Paul van Gerven
24 February

By fostering collaboration between companies focusing on atomic layer deposition as well as infusing them with the latest research results, a new TUE-led project aims to take the already strong Dutch presence in this field to the next level.

There’s a remarkable number of companies in the Netherlands that concentrate on atomic layer deposition (ALD). ASM International (ASMI) builds ALD equipment for the semiconductor industry. The tools of Levitech, Smit Thermal Solutions and Solaytec are supplied to the PV arena. Saldtech targets the flat panel display industry and SALD, a spinout from Solaytec, has set its sights on battery components and textiles, among other things. Additionally, two major institutes, Eindhoven University of Technology (TUE) and TNO, perform major ALD research activities.

Even though most of these companies share a common heritage – except for Smit, they trace their origins back to either ASMI or TNO – they tend to go it alone. Most collaborate with the research institutes but otherwise only interact on a superficial level. When push comes to shove, there are always bigger priorities to deal with than exploring ways to benefit from each other. Not to mention the risks involved with sharing too much with your, sort of, competitor.

SALD
Credit: SALD

This situation isn’t good for anyone, says Erwin Kessels, professor at TUE’s Plasma & Materials Processing group, which specializes in ALD. “Even though these companies focus on different markets, they all run into similar issues, especially if they’re trying to take the technology another step forward. I get asked the same questions over and over again, so to speak.” In other words, there’s a large common ground, yet no one is setting foot on it.

Longer-term, the lack of collaboration means commercial opportunities may be lost, Kessels continues. “ALD has a lot of potential applications, many end-users are interested in it. But the technology isn’t quite there yet. Most Dutch ALD companies are startups or SMEs – they don’t have the means to pursue these opportunities by themselves. By working together and sharing knowledge, we stand a much better chance of beating the competition to money-making applications. We have an excellent starting position, it would be a shame if we fail to capitalize on it.”

Kessels must have made a persuasive argument because he succeeded in uniting the above-mentioned companies, along with system supplier VDL ETG, Finnish thin-film test specialist Chipmetrics and US spectroscopic instrument manufacturer JA Woollam, in the “Spatial atomic layer deposition – more materials, more demanding applications” project. Co-funded by science funding agency NWO, Kessels’ group will take a deep dive into spatial ALD, a particular variety of ALD that most project partners focus on. The results will feed the companies with technology and, hopefully, foster more collaboration, thus giving the Dutch ‘spatial ALD hub’ a decisive boost.

Presumptuous

ALD involves sequentially exposing a substrate to different gases, which react with the surface in a self-limiting fashion, thus ‘coating’ it literally atomic layer by atomic layer. The difference between traditional (ie temporal) ALD and spatial ALD is that in the latter, the substrate is led through the process gases, whereas in temporal ALD, it’s stationary while a reactor is sequentially filled and emptied. The continuous character should make spatial ALD the superior technique for high-volume industrial applications.

As Levitech and Solaytec have found, however, temporal batch reactors can still put up a mean fight. Independently from one another, these two companies developed a spatial ALD tool to deposit an efficiency-boosting layer of aluminum oxide on solar wafers – a research field pioneered by Kessels, by the way. Both Levitech and Solaytec sold tools to pioneering PV manufacturers of high-end solar cells, but spatial ALD never became an industry standard. PV manufacturers currently prefer relatively inexpensive, large-scale batch reactors, which are less demanding in maintenance, have a smaller footprint and top the throughput of spatial ALD reactors.

Levitech Levitrack
Levitech’s Levitrack spatial ALD system (long machine on the left) in a solar cell production line. Credit: Levitech

“This should be an important lesson for everyone,” says Frank Verhage, who was responsible for starting Solaytec’s sister company SALD to match the company’s expertise with other markets and applications than solar. “It’s presumptuous to think that a small company, all on its own, can successfully develop and market such a complex machine. Solaytec’s fate shows the importance of connecting to others on a deep level, both to customers and partners.”

Like Kessels, Verhage noticed that many of his ALD peers didn’t see it that way: they tended to play their cards close to their chests. “They looked inwardly and primarily thought in terms of competition, not collaboration. I’m confident, however, that our project will change perspectives and bring us closer together,” says Verhage.

Untapped potential

Isn’t there another lesson to be learned here? Namely, that temporal ALD is a fierce competitor that isn’t easily bested by spatial ALD? “Depositing aluminum oxide on a solar wafer is just about the easiest ALD process you can think of. There are more demanding applications that are a much better match for spatial ALD,” says Bart Macco, project leader and researcher in Kessels’ group.

Crucially, spatial ALD technology doesn’t involve a vacuum. Macco: “This means spatial ALD is uniquely suited to handle foils or other flexible substrates, especially in high-throughput roll-to-roll and sheet-to-sheet processes.” Additionally, in spatial ALD, exposure to elevated temperatures can be limited. “This can be a key advantage in the production of sensitive products, such as perovskite solar cells and OLEDs. We also strongly suspect spatial ALD has the edge when it comes to depositing mixed or doped materials, which will expand the range of characteristics we can impart on the deposited layer.”

“Diversification of substrates and material palette are two core elements of our project. This is because these are the elements that enable spatial ALD to unlock new applications for ALD but also because that’s where the uncharted scientific territory is,” says Kessels. “The invention of ALD goes back to the 1970s, but it wasn’t until about a decade ago that spatial ALD started gaining attention. This means that there’s still a lot unknown about it,” adds Macco.

Kessels is happy to get a chance to uncover some more of spatial ALD’s secrets, while precompetitively supporting industrial activities at the same time. “It’s increasingly difficult to find public funding for this kind of technology research. Technologies like quantum computing may catch the imagination, but they’re further out in terms of commercial application. As a knowledge-based economy, we shouldn’t ignore our current industrial base and all the untapped potential that’s right there for the taking.”