The year 2019 is expected to be a year that has significant meaning in the history of South Korea’s semiconductor industry. This is because it will be the first year that South Korea will apply a brand new technology to start mass-production, beating all competitors. This ‘new technology’ is the EUV(Extreme Ultraviolet) lithography technology.
Many believe that Samsung Electronics’ introduction of the EUV technology has a strategic advantage since it allows them to expand their domains beyond the memory-based business model while strengthening their business in the field of system IC and foundry. However, this situation of them leading the world can also be a burden to Samsung at the same time. It is because that Samsung is now in a situation which they have to solve problems on their own, even when they face new, unreported technical difficulties.
Even right now, Samsung is facing a technical barrier that limits them from mass-production, the absence of EUV pellicles. Pellicles are used as a material to reduce wafer defects caused by mask contamination. Easily put, it is an anti-pollution component that is attached to the front side of the mask. It works by placing a highly permeable thin film with a little distance from the surface of the mask, the film will get contaminated instead of the mask, keeping the mask uncontaminated.
This way, the wafer does not get defected unless the size of the contaminant on the surface of the pellicle is bigger than 10µm. However, EUV pellicles, which require a completely different technology compared to the currently commercialized DUV pellicles, have yet to be commercialized due to technical difficulties.
There are many reasons why EUV pellicles have not been commercialized yet. First, it is due to the special nature of the EUV wavelength. EUV is a unique wavelength that absorbs almost all kinds of substances at a high rate. Not only does it absorbs solids and liquids, but it can even absorb gases at a very high rate. So to create a EUV pellicle in a form of thin film, the thickness of the film must be under 60µm, and materials that can be used are also extremely limited.
Even if a pellicle that has 90% permeability -which is the goal permeability- is created and applied, it will suffer a significant loss in productivity because the EUV has to pass the pellicle twice (because of the reflective mask) and will lose 19% of its light during the process. Currently, the permeability required by the industry is 88%, and there are no products that satisfy this level of permeability.
Second, the characteristics related to ‘heat.’ When EUV is absorbed by the pellicle, the absorbed energy is mostly converted into heat energy, which increases the temperature of the thin film sharply. And due to the structure of the pellicle and other environmental factors, it is not easy to cool it off. There are three principles for cooling: conduction, convection, and radiation; the EUV pellicle is so it has a very small cross-section, which limits its potential to transfer heat through conduction. Also, the EUV lithography equipment maintains a high vacuum state, so cooling through convection is also not easy. However, there is a small area around the mask (pellicle) that allows the flow of hydrogen radicals, so there may be some cooling effect from convection.
Finally, the mechanism of radiation can be used to provide cooling for EUV pellicles. Unfortunately, silicon (Si) materials, which have a high EUV permeability, does not have a high emissivity. The situation is even more troublesome since as the thickness goes down, the emissivity decreases even more. Therefore, when EUV lights are scanned during the lithography process, there will be repeated thermal shock, which means that the material will be heated up to 600-1200℃ and then cool down to room temperature in a short amount of time.
If this kind of thermal shock occurs repeatedly, the film of the pellicle will end up being deformed due to high heat, and it will also be vulnerable to fatigue failure. If this happens, the pellicle will not function properly, and even more, the lithography chamber can be contaminated due to failure, which will lead to a major disaster in the operation of the lithography equipment.
The third is related to the mechanical traits. Enough mechanical strength shall be required so that it can withstand various kinds of impacts such as the impact that occurs when transferring the mask with the pellicle equipped, the pressure difference that occurs on both sides of the pellicle during the pump-down process after it is equipped on the EUV lithography equipment (which is for creating a vacuum environment), and the mechanical impact that is caused by the fast-moving mask stage. However, this kind of mechanical strength is hard to obtain when using an ultra-thin film.
Currently, various attempts are being made to secure EUV pellicle materials that can show the ideal performance. ASML is the leading company in this field. Although ASML is not manufacturing its own pellicles, they are putting more than 140 people to pellicle R&D to lead related infrastructure technologies and set a guideline for pellicle creation.
The company that is actually developing EUV pellicles under the technological direction of ASML is a Canadian company called Teledyne DALSA. This company owns numerous membrane-related technologies while working on the MEMS foundry business. They are currently developing EUV pellicles based on their own polysilicon thin film. With a multi-layer structure that consists of Ru/SiN/Poly-Si/SiN, it currently has a permeability of 83%. However, it is still unclear whether they can fulfill the needs of a better permeability or a pellicle that can perform well on light sources with a higher power.
Belgium’s IMEC is currently developing porous pellicles using CNT (Carbon Nano Tubes), which have higher average permeability and is stable to pressure changes due to the characteristics of a porous structure. But the CNT itself is not stable in a hydrogen plasma environment, which is the usage environment, and thus requires additional layers of protection. However, there has not been an announcement of new results recently, which suggests that there may be a problem that occurs during this process of adding protective layers.
In South Korea, Samsung Electronics is currently researching nano-graphite material film materials for pellicle purposes. This is an attempt to utilize the strength of a special material called graphene, which is a type of carbon material. But even in this case, they are currently facing obstacles on producing it on a large scale.
FST, which sells pellicle products for DUV, has also been carrying out a research and development project to create EUV pellicles based on silicon carbide (SiC). Their product has a higher degree of strength and permeability but they are also facing difficulties in reducing the wrinkles when producing through the existing pellicle creation process (film production - frame bonding)
South Korea’s only blank mask manufacturer, S&S Tech, has also started developing EUV pellicles by using single crystal silicon and announced their product at the EUVL symposium which was held in the U.S at September 2017. Although it still has its limitations (e.g. more technology for improved yield rate, ways to solve heat emission issues), their products have shown an 88% permeability by using single crystal silicon layers and heat emission structures made of Ru, while also satisfying the specifications required for mass production. Currently, only products from S&S Tech has shown this kind of results and is drawing a big amount of interest from around the world
Hanyang University, which has been working on EUV lithography for the past 20 years, is developing a variety of EUV pellicle related technologies that are available for companies and has the infrastructure to evaluate the performance and characteristics of different EUV pellicles. Although they currently have the technology to create a full-size pellicle with 83% permeability by using silicon nitride, and also the technology to create a heat emission layer by using Ru, they are now focusing on the technology to create new materials for pellicles that can be used in high power EUV lithography technologies.
Currently, some researchers are also considering a way to proceed EUV lithography without a pellicle, but this doesn’t mean that pellicles will become unnecessary. The development of EUV pellicles that can satisfy specs which are required for mass production will open up new possibilities and business areas, so it is hoped that Korean companies can get a shot to dominate the global EUV pellicle market in the near future.
ASML is selling their EUV pellicles, which is under development, to institutes for R&D purposes, but it is known that they are asking for a hefty price of 100 million KRW ($89000 USD) per sheet. Although the price will definitely drop when it becomes possible for mass-production, there is a high possibility that the price will remain expensive due to the high technical difficulty. With most of the EUV-related business areas and markets becoming into a monopoly (or oligopoly), the EUV pellicle market is also expected to have a scenario of ‘winner takes it all.’