Europeans aim to complete GaN supply chain

Engineers are eagerly awaiting the results of the first batch of GaN wafers fabricated under a new European research project that could kick-start the region s development of the wide-bandgap semiconductor.

The "GREAT2" project, which officially began in June 2008 with €8.6 million ($11.4 million) backing from the European Space Agency (ESA), is scheduled to deliver its first results this spring.

According to the project co-ordinators, round-table workshops held over the past couple of years had highlighted a "missing link" in Europe s semiconductor expertise "“ the lack of a reliable GaN device manufacturing process and associated space-compatible component supply chain.

With GaN development handsomely funded in the US in particular, that "missing link" has become a strategic issue, especially as ITAR (International Traffic in Arms Regulations) trading restrictions could make it impossible to source GaN components manufactured in the US:

"If Europe s space industry is to maintain its competitiveness, then it needs guaranteed access to leading-edge GaN technology," said ESA.

Reliability drive
Talking to compoundsemiconductor.net, ESA s Andrew Barnes and Mikko Nikulainen highlighted that the major challenge at the moment is providing sufficiently reliable devices.

"We are trying to correlate material issues and defects in substrates, and in epitaxy and wafer processing, and link these factors together in order to improve device reliability," said Barnes.

As a result, many of the key technical milestones in the first, three-year, phase of the project relate to improvements in the mean-time-to-failure (MTTF) metric.

At the moment, this stands at around 1000 hours, but the project is ultimately aiming for a thousandfold improvement.

Barnes highlighted that understanding the device physics and failure mechanisms is very important for reliability of GaN devices. Accurate assessment of channel temperature is central to this improvement, and this is where one of the project partners "“ the UK s University of Bristol "“ will provide expertise through a high-resolution thermal characterization technique.

Performance targets
Power output and power-added efficiency (PAE) targets have also been set as key milestones under GREAT2, although precise goals for components operating in the L, S, X and Ku bands of the microwave spectrum are being kept under wraps for now.

While Barnes acknowledges that Europe is playing catch-up to US and Japanese teams as far as power and PAE are concerned, he has been encouraged by recent progress outside the project.

"We are starting to catch up regarding pure performance, but the biggest problem remains the reliability, particularly for space applications," said the senior microwave component engineer.

By mid-2011, when the first phase of GREAT2 ends, he anticipates Europe will have access to a reliable GaN process that will also have many applications in commercial sectors.

"An important difference within GREAT2 is that we have combined together some of Europe s leading research institutes together with the industrial component manufacturer UMS (foundry partner United Monolithic Semiconductors)," said Barnes.

"The aim is for UMS to identify the best process options that are compatible with achieving a space compatible GaN foundry process and industrialise them."

"By 2011, we hope to have a preliminary space-compatible process, and to make the process available for other defense and aerospace applications. The aim is to build a European supply chain for GaN."

Supply chain gaps
Nikulainen, head of the European Component Initiative (ECI), cites the lack of a suitable substrate supplier as one of the more obvious gaps in the European supply chain, as things stand currently.

Although no target date has been set for the first ESA mission to deploy GaN technology, Nikulainen says that the agency will soon be engaged in active talks with end users. Applications of the GaN components will also take place using technology demonstrator spacecraft initially, however these in-orbit demonstration missions have yet to be defined.

Ultimately, GaN technology should result in major benefits for ESA. Chief among these is the increased RF output power capability for microwave signal transmission backto earth. This can allow system benefits such as improved transponder EIRP (equivalent isotropically radiated power), or, since the power density of GaN is five times that of GaAs, a satellite transponder with equivalent performance that is far smaller and lighter.

On top of that, explained Barnes, GaN s ability to deal with much higher temperatures will simplify thermal management, while the wide-bandgap semiconductor s inherent radiation hardness could reduce the degree of radiative shielding that is required.

Assuming that phase one of the project goes according to plan, phase two is expected to begin in 2011, with anticipated support of €4 million.

The full list of GREAT2 project partners comprises: TESAT Spacecom (project management); IMEC (epitaxy optimization); the Ferdinand Braun Institute for High-Frequency Technology (reliability and transistor processing); the Fraunhofer Institute for Aplied Solid-State Physics (epitaxy and MMIC processing); United Monolithic Semiconductors (technology transfer and process development); the University of Bristol (thermal analysis) and the University of Rome Tor Vergata (device physics modelling).