing improves crystallinity by reducing defects and merging smaller domains.
Physical vapor deposition, including sputtering and pulsed laser deposition (PLD), offers an alternative for hBN synthesis, particularly for applications requiring lower growth temperatures or compatibility with non-metallic substrates. Sputtering employs a boron target in a nitrogen-rich plasma, with substrate temperatures ranging from 300–800°C. The stoichiometry and crystallinity of hBN films depend on the nitrogen-to-argon ratio and substrate bias. PLD uses a high-power laser to ablate a boron nitride target in a nitrogen environment, producing highly adherent films with controlled thickness. However, PVD-grown hBN often exhibits poorer crystallinity compared to CVD, requiring post-deposition annealing to enhance ordering. The main advantage of PVD is its versatility in depositing hBN on diverse substrates, including oxides and semiconductors, without the need for catalytic metals.
Epitaxial growth on single-crystal substrates like rhodium (Rh) or ruthenium (Ru) enables the synthesis of wafer-scale single-crystal hBN films. The lattice matching between hBN and the substrate minimizes strain, promoting aligned domain formation. For instance, hBN grown on Rh(111) at 1100°C under ultrahigh vacuum forms well-ordered monolayers with near-perfect alignment. However, the high cost of single-crystal metal substrates limits industrial adoption. Quasi-epitaxial growth on polycrystalline metals like platinum (Pt) has been explored as a compromise, where textured substrates promote larger domain sizes compared to conventional CVD on copper or nickel.
Substrate selection plays a crucial role in determining film quality. Metallic substrates facilitate catalytic decomposition of precursors and enable layer-by-layer growth, but they require transfer for electronic applications, which can introduce defects. Insulating substrates like sapphire or silicon carbide allow direct growth of hBN without transfer, but achieving high crystallinity is challenging due to weaker surface interactions. Recent advances have demonstrated the use of sacrificial layers or buffer coatings to combine the benefits of catalytic growth and transfer-free processes.
Nucleation control is essential for uniform hBN films. Pre-patterning substrates with catalytic nanoparticles or defects can guide nucleation sites, enabling ordered arrays of single-crystal domains. Conversely, suppressing random nucleation is achieved through substrate polishing, hydrogen etching, or precursor dilution. For example, introducing argon as a carrier gas reduces precursor flux, lowering nucleation density and promoting larger domain sizes. Temperature gradients across the substrate can also be exploited to create spatially controlled nucleation zones, improving film homogeneity.
Uniformity challenges arise from gas flow dynamics, temperature variations, and substrate surface conditions. In CVD, laminar gas flow ensures consistent precursor delivery, while rotational substrate holders mitigate temperature inhomogeneities. Edge effects, where films are thicker or thinner near substrate edges, are minimized by optimizing reactor geometry and gas injection designs. PVD systems face similar issues, with uniformity depending on target-to-substrate distance and plasma distribution.
Single-crystal hBN films exhibit superior electronic and thermal properties due to the absence of grain boundaries, making them ideal for high-performance devices. However, their synthesis requires precise control over nucleation and growth kinetics, often at the expense of scalability. Polycrystalline films, while easier to produce over large areas, suffer from reduced thermal conductivity and increased defect density. The trade-off between crystallinity and scalability depends on the intended application; for instance, polycrystalline hBN may suffice as a dielectric spacer in heterostructures, whereas single-crystal films are necessary for quantum light sources.
Comparative analysis of growth techniques reveals distinct performance metrics:
+--------------------------------+---------------------+---------------------+---------------------+
| Growth Method | Crystallinity | Scalability | Substrate Options |
+--------------------------------+---------------------+---------------------+---------------------+
| CVD | High (single-crystal| High | Metals (Cu, Ni) |
| | or polycrystalline) | | |
+--------------------------------+---------------------+---------------------+---------------------+
| PVD | Moderate | Moderate | Metals, oxides |
| | (polycrystalline) | | |
+--------------------------------+---------------------+---------------------+---------------------+
| Epitaxial Growth | Very High | Low | Single-crystal metals|
| | (single-crystal) | | |
+--------------------------------+---------------------+---------------------+---------------------+
Future advancements in hBN synthesis may focus on hybrid approaches, combining the scalability of CVD with the precision of epitaxial techniques. For example, pre-seeding substrates with aligned hBN nuclei could enable large-area single-crystal growth on cost-effective polycrystalline metals. Additionally, the development of novel precursors with lower decomposition temperatures may reduce energy consumption while maintaining film quality.
In summary, the synthesis of large-area hBN films involves careful optimization of growth techniques, substrates, and nucleation conditions to balance crystallinity and scalability. CVD remains the most practical method for industrial-scale production, while epitaxial growth offers the highest quality for specialized applications. Overcoming uniformity challenges and improving single-crystal domain sizes will be key to unlocking the full potential of hBN in next-generation technologies.