Author Topic: EUV light source - didn't know that :)  (Read 2953 times)

0 Members and 1 Guest are viewing this topic.

Offline daqqTopic starter

  • Super Contributor
  • ***
  • Posts: 2321
  • Country: sk
    • My site
EUV light source - didn't know that :)
« on: November 09, 2020, 09:18:16 pm »
So, apparently a normal light bulb ain't gonna cut it, so to get the silly wavelengths required for today's EUV processes ( ~13.5 nm ) you need something special.

What they do is:

Shoot a microscopic droplet of tin (~30 um) into a vacuum chamber.
Hit the droplet at precisely the right time with a CO2 laser.
The droplet becomes crazy hot, forms a dense plasma.
Plasma blinks at the required wavelength.
Light is gathered by insanely polished special mirror.
Light travels through the rest of the machine.
Rinse and repeat thousands of times per second.



Just wanted to share, pretty interesting stuff :)
Believe it or not, pointy haired people do exist!
+++Divide By Cucumber Error. Please Reinstall Universe And Reboot +++
 
The following users thanked this post: tooki, DuPe

Offline jogri

  • Frequent Contributor
  • **
  • Posts: 405
  • Country: de
Re: EUV light source - didn't know that :)
« Reply #1 on: November 09, 2020, 10:02:36 pm »
I wouldn't want to be the guy responsible for maintaining this thing... A mirror that has to be polished enough to reflect 13 nm light inside what is basically a sputtering system sounds like a fun combo.
 

Offline Cerebus

  • Super Contributor
  • ***
  • Posts: 10576
  • Country: gb
Re: EUV light source - didn't know that :)
« Reply #2 on: November 09, 2020, 10:42:39 pm »
I wouldn't want to be the guy responsible for maintaining this thing... A mirror that has to be polished enough to reflect 13 nm light inside what is basically a sputtering system sounds like a fun combo.

That was my immediate first thought: "Hold on, where does all that tin go? How do they keep it off the mirror?"
Anybody got a syringe I can use to squeeze the magic smoke back into this?
 

Offline S. Petrukhin

  • Super Contributor
  • ***
  • Posts: 1273
  • Country: ru
Re: EUV light source - didn't know that :)
« Reply #3 on: November 10, 2020, 12:25:20 am »
I wouldn't want to be the guy responsible for maintaining this thing... A mirror that has to be polished enough to reflect 13 nm light inside what is basically a sputtering system sounds like a fun combo.

That was my immediate first thought: "Hold on, where does all that tin go? How do they keep it off the mirror?"

Maybe they blow it off with a smoke catcher?  :)
And sorry for my English.
 

Offline daqqTopic starter

  • Super Contributor
  • ***
  • Posts: 2321
  • Country: sk
    • My site
Re: EUV light source - didn't know that :)
« Reply #4 on: November 10, 2020, 07:16:51 am »
I'm pretty sure that the tin removal is one of those things that are the secret sauce of the manufacturer :) That said, if you shoot the tin droplet from behind, won't it splatter in the forward direction?
Believe it or not, pointy haired people do exist!
+++Divide By Cucumber Error. Please Reinstall Universe And Reboot +++
 

Offline filssavi

  • Frequent Contributor
  • **
  • Posts: 433
Re: EUV light source - didn't know that :)
« Reply #5 on: November 10, 2020, 07:29:36 am »
The main reason EUV took so long to get to market (but by far not the only one) is precisely the light source, for many many years there was no light source strong enough to make commercial sense, and even today EUV is being deployed only in the most critical steps of the process, as EUV have a fraction of the throughput with respect to normal DUv machines, mainly due to light source issues
 

Offline Berni

  • Super Contributor
  • ***
  • Posts: 5050
  • Country: si
Re: EUV light source - didn't know that :)
« Reply #6 on: November 10, 2020, 07:39:45 am »
And it doesn't even end there. You need optics to handle the light.

Only issue is that we don't have a lens material that works with these wavelengths so all the optics have to be in the form of polished reflective surfaces. This includes optics all the way down to focusing it onto the mask.
 

Online NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9319
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: EUV light source - didn't know that :)
« Reply #7 on: November 10, 2020, 12:43:33 pm »
What's stopping them going to Xrays which are even shorter wavelength but relatively easy to generate?
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline Cerebus

  • Super Contributor
  • ***
  • Posts: 10576
  • Country: gb
Re: EUV light source - didn't know that :)
« Reply #8 on: November 10, 2020, 12:56:52 pm »
Argh! I'm being a putz. As far as keeping the mirror clean while ablating tin droplets is concerned it's obvious what's going on once you think about it.

The droplets have momentum, so dealing with residual unablated tin is easy, just let it hit a collector on the other side of the chamber. That bit was never too much of a problem as I saw it.

As for the stuff that gets really hot and turns into a plasma? Well, that's it, it's not just metal vapour in a vacuum, it's a plasma, it has charge. You can just push it where you want with a electric field.

I'm sure that there's a residual bit of tin floating about that eventually reduces the efficiency of the mirror but now I can see how the majority of the tin can be dealt with, rather than the machine being one large chamber for vacuum metallizing everything in it.
« Last Edit: November 10, 2020, 12:58:47 pm by Cerebus »
Anybody got a syringe I can use to squeeze the magic smoke back into this?
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: EUV light source - didn't know that :)
« Reply #9 on: November 10, 2020, 01:38:36 pm »
Hmm, that could perhaps do it.  The ions are certainly not negative (at least, not many of them?), the energy levels are several times the electron affinity of anything on the Periodic Table.  And with the energy levels being around 90eV as mentioned, it should really only take a few hundred to do that, or even thousands of volts, just because.  Grids could even be hanging to enforce specific electric fields in the drift area, without sacrificing the optical path too much I think.

What's stopping them going to Xrays which are even shorter wavelength but relatively easy to generate?

X-rays tend to cause a lot of damage, both to the resist and the substrate.  The beam resolution can be very good indeed (a few atoms, even), though the masks can't be magnified any, they have to be made at 1:1 scale (well, unless you want to take up many meters of beam length with grazing-incidence mirrors, which, eh, it could happen?).  But the exposure resolution is limited by the movement of ionized (Auger) electrons in the substrate and resin, some 10-20nm depending.  Electron beam writing, same idea, plus the added restriction that you can't really create a mask at all, so you have to scan vector or raster -- very slow indeed!

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline Berni

  • Super Contributor
  • ***
  • Posts: 5050
  • Country: si
Re: EUV light source - didn't know that :)
« Reply #10 on: November 10, 2020, 03:06:18 pm »
What's stopping them going to Xrays which are even shorter wavelength but relatively easy to generate?

I think the problem with those is that they start ignoring the reflective optics and going straight trough it, even if you can get them focused down into a mask then they might go trough the mask itself. Those Xrays are notorisuly hard to wrangle.

And yes also possible that large quantieis of Xrays might mess up the silicon chip that is being manufactured, not sure how much energy is needed for that. Maybe they instead move towards shooting electrons at it with an electron gun, tho those can also get destructive once you give them enugh energy.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: EUV light source - didn't know that :)
« Reply #11 on: November 10, 2020, 04:09:31 pm »
I wonder why they could not use a synchrotron source.  Are they too large to be economical?
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7043
  • Country: nl
Re: EUV light source - didn't know that :)
« Reply #12 on: November 10, 2020, 07:10:16 pm »
Apparently TSMC is the only company which can make it work with decent yields.

The debris from the tin ball screws everything up, I wonder how they fix it ... electrostatic filter perhaps?
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7043
  • Country: nl
Re: EUV light source - didn't know that :)
« Reply #13 on: November 10, 2020, 07:13:01 pm »
I wonder why they could not use a synchrotron source.  Are they too large to be economical?

AFAIK the industry refused to invest significantly invest in researching Free Electron Lasers mainly because of space requirements. Of course given the delays and yield effects from the tin light sources it might not have been the smartest decision in retrospect.
 

Offline Twoflower

  • Frequent Contributor
  • **
  • Posts: 742
  • Country: de
Re: EUV light source - didn't know that :)
« Reply #14 on: November 10, 2020, 07:56:57 pm »
This is provides a bit more information about the EUV process:
https://www.youtube-nocookie.com/embed/f0gMdGrVteI?rel=0

About the EUV source: At about 6:00 there is some more information. They actually hit the tin droplet twice. Even is a bit marketing-ish with photoshop-fail like video sequences it still contain some impressive information.
 

Offline filssavi

  • Frequent Contributor
  • **
  • Posts: 433
Re: EUV light source - didn't know that :)
« Reply #15 on: November 10, 2020, 10:18:08 pm »
What's stopping them going to Xrays which are even shorter wavelength but relatively easy to generate?

Designing the optical train and masks for EUV is already extremely difficult ( you can’t use lenses as there is no material transmissive enough to make them with and you need to use only mirrors which is much harder). Doing it for x rays will be nigh on impossible

also x rays have this pesky ability to pass straight through basically almost any material known to man, thus you will need to use mask sets made out of lead(which does not seem he ideal material).
 

Offline filssavi

  • Frequent Contributor
  • **
  • Posts: 433
Re: EUV light source - didn't know that :)
« Reply #16 on: November 10, 2020, 10:30:54 pm »
I wonder why they could not use a synchrotron source.  Are they too large to be economical?

I don’t know for shure, but I suspect the sheer amount of money needed to be sunk into such a project are probably a complete showstopper, not only in term if of building the machine itself, but also into the research of how to run it as a factory while turning a profit

As far as I know most, if not all currently operating synchrotrons are research type facilities, and are operated as such (even if they sell beam time), with running periods alternated with periodic downtime for maintenance that can last for month, this is really incompatible with the requirement to get an ROI in term of few months/years, needed in the foundry business
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: EUV light source - didn't know that :)
« Reply #17 on: November 11, 2020, 07:59:44 am »
Designing the optical train and masks for EUV is already extremely difficult ( you can’t use lenses as there is no material transmissive enough to make them with and you need to use only mirrors which is much harder). Doing it for x rays will be nigh on impossible

also x rays have this pesky ability to pass straight through basically almost any material known to man, thus you will need to use mask sets made out of lead(which does not seem he ideal material).

This is all solved problems, if perhaps not on the scale of semiconductor production.  Any heavy element will do, tungsten, gold, etc.  I suppose gold and platinum would be particularly desirable for their corrosion resistance.  The mask is made of a low-Z, rigid material, like silicon carbide.  I suppose beryllium would be wonderful, but too difficult to handle (it is still used I think, for certain applications like linear accelerator windows, which are used for food sterilization, medical treatment, etc.).

X-ray optics use grazing incidence and achieve good efficiency.  Chandra X-ray Observatory perhaps most famously, but they see use in synchrotron labs and so on as well.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline iteratee

  • Contributor
  • Posts: 41
  • Country: us
  • Computer Engineer
Re: EUV light source - didn't know that :)
« Reply #18 on: November 15, 2020, 03:08:16 am »
X-ray optics use grazing incidence and achieve good efficiency.  Chandra X-ray Observatory perhaps most famously, but they see use in synchrotron labs and so on as well.
I thought they were using the Chandra-like mirrors for EUV. This is the first I've heard of the conventional reflective mirror use. They need dense neutronium mirrors for better smoothness, or a chemical phosphorescence-like effect for dropping the xray energies down to to EUV levels. :)
« Last Edit: November 15, 2020, 03:09:47 am by iteratee »
 

Offline dmills

  • Super Contributor
  • ***
  • Posts: 2093
  • Country: gb
Re: EUV light source - didn't know that :)
« Reply #19 on: November 15, 2020, 02:08:35 pm »
Xray down to EUV is one HELL of a Stokes shift!


 

Offline Doctorandus_P

  • Super Contributor
  • ***
  • Posts: 4004
  • Country: nl
Re: EUV light source - didn't know that :)
« Reply #20 on: November 22, 2020, 07:46:20 pm »
It's completely insane technology.

Some time ago I heared that each mirror has only about 30% or so refectivity, and there are over 10 mirrors in the system, and after that you still have to have enough light left to develop some kind of photomaterial in ms.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf