Can’t use X rays, they don’t focus easily and blow through the wafers. The whole technology of semiconductors is hitting a wall anyway at <4nm because too dense and electrons will jump the transistors. This scale on nano fabrication is incredible and very cheap for what it is, but we are hitting a limit.
Photolithography is basically shining some kind of electromagnetic radiation through a stencil so that specific lines are etched into the top “photoresist” layer of a silicon wafer. The radiation causes a chemical change wherever a photon hits, so that stencil blocks the photons in a particular pattern.
Photons are subject to interference from other photons (and even itself) based on wavelength, so smaller wavelengths (which are higher energy) can fit into smaller and finer feature size, which ultimately means smaller transistors where more can fit in any given area of silicon.
But once the energy gets too high, as with X-ray photons, there’s a secondary effect that ruins things. The photons have too much leftover energy even after hitting the photoresist to be etched, and it causes excited electrons to cause their own radiation where high energy photons start bouncing around underneath, and then the resulting boundaries between the photoresist that has been exposed to radiation and the stuff that hasn’t becomes blurry and fuzzy, which wrecks the fine detail.
So much of the 20 years leading up to commercialized EUV machines has been about finding the perfect wavelength optimized for feature size, between wavelengths small enough to make really fine details and energy levels low enough not to cause secondary reactions.
We’re getting close! EUV is currently ~13nm, and soft xrays start at ~10nm (but go all the way down to ~0.01nm for hard xrays.)
Sadly there a lots of challenges in transitioning to smaller wavelengths. For example, to get the EUV light in the existing process, we’re already resorting to, essentially, exploding tiny droplets of liquid tin using lasers.
If you were wondering what the hell EUV (Extreme Ultra-Violet) stood for:
https://waferscope.com/duv-vs-euv-whats-the-real-difference-in-chipmaking/
This table from the link sums it up pretty well:
Using shorter and shorter wavelengths of light to etch chips with a higher density of transistors.
Less material for greater performance, for those that want more simplicity.
I wonder what the next generation will be called…EUV-2TM (EUV-2 THE MAX) or SEUV (Super EUV) perhaps?
I assume they will move from UV to X-rays.
Can’t use X rays, they don’t focus easily and blow through the wafers. The whole technology of semiconductors is hitting a wall anyway at <4nm because too dense and electrons will jump the transistors. This scale on nano fabrication is incredible and very cheap for what it is, but we are hitting a limit.
No, X-rays are too energetic.
Photolithography is basically shining some kind of electromagnetic radiation through a stencil so that specific lines are etched into the top “photoresist” layer of a silicon wafer. The radiation causes a chemical change wherever a photon hits, so that stencil blocks the photons in a particular pattern.
Photons are subject to interference from other photons (and even itself) based on wavelength, so smaller wavelengths (which are higher energy) can fit into smaller and finer feature size, which ultimately means smaller transistors where more can fit in any given area of silicon.
But once the energy gets too high, as with X-ray photons, there’s a secondary effect that ruins things. The photons have too much leftover energy even after hitting the photoresist to be etched, and it causes excited electrons to cause their own radiation where high energy photons start bouncing around underneath, and then the resulting boundaries between the photoresist that has been exposed to radiation and the stuff that hasn’t becomes blurry and fuzzy, which wrecks the fine detail.
So much of the 20 years leading up to commercialized EUV machines has been about finding the perfect wavelength optimized for feature size, between wavelengths small enough to make really fine details and energy levels low enough not to cause secondary reactions.
We’re getting close! EUV is currently ~13nm, and soft xrays start at ~10nm (but go all the way down to ~0.01nm for hard xrays.)
Sadly there a lots of challenges in transitioning to smaller wavelengths. For example, to get the EUV light in the existing process, we’re already resorting to, essentially, exploding tiny droplets of liquid tin using lasers.
XXXUV
With added sexual excitement to keep things interesting?