Researchers grow single-crystal GaN films on amorphous glass using a chemically converted molybdenum nitride buffer, removing the need for crystalline substrates in epitaxy
(Nanowerk Spotlight) The integration of high-performance semiconductors into materials such as glass, plastic, and other non-crystalline surfaces has remained constrained by one technical bottleneck: the inability to grow defect-free, single-crystal films without a lattice-matched substrate. These surfaces lack an orderly atomic structure, which is essential for guiding crystal alignment during film growth.
For compound semiconductors like gallium nitride (GaN), which are widely used in power electronics, optoelectronics, and high-frequency devices, epitaxial growth depends on stacking one crystal layer precisely on top of another. This process requires a substrate with a well-ordered atomic lattice. If the underlying material is amorphous or structurally incompatible, the resulting film forms with misaligned regions and high defect density, which limits its electronic performance.
This substrate dependence has limited where and how GaN and similar materials can be used. Traditional substrates such as sapphire and silicon offer the necessary crystal order but are rigid, expensive, and poorly suited for emerging device formats that demand flexibility, transparency, or compatibility with unconventional surfaces.
Various workarounds have been explored, particularly the use of two-dimensional (2D) materials like graphene or molybdenum disulfide as buffer layers. These ultrathin materials can be placed on arbitrary substrates to provide atomically flat surfaces. However, because they interact only weakly with the growing film, they offer little control over in-plane orientation. As a result, films grown on these buffers typically consist of many misaligned crystalline grains and fail to achieve the uniformity required for advanced device applications.
No existing method has demonstrated wafer-scale growth of single-crystal GaN directly on amorphous substrates with the crystalline quality necessary for practical use. This limitation has remained a major barrier to expanding the use of GaN-based technologies into new device formats that require mechanical flexibility, optical transparency, or low-cost fabrication on unconventional surfaces.
A new study published in Science Advances (“Two-dimensional buffer breaks substrate limit in III-nitrides epitaxy”) introduces a method for growing single-crystalline GaN directly on amorphous silicon dioxide (SiO₂), a substrate previously considered incompatible with high-quality epitaxy. Researchers from institutions in China and Saudi Arabia demonstrate that by chemically transforming a two-dimensional MoS₂ buffer into molybdenum nitride (MoN), they create a covalently bonded interface that supports ordered crystal growth across the wafer.
“This is a paradigm shift in epitaxy,” Husam Alshareef, Professor of Materials Science and Engineering at King Abdullah University of Science and Technology (KAUST), tells Nanowerk. “By transforming a van der Waals bonded 2D material to a covalently bonded 2D material, we’ve engineered a surface that allows for the growth of single-crystalline films on amorphous substrates, which was previously thought impossible.”
The conversion is achieved by exposing the multilayer MoS₂ to high-temperature nitridation, which replaces sulfur atoms with nitrogen and forms a thin MoN film. This process, which the researchers term a “chemical bond transition,” produces a buffer layer with high surface energy and defined crystallographic orientation. The MoN inherits the hexagonal symmetry of the MoS₂ and preserves lattice coherence across its surface.
Comparison of MoN and other buffers for the epitaxial growth of AlN. (A and B) Schematic diagram of the nucleation and epitaxy of III-nitrides on an amorphous SiO2 with MoS2 (A) and MoN (B) buffers. The blue dashed arrows with arbitrary orientations and the solid arrows with consistent orientation represent the random and unidirectional in-plane orientation of the nuclei, respectively. (C) Binding energies of N-polar AlN on MoN, MoS2, graphene (Gra), and h-BN under a fixed number of AlN units. (D and E) Distributions of charge density difference in AlN/MoS2 and AlN/MoN heterostructures (red, positive; green, negative). The isosurface level is 0.006 e bohr−3. (F) Total energy difference versus in-plane orientation angles from 0° to 120° with 15° increments for the AlN cluster on MoS2 and MoN buffers. The AlN cluster at an in-plane orientation of 0° is used as a reference. (G and H) Out-of-plane (top) and in-plane (bottom) EBSD IPF maps of GaN grown on MoS2 (G) and MoN (H) buffers. (Image: Reprinted from DOI:10.1126/sciadv.adw5005, CC BY) (click on image to enlarge)
To enable GaN growth, the researchers first deposit a 30-nanometer-thick aluminum nitride (AlN) layer onto the MoN buffer. Simulations using density functional theory predict that AlN bonds more strongly to MoN than to typical van der Waals materials and favors a single, well-defined in-plane orientation. This prediction is confirmed by electron backscatter diffraction (EBSD) mapping, which shows a consistent orientation across the GaN film when grown on MoN. By contrast, growth on unconverted MoS₂ results in misaligned grains and polycrystalline structures.
“It was once believed that perfect single-crystal substrates were essential for the quality of semiconductor films,” notes Xiangming Xu, a KAUST researcher and incoming principal investigator at the Chinese Academy of Sciences in Shanghai. “After the discovery of lattice orientation of 2D MoN films perfectly inherited from its father—wafer-scale single-crystal MoS₂—we fundamentally liberated epitaxial growth from substrate constraints, enabling scalable, high-quality wide-bandgap films on even amorphous substrates, ready for industrial deployment.”
The GaN films were grown using a three-step epitaxy process. An initial high-temperature AlN layer was followed by a 3D island formation stage in which discrete GaN crystallites nucleated and began to coalesce. A final two-dimensional growth step smoothed the film, resulting in continuous coverage with atomically flat terraces. X-ray rocking curve measurements indicate narrow full width at half maximum (FWHM) values—0.19° for the (002) reflection and 0.32° for the (102)—comparable to GaN grown on silicon or sapphire. Across 15 fabrication runs, these values remained consistent, underscoring the reproducibility of the method.
Transmission electron microscopy reveals a threading dislocation density of 2.2 × 10⁸ cm⁻² in the top GaN layer. Most dislocations are filtered out during the early stages of growth or terminated at the AlN/GaN interface. The relatively low lattice mismatch between AlN and MoN, combined with the uniform orientation induced by the buffer, helps suppress defect formation.
To validate the electronic performance of the material, the team fabricated high-electron-mobility transistors (HEMTs) on the GaN/AlN/MoN/SiO₂ structure. The transistors feature a standard AlGaN/AlN/GaN heterostructure that supports a two-dimensional electron gas (2DEG) at the interface. At room temperature, the devices achieved a 2DEG mobility of 2240 cm²/V·s and a sheet carrier density of 7.7 × 10¹² cm⁻². Cooling to 5 K increased mobility to over 13,000 cm²/V·s. These values are comparable to the best reported for GaN devices grown on conventional substrates.
The device-level metrics confirm the material’s applicability. A typical HEMT fabricated on the amorphous substrate showed a threshold voltage of −2.8 V, a subthreshold swing of 83 mV/decade, and an on/off ratio exceeding 10⁹. Across 200 devices tested on the same wafer, the performance metrics remained tightly clustered. The results show not only feasibility, but uniformity at the scale required for commercial production.
Optical microscope image and the cross-sectional SEM image of the fabricated HEMT devices on the amorphous SiO2. (Image: Reprinted from DOI:10.1126/sciadv.adw5005, CC BY)
“This new epitaxy strategy has the potential to revolutionize the way we fabricate electronic and optoelectronic devices,” Zhe Zhuang, Associate Professor at Nanjing University, points out. “It opens up the possibility of integrating high-performance III-Nitride semiconductors on a wide range of non-crystalline substrates, which could lead to more versatile and cost-effective device fabrication.”
The technique could enable integration of wide-bandgap semiconductors on glass, polymer films, or other unconventional materials that cannot support conventional epitaxy. This expands the range of applications to include flexible electronics, transparent displays, and low-cost optoelectronic modules. It may also facilitate integration with silicon photonics platforms and other chip-scale technologies where lattice mismatch has traditionally blocked direct growth.
By removing the constraint of crystallinity in the underlying substrate, the MoN buffer strategy offers a path to functional, scalable, and cost-efficient semiconductor integration. While further engineering is needed to adapt the method for complex multilayer stacks and other material systems, the core concept—chemically transforming a 2D buffer to enforce epitaxial order—provides a general framework that may apply beyond GaN.
The work was conducted by a collaborative team from Nanjing University, KAUST, and Hefei University of Technology.
Get our Nanotechnology Spotlight updates to your inbox!
Thank you!
You have successfully joined our subscriber list.
Become a Spotlight guest author! Join our large and growing group of guest contributors. Have you just published a scientific paper or have other exciting developments to share with the nanotechnology community? Here is how to publish on nanowerk.com.
