Following federal funds, BluGlass is eyeing GaN on silicon LED markets.
BluGlass now hopes to take its remote plasma CVD process to LED and power electronics markets. [Credit: BluGlass]
In July this year, Australia-based semiconductor process developer, BluGlass, won Aus$3 million in federal government funding to demonstrate high efficiency, low cost GaN LEDs.
The start-up has spent years honing the low temperature growth of GaN, and more recently the growth of p-GaN layers, on MOCVD GaN templates but will now ramp up efforts to grow entire LED structures on silicon substrates as well as sapphire wafers.
“We’ve always wanted to get stuck into GaN on silicon, particularly with larger wafers and now intend to develop full LED structures,” explains company chief technology officer, Ian Mann. “Our lower temperature remote plasma CVD has the potential to reduce the bowing and cracking problems of traditional high temperature MOCVD… and we can see emerging opportunities in GaN on silicon markets.”
To achieve low temperature growth, BluGlass replaces the ammonia source of a typical MOCVD system with nitrogen gas, passed through an electrical coil to generate a plasma.
During conventional MOCVD, temperatures of up to 1200ºC dissociate nitrogen from ammonia, and GaN layers are grown. But by directly supplying nitrogen via a plasma, BluGlass deposits the semiconductor layers at much lower temperatures.
To date, the company has not revealed actual growth temperatures but earlier this year it showcased p-GaN films, grown via RPCVD on MOCVD GaN templates, that met industry electrical performance benchmarks. Mann claims the process is cheaper than MOCVD as large amounts of ammonia are not used, although a RPCVD deposition tool will cost about the same as an MOCVD system.
His team is now working towards growing p-GaN layers at low temperatures and aim to grow at temperatures matching those used in MOCVD growth of multi-quantum wells.
“Today we buy multi-quantum well layers made on an MOCVD machine and we grow p-GaN layers on top of these,” says Mann. “But with our technology, some of the real value will come from growing the multi-quantum wells and then the p-type GaN layers at the same temperature all via RPCVD. We believe this is achievable.
And crucially, as Mann highlights: “A low temperature p-GaN layer has the potential to be less damaging to the MQW and therefore improve the efficiency of an LED.”
The recent federal funding has enabled the company to buy a second, larger MOCVD system, that Mann and colleagues will now retrofit to a RPCVD system. The CTO emphasises that the system will be smaller than the size of a production facility in a tier one fab’, but will allow the team to work with single, eight inch wafers.
“The system will allow us to demonstrate eight inch silicon, which is a must-have and isn’t something we can do with our current development equipment,” he says.
The team is currently developing nucleation and multi-quantum well layers, and while the RPCVD process will differ from conventional MOCVD, Mann asserts: “We will use RPCVD to grow a full structure in the same way that MOCVD can grow a full structure.”
“At the same time we are also looking to commercialise as fast as we can and will draw on other resources, potential partnerships and other funding to accelerate this,” he adds.
But what of the company’s starting technology? In the summer of 2007, BluGlass demonstrated the world’s first blue light emission from GaN deposited on glass. But as Mann points out, BluGlass is now focused on the conventional substrates that the LED industry is familiar with. And the company doesn’t expect to stop at LEDs.
Looking to the future, and beyond the latest federal funding, BluGlass also intends to take its technology to power electronics markets. “Clearly the power electronics market is an emerging market as well,” highlights Mann. “And from a technology point of view, getting good quality GaN growth on silicon is important for both power electronics devices and LEDs.”