My main teaching task at the Eindhoven University of Technology takes place in the fall when I teach the MSc course “Physics of Plasmas and Radiation”. This course is mandatory for all our MSc students in Applied Physics that follow the “Plasma & Beams” track (the other tracks within our department are “Nano, Quantum, Photonics” and “Fluid, Bio and Soft Matter”). This track draws typically 50 students per year although we also had a peak of 90 students a few years ago. I teach this course with a colleague and I take responsibility for 2/3rd of the course in which I cover the introduction to plasma physics part. This basically encompasses Chapters 1-6 and Chapter 8 of the book Principles of Plasma Discharges and Materials Processing by Michael A. Lieberman and Allan J. Lichtenberg.
This academic year teaching of this course was special: I did not only teach it to our MSc students but also to employees of ASML. I did this by giving dedicated lectures, so independent of the lectures taken by the students. For a period of 5 weeks, I gave my lectures during a full afternoon to 35 employees of ASML, mostly employees working in “ASML Research” division but also for some people of “Development & Engineering”. Some knowledge about plasma physics is very important for those working on the EUV lithography systems (as I will elaborate about below) but many of the employees have no background in plasmas. The aim of this course for ASML was to bring those people up to speed in this respect.
Photos: Bart van Overbeeke
Also for me it was an interesting experience to give this course to ASML employees even though I’m rather used to teaching to professionals as a founder of the ALD Academy . I got many more questions than I normally get from students and the questions were also much more diverse: from very basic general physics questions (certainly not always easy to answer!) to very detailed questions which very much challenged my broader plasma physics knowledge. The ASML employees were also very much engaged during the lectures and everyone kept coming back without complaining about the duration of the lectures which took up to 4 hours every week. Overall, it was really fun to do! Luckily enough the feedback that I received was very positive as is also clear from the outcome of the questionnaire that was filled in after the course (see below).
But why is it so important for the ASML employees to have knowledge about plasma physics?
The first answer that might come to mind is that the EUV radiation is generated by a laser-produced plasma (LPP). The radiation at 13.5 nm is created by focusing a high-power infrared laser onto microdroplets of Sn. First the spherical tin droplet is hit by a laser prepulse. A plasma is generated and the droplet is propelled and reshaped into an extended disk-shaped target. This target is subsequently hit by the main pulse laser irradiation creating a highly-ionized tin plasma that emits the EUV radiation. This EUV radiation is collected by the collection mirror and the 13.5 nm radiation is focused into the scanner optics region of the EUV lithography system. You can read more about the physics of laser-driven tin plasma sources for EUV radiation in the review published by Oscar Verslato of ARC NL in Plasma Sources Science and Technology: Physics of laser-driven tin plasma sources of EUV radiation for nanolithography:
Well, this would be one reason, but it is not the most important reason for ASML employees working in Veldhoven because ASML San Diego is the epicenter for EUV light source development. The main reason is that the EUV radiation in the scanner optics region of the EUV system also induces a plasma by the method of photoionization. The scanner optics region is not kept at very low vacuum but it is flushed by hydrogen gas at a background pressure of approximately 5 Pa. The hydrogen gas is ionized by the 92.4 eV photons of the EUV radiation, mostly through the single-photoionization process (ionization potential of H2 is 15.8 eV). Subsequently, the ejected electrons carry a high excess energy (76.4 eV) which leads to additional electron-impact ionization further contributing to the plasma generation. Moveover secondary electrons can be generated by the photo-electric effect. Overall, the plasma dynamics is pretty complex. More information can be found in the review papers first-authored by my TU/e colleague Job Beckers (EUV-Induced Plasma: A Peculiar Phenomenon of a Modern Lithographic Technology) and Mark van de Kerkhof (EUV-induced hydrogen plasma and particle release) who works at ASML and who is part-time associated at the TU/e. I also used these resources in my teaching.
The hydrogen gas in the scanner optics region and particularly also the H2 plasma created by the EUV radiation, serves several important purposes in the EUV system. First of all, the use of a small background of H2 contributes to maintaining purity and quality of the vacuum environment (better than just trying to work at base vacuum) while H2 has a high EUV transmission. Moreover, H2 has anti-oxidation and cleaning properties. It especially removes carbon contamination (as present in every vacuum system but also originating from e.g. the photoresist) from the optics. This also holds for potential Sn contamination that might diffuse in the system from the EUV source region. Contaminants, even in small quantities, can significantly absorb EUV light and degrade the performance of the optics. Hydrogen effectively reacts with contaminant species converting them into gaseous forms that are easily removed from the system. Furthermore, the high-energy photons in EUV lithography can cause the optics to oxidize, which would degrade their performance over time. Hydrogen acts as a reducing agent that helps to minimize or prevent this oxidation. Altogether, hydrogen extends the lifetime of the mirrors used in the EUV system. This is crucial because the mirrors are extremely expensive and difficult to manufacture and replace.
The EUV-induced H2 plasma in the optics region is, however, not only beneficial and its interaction with the surfaces in the scanner needs to be carefully managed to avoid serious adverse effects: undesired etching of materials, surface roughening, diffusion into materials making them brittle, etc. This means that the plasma load on the materials close to the EUV beam should be taken into account by proper understanding of the effects and by controlling them. In terms of elementary reactions, this includes the surface interaction of hydrogen atoms and hydrogen ions as well as all kind of synergistic effects (between the various hydrogen plasma species and between the plasma species and EUV photons etc). These topics are addressed by an extensive team of researchers within ASML research and obviously, this requires extensive and specialized knowledge of different fields of science: plasma science, materials science and surface science. With my course on plasma science, I have contributed to extending this knowledge to some extent.
With well over 200 EUV systems shipped, one could wonder why ASML is still putting increasing effort in understanding EUV-induced H2 plasma in their EUV lithography systems. Obviously, this is for a part related to enhancing the lifetime of the system components and hence to increasing the uptime of the systems etc. However, very importantly, also the power level of the EUV radiation at the wafer position needs to be increased. This is particularly important for the new high-NA systems: the next generation EUV systems will have a numerical aperture of 0.55 instead of 0.33. These high-NA EUV systems require higher power levels of EUV radiation to overcome various challenges related to optical design, absorption, efficiency, depth of focus, stochastic effects, and photoresist sensitivity. Only then sufficiently high throughputs and acceptable costs for chip manufacturing can be reached. Note that by the end of 2023, ASML has shipped their first high-NA system (the Twinscan EXE:5000 EUV scanner) to Intel
To conclude this blog with a nice figure, the illustration below shows the famous ne – Te typically used to illustrate that plasmas occurring in nature or prepared in the laboratory span several orders of magnitude in electron temperature and electron density. The data points for the plasmas in the EUV lithography system have been included too. As can be seen, a fairly large range of electron densities and electron temperatures are covered in this industrially very relevant tool!
I would like to thank Dr. Richard Engeln, Dr. Ruud van der Horst, and Dr. Seth Brussaard of ASML for taking the initiative for this course and their support.