Twenty years ago, a small Finnish company called Microchemistry held the key to continuing Moore’s Law. From twelve contenders, it picked ASM International to scale up – to the great delight of Intel.
On 22 January 1999, a single visit radically overturned ASM International’s future. On that date, ASMI chief technology officer Ernst Granneman received two visitors from Finland: Kari Lampinen and Matti Ervasti. The former was a manager for mergers and acquisitions at a Finnish energy company called Fortum. The latter was a manager of a small subsidiary called Microchemistry, located in Espoo, Finland. Both men visited Bilthoven in order to find out whether ASMI was interested in acquiring Microchemistry and its technology.
Their pitch was about atomic layer deposition (ALD), a process developed by the founder of Microchemistry. ALD entails the controlled deposition of films in layers of single atoms. The process enabled the formation of the thinnest films with unprecedented conformity and step coverage while offering low contamination of unwanted impurities. Through the prevalence of Moore’s Law, technologies enabling the construction of materials at the smallest dimensions – atoms in this case – enjoyed the particular interest of the chip industry. It was just a matter of time before ALD would become indispensable for further miniaturization.
At first sight, ALD has some resemblance to chemical vapor deposition (CVD). In the classical CVD process, two or more reactants form a film of the desired material by introducing them in the reactor at the same time. In an ALD process, however, the desired reactant gasses are inserted sequentially into a reaction chamber. Taking turns, the reactant gasses react with the exposed surface until all available bonds are saturated.
Crucially, in ALD, the chemical reaction on the exposed surface automatically stops after all the available sites of the surface have been ‘occupied’, causing the substrate surface to be ‘saturated’. Adding more reactants wouldn’t result in more reactant on the surface. This principle of self-saturation is critical and enables the sequential deposition of layers of the same or different composition.
After saturation, a purge of an inert gas like nitrogen or hydrogen removes possible reaction products and all the remaining or surplus reactant molecules in the chamber. Next, the second reactant is inserted, which reacts with the first reactant until – again – all available sites of the substrate have reacted. Once more, the remaining reaction products and the redundant reactant molecules are purged out of the reactor. This complete process, called a “cycle”, adds one layer to the thin film being deposited. This cycle can be repeated until the desired film thickness is obtained.
The missing piece
Apart from the ALD process, the small Finnish company triggered the interest of Granneman with its treasure trove of processes. Over the course of many years, it had developed several pieces of equipment and various ALD chemistries. Through its tools, process knowledge and interactions with some chip manufacturers, Microchemistry singled out various applications in semiconductor production.
The most prospective was the gate stack. Microchemistry knew how to use ALD for the deposition of new exotic materials for the gate stack, for instance for the creation of high-k dielectrics and metal electrodes. By 1999, the first alteration of the CMOS gate stack in forty years of semiconductor manufacturing was already looming large on the horizon. It was this application that stirred the interest of chip manufacturers like Intel.
Microchemistry’s work on gate stack materials also perfectly aligned with ASMI’s work done in an earlier European research project and a joint development program with Siemens on integrated deposition processes in multi-chamber systems. Process and material-wise, Microchemistry’s techniques formed the missing piece. Moreover, some of its tools were designed according to the industry’s Material and Equipment Standards and Code (MESC). In other words, they could be integrated into multi-chamber tools developed by ASMI. All things fell in place.
One week later, after consulting ASMI’s founder and CEO Arthur del Prado, strategic marketing manager Chris Werkhoven, the newly appointed CTO Ivo Raaijmakers and resigning CTO Granneman e-mailed Lampinen to confirm ASMI’s interest in a joint future with Microchemistry. He argued that the technology clearly matched ASMI’s technology roadmap and that the company would fit in the ASMI organization. The organizational structure of ASMI as designed by Del Prado allowed the Finnish start-up to grow into an established original equipment manufacturer rather independently.
After the initial meeting in January, things moved fast, in particular within ASMI. Werkhoven and Granneman visited Microchemistry to learn about the technology. Del Prado, Granneman, and ASMI’s CFO Rinse de Jong worked out the financial details with Ervasti and Lampinen. By July, Del Prado and De Jong succeeded in finding “an acceptable creative financial solution” and finalized the acquisition. The firm was now called ASM Microchemistry.
In 1999, Microchemistry employed 42 staff, of whom 14 in development, 12 in engineering and 8 in manufacturing. It had four types of machines, including the F120 and the F200 for semiconductor production. The F120 was a mere R&D reactor, only capable of processing very small substrates and primarily meant for process development. The F200 was designed for semiconductor manufacturing. Technology development agreements were in place or were negotiated with Intel and IBM, while others, such as Imec, were considered.
Soon after the acquisition was finalized, positive signals were received. Major chip manufacturers congratulated ASMI after Werkhoven contacted prominent customers surrounding ASMI’s due diligence of Microchemistry. They had anxiously awaited the outcome of the acquisition process. “A few hours after they called. Also about Microchemistry, congratulating us they called a very good buy, quickly concluded and successful in competition with 12 (!) other contenders [sic]. So, a job well done,” Werkhoven faxed to headquarters.
Key supplier
Chip manufacturers had a reason to be happy. As some Intel engineers involved in the process development of the gate stack later recalled in IEEE Spectrum, “For the first two years, all the dielectrics we tried worked poorly. (…) You want a transistor to operate exactly the same way every time it switches, but these gate stack structures behaved differently each time they were charged up. The results were very discouraging.”
“To make the dielectric layer, we were using one or two different semiconductor manufacturing techniques: reactive sputtering and metal-organic chemical vapor deposition. Unfortunately, both processes produce surfaces that, though remarkably smooth by most standards, were nevertheless uneven enough to leave some gaps and pockets in which charges could get stuck. We needed something even smoother – as smooth as a single layer of atoms, actually.”
Depositing such thin films in a controlled manner wasn’t possible with any other method. Process-wise, ALD had the best papers, even though reproducible high-volume manufacturing still had to be established.
The acquisition of the rather unknown company and its novel technology set ASMI on a trail packed with uncertainties but with high expectations as well. The positive comments from customers, some of whom were the most advanced semiconductor manufacturers at the time, stimulated confidence in the path chosen. “The announcement of the intended acquisition has created a lot of positive momentum for ASMI from customers like Intel, IBM, Philips and STMicroelectronics. The combination of Microchemistry’s technology with ASMI’s global infrastructure solves the main issue from these customers that Microchemistry on its own was too small to support the technology and create a production-ready solution,” reported management to the somewhat skeptical supervisory board of ASMI. The Bilthoven company’s introduction in the process did place the decision-making on a different level at the customer. For ASMI, this was a very positive signal.
After the formalization of the acquisition of Microchemistry, it was up to ASMI’s management to capitalize on the momentum. A new approach had to be developed to enact the newly obtained ALD technology. The matter was quite urgent indeed. ASMI had been propelled to the lead in ALD and gate stack technology. Moreover, the major chip manufacturers were impatient. Moore’s Law defined the pace of their technological development and the gate stack constituted a rapid approaching obstacle on their route.
Microchemistry was a gift from heaven: unexpected, but most welcome. It was up to ASMI to fulfill its promise.
This is an excerpt from the PhD dissertation entitled “Fortunes of high tech: a history of innovation at ASM International, 1958-2008” by Jorijn van Duijn, which is now available at Techwatch Books.