With its first tool sales in the books, Eindhoven-based ‘startup with experience’ SALD is ready to introduce atomic layer deposition into a wide range of new markets.
In the world of materials deposition, nothing trumps atomic layer deposition (ALD) when it comes to precision. The technique offers control over film thickness and composition at the atomic level, while completely following the contours of a surface. This precision comes at a price, though: in its traditional form, ALD is incompatible with a continuous process. This limits the manufacturing volumes that can be handled cost-effectively, rendering ALD unsuitable for a large number of commercial applications.
SALD has set out to change that. The company founded in 2019 has developed equipment for continuous ALD on a wide range of substrates, enabling a number of industries that previously wouldn’t dream of using ALD to apply the deposition technology. “We’re working with companies and research institutes from all over the world to see how they can take advantage of ALD’ed functional nanolayers,” says Peter Visser, business developer at SALD.
Several customers are already convinced and purchased an R&D tool from the Eindhoven-based company. Once these and future customers have perfected their film recipes, SALD is confident that it can swiftly follow up with matching production equipment. “Not only can we rely on the vast system integration and equipment manufacturing expertise of the Eindhoven Brainport area, SALD is a startup with experience,” says Visser.
To see how this can be true, we need to go back to 2010, when research institute TNO spun off Solaytec to commercialize ALD equipment for solar cell manufacturing. Or, to be more precise, to commercialize spatial ALD equipment for solar cell production.
ALD, in general, 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. In traditional (ie temporal) ALD, the substrate remains stationary while a reactor is sequentially filled and emptied with the appropriate gasses. In spatial ALD, on the other hand, the substrate is led through alternating zones of gases, building up a film as it moves.
Solaytec’s spatial ALD equipment was designed to deposit a nanolayer of aluminum oxide on solar wafers, which resulted in additional efficiency compared to other deposition techniques. Several high-end solar cell manufacturers bought the company’s SALD tools, and these are still running today, but as the aluminum oxide layer became more and more mainstream in solar, so did much cheaper Chinese batch ALD tools.
This forced Solaytec to investigate whether spatial ALD could expand its horizon. When the answer turned out to be a resounding yes, sister company SALD was founded. With a decade of experience under its belt, however, it obviously doesn’t have to start from scratch. “When three billion wafers have passed through your equipment, in a part of the world where maintenance isn’t necessarily high-priority, you learn a thing or two,” assures SALD CTO Erik Kremers.
That doesn’t mean that Solaytec’s machines can be used for other applications. After all, they were designed for a very specific purpose: deposition of aluminum oxide on solar wafers. Processing differently sized, thicker or thinner or flexible substrates wouldn’t work, and the technology was never intended for deposition of any other material than aluminum oxide.
And so, for the past two years, SALD has been developing a much more versatile version of spatial ALD. “This all boils down to the deposition head, which we turned into a modular and scalable concept,” explains Kremers. A deposition head leads a gaseous precursor through parallel channels onto the substrate moving underneath. Since the ALD process consists of two complementary half-reactions, two heads in sequence are required to complete the deposition of a single ALD layer in one pass. ‘Curtains’ of inert gas built into each head prevent the reactants from intermixing.
Kremers: “To deposit more layers, as many heads as needed are installed to reach the desired layer thickness. That type of configuration is only relevant for high-volume production, though. In our R&D tool, for which a small footprint is preferred, the substrate oscillates underneath three alternating heads in an A-B-A configuration. Crucially, we employ the same deposition heads in our R&D and production tools. So turning R&D into an industrial process will be smooth.” SALD is currently assembling its R&D tools in-house but plans to outsource production of industrial-volume tools to a system integrator in the Brainport region.
The deposition head’s width, typically a few centimeters, is determined by the deposition characteristics of the particular gas flowing through it – SALD’s system can be used to deposit nanolayers of many different materials, even mixing materials and layers is an option. The head’s length, however, is flexible to accommodate a range of substrate widths, which vary for different applications. “We switched to 3D printing to manufacture our rather complex deposition heads. That way, we can easily change the head’s length to the customer’s specification. And it’s a very reproducible method to boot,” reveals Visser.
“Solaytec’s heads are made from aluminum. We switched to stainless steel, which is inert and hence easier to clean by chemical means. Thanks to our new deposition head design, contamination issues are greatly reduced, but should the need arise to clean a head, it’s easy to remove and replace it with another for offline cleaning,” adds Kremers.
The last piece of the puzzle to broaden the scope of spatial ALD is substrate handling. Visser: “Solaytec’s systems used gas cushions to transport the wafers. That method isn’t compatible with many substrates, like battery electrodes or plastic foils. So we engineer appropriate solutions to move our customers’ substrates past the deposition heads. Our R&D tool features an oscillating stage, while we expect high-volume applications to embrace roll-to-roll production. We plan to demonstrate that next year.”
There’s more on the roadmap of the company currently employing 25. “Many ALD processes require elevated temperatures, if only to prevent reactants such as water to condense. That’s why we’re working on adding plasma-ALD to our offerings. Plasmas are more reactive, which is useful in its own right, but it can also be used to lower the process temperature,” says Kremers.
In sampling for customers, SALD is learning new things every day. “We have two people sampling for customers day in, day out. We get orders for all sorts of combinations of substrates and nanolayers. This provides us with a wealth of experience, which we take advantage of to help our customers get the most out of our tool,” says Visser. To encourage application development even more, SALD is working with the Plasma Materials Processing Group at Eindhoven University of Technology and is participating in a national spatial ALD research program.
The big question, then, is when customers take the plunge and start to incorporate SALD’s spatial ALD technology in their manufacturing processes. That’s difficult to predict, but Kremers has a hunch which market will be first. “Next-generation thin-film solar cells feature layers that realistically can only be deposited using ALD. Additionally, the solar market is already familiar with this technology. So my guess would be that solar will get the ball rolling, but obviously, I wouldn’t mind to be surprised when another green market proves to move even faster.”
This article was written in close collaboration with SALD.