The promised land of future ALD research
Last spring saw the conclusion of the Industrially Functional Surfaces (IFS) project focusing on the experimental and applied research of thin films at JAMK University of Applied Sciences. Scientific thin film research has a long-standing tradition in the Finnish academic field. Internationally-acclaimed thin film research has been carried out at the University of Helsinki, Helsinki University of Technology (Aalto University) and the University of Jyväskylä, just to name a few examples. Among universities of applied sciences, on the other hand, applied thin film research has been somewhat scarce. The development of atomic layer deposition (ALD) technology, in particular, which was even granted the Millennium Prize, has elevated the applicability of thin film technology to an entirely new level. Now is the time for engineers, too, to roll up their sleeves and dive into the realm of thin films to help Finnish industry leverage the fruits of the research on a wider scale.
One goal of the IFS project was to develop coating-related collaboration between JAMK University of Applied Sciences, the University of Jyväskylä and businesses in Central Finland. The project’s yields include the development of a new kind of thin film ecosystem, the ALD CoCampus, that utilises the equipment and leading expertise of the University of Jyväskylä and JAMK University of Applied Sciences.
With the establishment of the new CAMS Centre, JAMK is investing heavily in applied material technology. The three cornerstones of the CAMS Centre’s operations are the fatigue testing of materials, 3D printing and thin film technologies. Projects, training and business focused on thin films will be an essential element of the centre’s operations. The CAMS Centre’s operating model involves business-centric R&D projects that aim to generate new information and business.
Hard work pays off!
ALD is currently the world’s fastest-spreading thin film technology. Unsurprisingly, the developer of the method, Tuomo Suntola, was granted the Millennium Technology Prize worth one million euros in the spring of 2018. The benefits of the ALD method, which include near-flawless, atomically smooth coatings that accurately adhere to the surface contours and the opportunity to control the thickness of the coating down to a single atomic layer, have made it irreplaceable to the electronics industry. ALD is the most important reason for the fact that Moore’s Law is still alive and well as we near the 2020s. According to the law presented by Gordon E. Moore in 1965, the number of transistors on a microchip doubles every two years.
ALD is an excellent example of the power of persistent technological development efforts over a long period of time. While Suntola & Co. patented the method as early as the 1970s, the general consensus in the 1990s was still that it was far too slow and cumbersome to ever be feasible for mass production. In the 2000s, however, the method’s value for microelectronics was finally understood and it broke through in the production of memory chips and microprocessors. Intel was the first to adopt the method in its 45 nm processors in 2007.
Atomic layer deposition – like laying a brick wall
As the name suggests, atomic layer deposition involves generating coatings a single atomic layer at a time, which is not that far from laying bricks to build a wall – only on an atomic scale. The mortar in the process is chemistry. Traditional ALD coatings have been successfully produced at temperatures only slightly above room temperature but, generally speaking, the temperatures required are higher. As such, the method has been dubbed ‘thermal ALD.’ The requisite temperatures depend on the material being coated and the reactants (precursors) used. The same coating material can often be manufactured from a variety of precursors. In this context, we refer to varying ALD processes for the same coating material. The ALD processes are viable within a specific limited temperature range, called the ALD window. Outside this window, the “atomic brick wall” either cannot be built at all or the process falls apart as the sloppy mason throws in multiple bricks at a time.
The thermal ALD equipment function in cyclical pulses (Figure 1). The coating process starts (phase 1) with the ALD reactor being fed with precursor 1, which begins to react with the substrate surface. Once precursor 1 has covered (saturated) the entire surface, the reaction ends, because the precursor does not react with itself and there are no available reactive spots left on the surface for the precursor to start another reaction. Any remains of precursor 1 are flushed out of the chamber with an inert gas, which is typically nitrogen or argon. The purge ends the first half-cycle. In the next phase, precursor 2 is injected into the chamber to react with precursor 1 and form the desired compound. As before, the reaction continues as long as the surface has available reaction spots (i.e. until the compound has covered or saturated the surface). The growth cycle, and the second cycle, ends with the remaining precursor being flushed out of the reactor. The new cycle begins with precursor 1 being injected into the reactor. This coating process is described as self-limiting and surface limited. The coating process is continued until the desired number of cycles/atomic layers on the surface has been reached.
Benefits and limitations of ALD
The most significant benefit of the ALD coating method is a direct result of its surface limited nature. The coating adheres accurately to the contours of the surface being coated. In other words, the coating is very conformal to the underlying substrate. No other coating method comes even close to the level of conformality enabled by ALD. Applying the coating carefully one atomic layer at a time also provides other clear benefits. The films are uniform in their thickness, dense and feature no pinholes whatsoever. The price of this precision is that the traditional thermal coating process is significantly slower than other gas phase methods. In certain applications, on the other hand, ALD coatings only a few dozen atomic layers thick are sufficient... and if more irregular and sharp contours need to be coated with an even film, ALD is just the solution for the job.
High-quality protective layers with ALD
Currently, the most important applications for ALD coatings are insulation layers in microelectronics as well as various barrier layers that protect surfaces against moisture, oxygen and unwanted diffusion. The requirements that the electronics industry places on the quality of the layers are extraordinary to say the least. For example, bendable OLED displays are particularly susceptible to the effects of moisture and need a functional surface layer to protect them. In terms of its protective effect, the quality of the layer must be extremely high: equal to ensuring that no more than a drop of water is spread across a plastic sheet the size of a football pitch each month. As an example, this can be achieved with a glass layer about 50 micrometres thick (the thickness of a human hair is about 100 micrometres). The thickness of the requisite ALD layer, in turn, is roughly one-thousandth of this, 50 nanometres!
Plasma-assisted deposition at lower temperatures and more coating materials
The temperatures required by thermal ALD limit the method’s feasibility for substrates that cannot handle high temperatures. These materials include several plastics, for example. In order to reduce the coating temperatures, energy must be introduced to the surface in some other way than heating up the reactor. This is where the fourth state of matter, energetic ionised gas or plasma, comes in. In plasma-assisted ALD-coating, oxygen, hydrogen or ammonia plasma, or various mixtures of them, are typically used as the precursor for the second half-cycle, depending on the desired reaction.
Plasma assistance increases the applicability of ALD for both substrates and coatings but also introduces a number of limitations. The conformality of plasma-assisted coatings falls short of thermally produced coatings. The plasma systems also make the equipment more complex and hamper the industrial scaling of the ALD processes. In certain ALD processes, the energetic plasma may also damage the growing film.
Spatial ALD boosts coating speeds to a new level
Tuomo Suntola also patented another way of manufacturing ALD films in the 1970s. Whereas traditional ALD coating is a timed sequence of precursor pulses and flushes, spatial ALD entails continuous gas flows and the substrate is moved through distinct flow zones (see figure). When moving through zone 1, the substrate is saturated with precursor 1. After this, the substrate moves to a zone that features a flow of inert barrier gas, typically nitrogen. Following treatment with inert gas, the substrate transitions to the second half-cycle zone where the resulting chemical reaction creates the desired material. The coating process is then continued by moving the substrate back through the barrier gas flow to the precursor 1 zone. The coating process persists until the desired number of cycles/atomic layers has been accumulated onto the surface.
The time-consuming purge phases are not needed in spatial ALD (SALD), which means that the growth rates of the film can be increased to as much as 100 times that of traditional ALD. The limiting factor is the retention time in the various zones to ensure sufficient saturation. The substrate must not leave the reaction zone too quickly. Increasing the substrate’s movement speed also increases the risk of unwanted precursor gases travelling through the barrier zone from one reaction zone to the next. The coating speeds are competitive with other gas phase methods. The method can also be easily scaled to pilot and industrial contexts and roll-to-roll (R2R) manufacturing. SALD doesn’t even require a vacuum chamber! Back in the 1980s, Suntola calculated that spatial ALD would be financially viable, without unreasonably high cost-increasing gas flows, even at normal air pressure. This method is referred to with the acronym AALD (Atmospheric Atomic Layer Deposition). During the ALD research boom of the 2000s, researchers at the Eastman Kodak Company were the first to introduce a functional technical solution for AALD. The method involves bringing the substrate very close to the coating head, to a distance of about 30 micrometres, which is why it is referred to as close proximity SALD.
Since SALD retains the benefits of ALD – unparalleled conformality, dense and smooth coatings, and thickness control at the level of a single atomic layer – it is no surprise that the development of SALD has led to a new surge in ALD research.
Since the technique is a newcomer in the realm of ALD, there are currently a limited number of feasible SALD processes compared to traditional ALD. So far, ten viable oxide processes have been introduced, including aluminium oxide (Al2O3), zinc oxide (ZnO), titanium oxide (TiO2), hafnium oxide (HfO2), zirconium oxide (ZrO2) and niobium oxide (Nb2O3). Doped films, which are essential to the semiconductor industry, can also be produced spatially. Metal processes have been developed for silver (Ag) and platinum (Pt). In total, there are currently some twenty workable SALD processes, which means that there is plenty of room for more.
Suntola’s three initial ALD reactor models utilised spatial technology. That said, the development of SALD did not get into full swing until after the turn of the millennium. The first scientific publications on the topic were released around 2005, and currently dozens are published each year. According to ALD researcher Dr David Muñoz-Rojas (2015) based in Grenoble, research related to SALD is being conducted at 16 laboratories worldwide and there are six commercial equipment manufacturers. In Finland, accomplished development of the method took place in the early 2010s in Mikkeli at the Astral laboratory of Lappeenranta University of Technology, under the leadership of Professor David Cameron.
SALD in Finland
Both Finnish ALD equipment manufacturers, Beneq and Picosun, have put together SALD solutions suitable for research purposes. Beneq was the first to introduce pilot-scale R2R SALD equipment to the market. The first development version WCS 500 (WCS, Web Coating System) was installed in Mikkeli in 2013. Since then, two more recent WCS 600 systems have been sold for research purposes to the UK and Japan. After the discontinuation of the Astral laboratory in Mikkeli, ownership of the WCS 500 system and two other research-oriented Beneq ALD systems – the TFS 200r SALD and the TFS 500 thermal ALD systems – was assumed by the University of Jyväskylä and the equipment was moved to Jyväskylä in a collaborative effort between the University of Jyväskylä and JAMK University of Applied Sciences. In order to foster closer cooperation related to coating research, the university donated the WCS 500 system to the university of applied sciences. In early 2019, the device was moved to the JAMK premises on Rajakatu and it is currently being set up for operation. After the move, the TFS 200r system was once again in coating form in the autumn of 2018 and SALD coatings have been applied in Jyväskylä for about a year now.
Thin film and ALD research at the University of Jyväskylä
The University of Jyväskylä has been engaging in high-quality scientific research on nanotechnology for quite some time. The university’s Nanoscience Centre boasts coating systems utilising a wide variety of gas phase methods, including CVD, electron beam evaporation and pulsed laser deposition (PLD).
In addition, the Pelletron accelerator of Professor Timo Sajavaara’s research team and the Nanoscience Centre’s helium ion microscope are exceptional workhorses in the field of thin film research, even on a global scale. Having taken up a researcher’s position at the University of Jyväskylä, Sajavaara has systematically moulded his accelerator laboratory team into one of the world’s foremost groups focusing on accelerator-based material physics. These efforts were rewarded when Sajavaara was appointed as a professor of the field at the University of Jyväskylä’s Department of Physics in 2015. The Pelletron accelerator and particularly the TOF-ERDA method are extremely well-suited to determining the element composition and depth distribution of thin films. Other important processes that utilise the Pelletron accelerator are RBS (Rutherford Backscattering) and PIXE (Particle Induced X-Ray Emission).
The modern range of equipment at the Physics Department and Nanoscience Centre includes a wide variety of electron microscopes, such as scanning electron microscopes (SEM), transmission electron microscopes (TEM), atomic force microscopes (AFM), X-ray diffraction equipment, an ellipsometer, a 3D profilometer, light microscopes, hyperspectral cameras, X-ray tomography equipment, and so on. The helium ion microscope housed in the Nanoscience Centre’s cleanroom is an especially useful and rare tool, which is uniquely suitable for the analysis of thin film coatings. Among the distinct benefits of the device is that it can be used to image electrically insulating samples without a special pre-treatment phase. In SEM microscopes, for example, insulating samples need to be coated for conductivity to enable imaging.
Sajavaara’s team has conducted ALD research since 2013, at which point the Beneq TFS 200 ALD research tool was acquired for the Nanoscience Centre. One of the key research areas has been plasma-assisted ALD of metal oxides. A doctoral dissertation on the subject was published by Mari Napari, who is currently continuing her accomplished research at the University of Cambridge in England.
ALD CoCampus: coating expertise for businesses from research to pilot projects
One aim of the IFS project was to develop coating-related collaboration between the University of Jyväskylä, JAMK and businesses in the region. This goal was achieved, as the cooperation is off to a good and smooth start. A good example is the transfer of the ALD equipment from Mikkeli to Jyväskylä. Joint efforts are under way to bring the equipment to working order and kick off relevant research and project activities. The parties strive to work as equal partners to leverage their various unique strengths, the university’s strong research expertise and JAMK’s close relationship with the business sector. In the spirit of the EduFutura collaboration between local universities, the parties involved are forging a new kind of ecosystem offering research, development and education services related to thin films. This ecosystem has been dubbed ALD CoCampus. It serves as a one-stop-shop provider of thin film expertise to those in need of it, from companies and research facilities to other parties that may benefit from this knowledge.
ALD CoCampus’ diverse range of equipment enables the development of ALD coating processes from recipes to finished products. The TFS 200 system is well-suited to the testing of new coating materials and various substrates. Thanks to its larger coating chamber, the TFS 500 can be used to test larger pieces. Both the TFS 200 and 500 are also equipped for plasma-assisted ALD, which means that the range basically covers the entire spectrum of ALD processes.
The spatial TFS 200r system can be used to test ALD processes for flexible materials, such as printable electronics, plastic and metal membranes, fabrics, paper and other fibre-based materials. Reference coatings for process tests can be easily manufactured with the TFS 200. Ultimately, the best and most functional spatial processes for flexible materials can be tested on an industrial scale with the WCS 500 R2R coating device.
ALD coatings have experienced a breakthrough in microelectronics. The development of SALD methods is ousting ALD processes from their microelectronics foxholes, making them available to other industrial sectors. JAMK University of Applied Sciences now has the opportunity to ride the emergent ALD wave to the top and profile itself first as the leading Finnish UAS in surface technology, but the international field is also up for the taking. Go ALD CoCampus!
- Esa Alakoski, Project Engineer, PhD, Industrially Functional Surfaces project (ended), CAMS, ALD CoCampus, Jyväskylä