Development of low dislocation density gallium nitride substrates

Informations projet

DENIS

N° de convention de subvention: G5RD-CT-2001-00566

Projet clôturé

Date de début 1 Mai 2002

Date de fin 31 Decembre 2005

Financé au titre de

• Programme for research technological development and demonstration on "Competitive and sustainable growth 1998-2002"

Coût total € 2 041 604,00

Contribution de l’UE € 1 549 816,00

Coordonné par

ACREO AB
Sweden

Livrables

High pressure annealing method for modification of structural, optical and electrical properties of Group-II nitride semiconductor structures.

High pressure has proven to be an important tool in nitride technology and research. At atmospheric pressure GaN decomposes at temperatures lower than 1000-1100oC. High pressure of nitrogen up to 15kbar (1.5GPa) makes it possible to extend the accessible temperature range up to about 1500oC. Such temperatures are required for an efficient growth of GaN single crystals from the N-solution in liquid gallium, as well as for making the defects/impurities mobile (diffusion, defect annihilation, etc.), what is necessary, for example, for efficient removal of implantation defects Specially designed high pressure chamber has been tested by the UNIPRESS team. We have used undoped heteroepitaxial layers of GaN on sapphire applying high pressure annealing in nitrogen gas and in the pressure range of 10-16kbar, temperatures 1000 - 1500oC. We have found and improvement of the structural and optical properties of the used layers at some regions of temperatures and pressures. Particularly, Full Width at Half Maximum (FWHM) of the Xray reflections decreases by factor of 30% at pressures P=13-14kbar and T=1300oC-1400oC. Photoluminescence intensity (of donor-bound-exciton) can increase by factor 3-4 . The above obtained results reflect very likely a decrease in the concentration of threading dislocations in GaN/sapphire. It demonstrates the potential of the method for use in other technology related activity. In particular, annihilation of implantation defects and doping.

Growth of GaN/AlGaN wafers for high electron mobility transistor (HEMT) by metaloorganic vapor phase epitaxy (MOVPE)

At UNIPRESS, Intsytut Wysokich Cisnien, two approaches to a design of epitaxial structures forming a base for high power, high frequency GaN/AlGaN transistors have been employed. First one utilizes metaloorganic vapor phase epitaxy (MOVPE) and GaN substrates grown by hydride vapor phase epitaxy (HVPE). This approach makes possible to produce wafers of reproducible physical properties (carrier concentration and electron mobility) suitable for fabrication of high performance high electron mobility transistors (HEMTs). A proper choice of the growth parameters, content of Al in AlGaN barriers, thickness of barriers etc. enabled us to tune and optimise parameters of the grown epi-structures. These structures were sent to ACREO for processing them as transistors, which show parameters comparable with devices fabricated by leading Japanes/Korean and US institutions. The second approach consists of employing molecular beam epitaxy (MBE) growth technique for preparation of GaN/AlGaN heterostructures. Here bulk GaN crystals (doped with Mg) are used as substrates. It leads to a reduction in dislocation density (charged scattering centers) by few orders of magnitude in comparison with structures grown on SiC or sapphire (two to three orders of magnitude with respect to structures grown on HVPE-GaN. Within the second approach record mobility structures were achieved. They are highly suitable for basic research e.g., studies of new effects in Quantum Hall Effect regime.

A reactor for GaN thick wafer and boule growth

Currently LEDs, lasers and high frequency electronic devices are typically grown on foreign substrates, like sapphire, SiC or Si. This gives rise to a very high defect density in the devices, limiting their performance and sometimes their operating lifetime. The alternative possibility involves growth on GaN substrates. Such substrates of reasonably quality are presently only available in very limited quantities from a few companies, using the HVPE technique. The growth of such GaN wafers requires the development of suitable growth reactors, at the moment there is no such reactor available for the market in series production, most (future) GaN producers build their own proprietary equipment. There is thus at the moment an opportunity to develop a sizeable business volume in the area of HVPE reactor for production of bulk GaN wafers. At LiU there is collaboration with the SME Epigress AB in Lund, Sweden concerning development of such a suitable GaN HVPE reactor. A new construction has been developed during 2005; an additional upgrade is due at the end of 2005. Promising results are already achieved, such as the growth of a 2 mm thick 2¿ GaN wafer of high quality.