Epitaxial GaN on Graphene substrate

Epitaxial GaN on Graphene substrate

Due to their excellent physical properties, III-nitrides are highly prized semiconductors for the fabrication of microelectronic and optoelectronic devices. Although many III-N-based devices are already on the market, their performances can still be impacted by a high density of defects in the active layers. This is due to the lack of a native III-N substrate available at low cost. These III-N layers are currently grown by heteroepitaxy, i.e. on a bulk substrate of different nature. It therefore appears important to identify an alternative substrate to improve the crystalline quality of the active layers. In this context, my PhD work aims at the elaboration of monocrystalline GaN domains of micrometric size that can be used to fabricate a single device on each of them. I used an innovative process based on two important steps: (i) molecular beam epitaxy (MBE) of nanostructures of excellent crystal quality on an ultimately thin and compliant substrate, i.e. a single graphene layer transferred onto silica; (ii) lateral regrowth of these nanostructures by organometallic vapor phase epitaxy (EPVOM). My results first illustrated the problem of GaN nucleation on graphene and, more generally, of a 3D material (with sp3 orbitals) on a 2D material (with sp2 orbitals). In particular, a very long incubation period is observed before nucleation of the first GaN seeds. I have tried to explain the origin of this incubation time and identified experimental conditions to shorten it. I have shown that the incubation results in important structural modifications of graphene. It is likely that these modifications create GaN nucleation sites. Then, we measured in situ the deformation of the first GaN nuclei at ESRF. At the very beginning of nucleation, GaN is under a tensile strain of 0.8%. A growth scenario based on all these experimental observations is proposed. I also showed that selective epitaxy of GaN nanowires is possible on very small graphene patches on silica. With optimized conditions, I could obtain a single nanowire per patch. My preliminary results on the lateral regrowth step demonstrate the possibility to obtain micrometer-size GaN crystals without extended defects. I studied their growth kinetics using AlN markers. A highly anisotropic growth regime corresponding to the progression of vertical facets can be reached. This represents an ideal condition for developing GaN micro-templates. This approach by selective epitaxy on patterned graphene offers very promising prospects for the transfer of micro-devices to a host substrate, different from the growth support material.

Thesis manuscript: https://theses.hal.science/tel-03469445

Top image: characterization and elaboration chamber for in situ sutdy of GaN nanowire on graphene surface. Atomic surface structure and nanostructure objects morphology can be determined by grazing incidence wide and small angle X-ray scattering (GIXD and GISAXS) during growth.