We ate with scalpels and tweezers, says Aixtron founder

The III-V community was the butt of quite a few jokes in the early 1980s. "GaAs makes papers and silicon makes money" was one popular wisecrack, alongside the famous one-liner: "GaAs is the material of the future and always will be."

Like all bitter jokes, they featured an element of truth. Back then research efforts overshadowed device manufacture and production was limited to fabrication of rather dim green, orange, yellow and red LEDs based on AlGaAs, GaAsP and GaAs – even infrared lasers for future CD players were still under development.

But this community was undergoing a transition, with research edging towards commercially viable applications. Success, however, would hinge on the development of stable, consistent manufacturing methods for making epiwafers. Into this breach stepped a couple of German researchers at the University of Aachen, and the rest, they say, is history.

Aixtron co-founder Holger Juergensen was one of these researchers. "By 1982, it became clear to Meino Heyen and me that the time was right to offer well developed, standardized equipment to researchers that allowed them to obtain reproducible results in a uniform manner on full wafers. Even in those early days, we had started to conceptualize MOCVD equipment growing III-V films on a full 2 inch wafer. This was revolutionary – at the time most researchers could only grow on a few square centimeters."

Premises for the company were secured in Aachen, which is close to the borders with Belgium and the Netherlands. "The proximity to the university was a key to Aixtron s success," said Juergensen. "Our first location was inside a university technology center that offered attractive conditions to start-up companies." That kind of set-up might seem normal these days, but incubator-style technology centers were a radical idea at that time.

Launching any company requires much more than a good idea and business premises – healthy finances underpin everything. This is where Heyen s father in law, the successful businessman Heinrich Schumann, stepped in as Aixtron s third founder.

Although this triumvirate offered the ideal mix of III-V expertise and business acumen, they still found it tough to generate the cash needed to start the company. Securing some public funding to build an initial prototype was the breakthrough that they needed. "It took us about a year to identify and qualify appropriate suppliers, get the first team together and fix the design," said Juergensen. According to him, the biggest challenge was converting a research tool into a user-friendly, reliable piece of kit that could be built and sold commercially.

The founders then received some great assistance from the III-V team at AEG in Ulm, Germany, who wanted to create faster GaAs-based communication chips. "They simply went ahead and bought Aixtron s first MOCVD equipment with an upfront payment," recalled Juergensen. In addition, the founders secured a small loan from a local bank. "The loan in itself was almost revolutionary, as we could not provide any of the typical securities."

Europe is well known for its 35 hour working week, with the focus on time with the family and vacations that knock out August. But the founders of Aixtron bucked this trend by working day and night, and through weekends and holidays.

This complete commitment to the company was exemplified by the relationship with AEG. Teething trouble with this lead company s reactor led to visits to Ulm, a city at the opposite end of Germany, four or five times a week. "We often worked 10–12 hours and drove 1100 km per day," said Juergensen. "I don t remember how we got any sleep."

Heavy workloads are not unusual for the founders of start-up companies, who can find themselves eating together at their premises in the evenings. For a semiconductor equipment firm, even this poses some unique challenges. Understandably tired, the Aixtron pioneers sometimes forgot to bring knives and forks through the lab s airlocks. "We had to enjoy the great variety of lab food using small medical scalpels and tweezers that were used to cut and handle GaAs wafers," reminisced Juergensen.

Legacy products
Aixtron s first shipment, which went to the researchers at Ulm, was a single 2 inch machine with an AIX 200 quartz reactor cell, which shares many similarities with the firm s current product range. It featured computer control, safety interlock systems and many other hardware details that were maintained in future designs. "That said, it s a bit like comparing an old and a new Mercedes E class," said Juergensen. "Although today s generation shares the same heritage as those first generation tools, the performance is of a completely different class."

The early years of a start-up often include times when money runs out, and founders have to beg for cash. But Aixtron managed to steer a far safer course and was profitable from its very first year. "I don t think we ever came close to a really dangerous situation," said Juergensen, "but that is due to the strong focus we put on financial management. We paid and took low salaries, keeping all of the cash within the company."

Ulm s breakthrough order was followed by others from SEL in Stuttgart (which later became part of Alcatel) and the Universities of Berlin and Duisburg. Later, Aixtron shipped a tool to Philips in Eindhoven, the Netherlands, to develop the first commercial red AlInGaP lasers for DVD players. This signaled the start of an international business, and it wasn t long after that when AIX 200 reactors were dispatched to AT&T for growing InP-based lasers and detectors at its sites in Murray Hill, NJ, and Reading, PA.

In 1988 Aixtron launched its first multi-wafer reactor. "For the first time, this reactor enabled high-volume, industrial-scale manufacturing of III-V compounds," remarked Juergensen.

The 1990s witnessed the launch of more tools and the emergence of what is now the firm s key application: nitride-based high-brightness LEDs.

The growing company targeted manufacturers of the first blue, green and white LEDs, and in late 1997 the Aachen outfit launched itself on Frankfurt s "Neuer Markt" stock exchange. The capital raised ultimately enabled expansion of the company through acquisitions of Thomas Swan Scientific Equipment and Epigress AB in the 2000s.

Diversification has been the latest theme. The focus on gas-phase deposition equipment remains, but the company is now looking to cover a broader range of materials. This includes the development of a deposition technology for making organic LEDs, and a product line for silicon-related materials via the further acquisition of the California-based equipment manufacturer Genus Inc in 2005.

"These are meaningful events," said Juergensen, "because they mark our commitment to developing technology to meet the inevitable convergence of compound and silicon technologies." Aixtron s tools are in fact already appearing in silicon fabs, where they are used to deposit hafnium-based gate material. Far from being the butt of the silicon community s jokes, MOCVD is now becoming a technique that is central to its future.

Who would have spotted that marriage on the horizon when Aixtron launched the first MOCVD tool 25 years ago, amid the taunts of the blinkered silicon community?   

View pdf of article