Kyma Responds to CS Article entitled “Bulk GaN: Ammonothermal Trumps HVPE”
“We have mixed feelings about responding in the public domain, but in the end we have chosen to do so,” said Dr. Keith Evans, Kyma president and CEO. “Kyma’s stakeholders can rest assured that the GaN HVPE growth process is here to stay, and that HVPE will remain the primary technology used to make GaN boules from which commercially available substrates can be sliced and polished.”
Kyma uses the hydride vapor phase epitaxy (HVPE) process for making GaN substrates. Along with others, Kyma has shown that the quality of HVPE GaN improves as it grows thicker after starting on a foreign substrate (which is usually sapphire). If high-quality GaN seeds were to become available, the foreign substrate could be eliminated and the value proposition of HVPE for GaN device applications would be further strengthened.
“We read the article as an affirmation that HVPE is the commercial leader in large part due to its high growth rate, high purity, and its demonstrated benefits in commercially important devices. HVPE’s primary challenge is the lack of availability of good seeds. In contrast, ammonothermal growth makes higher quality crystals, but at growth rates of 2-4 micron/hr. This is well below the 1mm/hr that is typical of all commercially successful semiconductor material manufacturing processes (including hydrothermal quartz), and therefore ammonothermal is not in the market,” stated Evans.
“We believe that the article represents the perfect argument for employing ammonothermal grown GaN seeds in an HVPE growth process,” stated Dr. Ed Preble, Kyma COO and VP Business Development. “Such a combination represents the ultimate in terms of GaN substrate product quality, cost, and time-to-market.”
Kyma hopes that ammonothermal grown GaN seeds will become commercially available, while in parallel it is developing new HVPE based seed generation techniques that potentially render the foreign substrate a non-issue.
“Our latest HVPE-only seed expansion and boule growth results show great promise and benefit due to recent innovations at Kyma,” added Preble. “Our intent is to develop seeded boule growth processes which effectively put foreign seed based GaN defects and the associated single wafer limited liftoff processes behind us. This will drive GaN prices from over $1000 per 2” wafer down to less than $100 per 2” wafer.”
Additionally, there are many other secondary parameters and growth technology subtleties that are also important with respect to the manufacturability of a GaN growth process, such as the controllability of n-type and compensation doping, the temperature dependence of substrate electrical conductivity, presence of unintentional background impurities, presence of native vacancies and anti-site defects, and the ability to expand the crystal in one or more crystallographic directions. Lastly, 100% recyclability of unused gallium in Kyma’s HVPE growth systems has allowed Kyma to trade off growth efficiency for process stability and speed, eliminating the need to have high efficiency gallium utilization in each growth run. Kyma believes these many secondary issues favor HVPE over ammonothermal growth.
click here to link to Bulk GaN: Ammonothermal trumps HVPE article http://www.compoundsemiconductor.net/csc/features-details.php?id=19645907
