Industry veteran aims to steer GaN from start-up to mass production

The shrewd investors who snapped up the shares of Australian GaN epitaxy start-up BluGlass must be basking in their blue-tinted glory. After the company was floated on the Australian Stock Exchange in mid-September, its share value jumped 67% from AUS$0.30 to AUS$0.50 (US$0.39). And flicking through their corporate presentation it s easy to see why. Over the past 10 years the GaN industry has grown more than a hundred fold and by 2009 forecasters predict it will be worth more than US$7 billion (Strategies Unlimited).

BluGlass is still in its infancy. The firm was conceived in July via the acquisition of Gallium Enterprises, a company originally set up by the commercial arm of Macquarie University in Sydney after a decade-long research project to develop low-cost GaN epitaxy. Immediately afterwards, BluGlass launched an initial public offering (IPO) of shares to raise AUS$6 million. However, as the deadline drew closer, unrelenting demand pushed BluGlass to release a supplementary prospectus to attract even more cash. All told, BluGlass raised AUS$10 million through the IPO.

This capital now rests in the hands of the new CEO, David Jordan, for whom the semiconductor industry is second nature. "Most of my time before BluGlass has been almost exclusively spent in the photovoltaics and silicon semiconductor field," he says.

"In my previous role at BP Solar I managed to interface research and development programs and high-volume manufacturing worldwide. I very much see that as the basis of what we re doing here at BluGlass. We are taking a very promising, university-based technology and examining its potential as a business case, with the hope of taking it on to full commercialization and high-volume production."

Most GaN material produced today is manufactured through MOCVD. Though an effective, industry-proven method, MOCVD requires high temperatures in the region of 1000 °C and the use of highly toxic ammonia gas. The technology that BluGlass is pioneering, called remote plasma chemical vapor deposition (RPCVP), requires no ammonia and proceeds at the slightly more benign temperature of 700 °C. This means that hardy but relatively expensive wafer materials such as sapphire or SiC can be substituted by glass (hence the company name) at a fraction of the cost. French firm Saint-Gobin Recherche, which supplies both glass and sapphire for LED production, has just signed an 18 month deal to co-develop specially engineered substrates for BluGlass.

RPCVP is the subject of two international patents and one Australian patent held by BluGlass. These describe passing nitrogen plasma over the heated gallium. "The gallium is transported into the reaction chamber in a nitrogen carrier gas stream through a trimethyl gallium buffer," explains Jordan. The BluGlass team has already reduced the surface roughness of its thin films from 9–13 nm down to 1 nm, thus improving crystallinity and suggesting that a high-quality GaN layer is produced.

"We are also attacking the scalability," Jordan says. "BluGlass can already deposit at 4 inch diameters, whereas most of the MOCVD industry deposits on a diameter of 2 inches. And there s no reason why we can t extend that to 8 inches, 16 inches or beyond."

If this scale-up can be achieved, it points to a high throughput of cheaper GaN devices, which have become ubiquitous in consumer technology. For blue LEDs, the mainstay of the GaN market, 68% of applications are in LCD backlighting such as mobile phones and digital cameras. Just 4% are in general lighting, but this could all change if BluGlass can make an impact. Every year, a fifth of the global electricity supply is used in lighting applications, most of which is still provided by inefficient bulbs. When it is possible for white LEDs to be mass-manufactured at a much lower cost, this huge market beckons.

"There is no doubt that in terms of efficiency white LEDs are currently on a par with compact fluorescent tubes and are significantly better than incandescent bulbs," says Jordan. "Compared with the light from fluorescent tubes, which is fairly harsh, white LEDs can be tuned to give a much softer light. In five years or so LEDs are set to take a very significant proportion of the general lighting market," he predicts. "All we have to do is to address the cost, which is still too high. Once the cost comes down, the market will open up."

There is another emerging market that could benefit from economic GaN optoelectronics: high-definition (HD) video. Found in Sony s new Playstation 3 games console and Blu-ray Disc players, HD discs can store 10 times more data than a regular DVD. The technology has been slow to take off because of problems making blue laser diodes and their resulting high cost. BluGlass s epitaxy could offer an attractive, lower-cost alternative to MOCVD-based laser diode production and help to ensure that HD discs fulfil their potential as the natural successor to DVDs.

Despite the three patents under its belt and ample investment, Jordan is revealing no definite indication of how exactly BluGlass is going to generate sales revenue. "We will be working with strategic partners, either licensing the technology once it s proven, or forming joint ventures with different groups," says the CEO. "But that s down the track, and we need to prove that we are able to hit the targets of cost-structures and device performance that we believe this technology offers us."

Having already demonstrated a functional single p-n junction LED using the RPCVD method, the next step for BluGlass will be to complete a pilot production plant in collaboration with cleanroom specialist M+W Zander. "This is an 18 month to two-year program," says Jordan. "We envisage moving into the pilot production plant at the beginning of next year, and then the commercial tools that we will use to implement the technology will be put in place in the first and second quarter. All of those things are required and they re coming together now."