Lecture 48 (CHE 323) Lithography Resolution

[@masio81]

Amazing material! Thanks for sharing them. And is crazy how you can be right on so many things, but people will comment on the one you are wrong!

[@chewcharntan9115]

I have just started my Precision Engineering last year and have been learning the skills until I came to your site. These have a very great impact on my learning, I can integrate the skills learn from all directions into one. The publications in IEEE is moderated for me and it has let me see in photolithography. These have benefit me lots.

[@lidarman2]

Here I am in the future regarding the EUV viability. I guess Congrats AMSL. It really shows that we can solve crazy problems if there are enough resources behind the effort.

[@ChrisMack]

PDF copies of all the slides in this course are available at: www.lithoguru.com/scientist/CHE323/course.html

[@thefastestturtle5596]

its crazy watching these now and knowing theres 5nm production now

[@ricardofontane2707]

5 nm is the node label and not the feature size. the feature size is bigger, but never the less very impressive

[@ToSamir]

EUV worked... !! you never know....

[@ashwin2112]

Hello Chris, These are wonderful sets of lectures. Thanks for sharing your knowledge. I have a question about the k1 relation with #-beam imaging. 1) A couple of lectures back, you mentioned that the theoretical limit for k1 with 3-beam is 0.5. Most of the current EUV specs in Wiki say k1 is below 0.5, depending on the feature size. Although it is not mentioned explicitly anywhere, does it mean that 2-beam EUV is mainly used in these machines? 2) Secondly, you mentioned that the Rayleigh DOF definition is only valid for 3-beam interference. Does it mean DOF in these new EUV technologies does not carry a strict meaning.? Thanks again.

[@danielgerchman9780]

With the current feature size around 22nm, what has been done to lower the resolution beyond the 36nm you have mentioned? Really great videos, thank you!

[@ChrisMack]

+Daniel Gerchman Two comments on this question. First, when Intel (or anyone else) says they are running a 22-nm process, this is the name of the process, not a description of the feature sizes used. There are no 22-nm features in a 22-nm process. The smallest half-pitch feature is about 40-nm for that process. That said, the way to go beyond the 36-nm resolution limit of today's 193 immersion tools is to use double (or multiple) patterning. See Lecture 59 in this course series. Chris

[@danielgerchman9780]

+Chris Mack Thank you for the answer! (I am currently in Lecture 53)

[@timgepoint1000]

I was told at a fair by Zeiss that they produce mirrors for a process that uses 13nm UV-light

[@THCynical]

Great Lectures Chris! Any comment on the current state of EUV technology? granted its been many years coming, very close now :)

[@MrAaaanil]

Thank You for your lectures, Sir!! As you said that adding water does not increase NA but it allows an increased NA lens to be designed and built. What "allows" means here? Why can not we make bigger lenses for Low refractive index systems like air?

[@chrismack783]

You can't make the sine of an angle bigger than 1. This limits the numerical aperture to be less than 1 when operating in air.

[@mrmendee]

Great lecture , thank you very much. I wonder half pitch's information is needed to this ppt.

[@nickr3272]

Chris, your videos and paper publications in IEEE have let me start my paper in photolithography. I was really stuck and this gave me a starting point. Thank you very much your knowledge and explanation are well appreciated. Also do you have any suggestions in writing a paper based on photolithography? Any certain subjects i should research? Again, thank you.

[@ChrisMack]

Nick, there are many interesting topics in lithography! Today, the areas that attract the most attention are EUV lithography and multiple patterning with 193 immersion.

[@SaiyedTasnimMdFahim]

What will happen if we immersed the photomask in the liquid?

[@ChrisMack]

Almost all lithography tools use reduction, so that the numerical aperture on the mask side is smaller than on the wafer side. As a result, there is no advantage to using an immersion fluid on the mask side (the low-NA side) of the optical system.

[@RebeliousSapien]

has the wavelength changed from 193 nm now in 2020?

[@ChrisMack]

193 nm light (from ArF laser) is still the dominant wavelength for leading-edge manufacturing of high resolution patterns. However, in 2019 the first chips made with EUV lithography (wavelength = 13.5 nm) were shipped. Only one or a few lithographic levels used EUV (most others used 193), but it was still a major accomplishment. See Lecture 60 (kzbin.info/www/bejne/gnnch5Jjbq-Lq5Y).

[@yasir2nd]

great work, benefited a lot!

[@cinzhe]

You were wrong about EUV!

[@ChrisMack]

I am frequently wrong.

[@paulaao]

We have EUV now :)

[@JG-ol2bz]

better at 1.25 speed