Photolithography from first principles
Writing with light.
Lithos is Greek for stone. Today the stone is a 300 mm disc of silicon, and the pen is light with a wavelength shorter than a virus is wide.
This page takes the machine apart one idea at a time. Each chapter has a scene you can turn, and most have something to change. Scroll to go on.
The slit of light crossing the wafer is the whole story in miniature. The rest of this page explains why it has to be a slit, why the light is that colour, and why it moves.
01 The principle
A projector running backwards.
A cinema projector throws a small film frame large onto a screen. A lithography scanner does the opposite. It takes a mask, the reticle, and throws its image 4× smaller onto the wafer.
Look at the F. It lands upside down and mirrored, like any image through a lens. The chip designer never sees this. The mask is simply drawn the other way round.
Why shrink at all? A speck of dust or a wobble on the mask shrinks by four as well. The expensive precision sits on a big part, where it is easier to make.
02 What the light actually does
The print is a stencil, not ink.
Light deposits nothing. It changes a thin light-sensitive film called photoresist. Where light lands, positive resist becomes soluble and washes away. What stays behind is a stencil for the real work: etching, implanting, depositing metal.
Then the resist is stripped and the next layer begins. A leading chip goes round this loop several dozen times.
This is why a fab's lithography bays are lit yellow. Resist ignores amber light, so people can work in them without exposing every wafer in the room.
03 Wavelength
Shorter waves draw finer lines.
You cannot draw a line thinner than your pen. For light, the pen is its wavelength. For forty years the industry has walked down the spectrum: mercury-lamp lines, then excimer lasers, then EUV.
The waves are drawn to scale against each other. The grey lines below them are the finest lines each generation printed in production.
04 Diffraction
Where lines blur into grey.
Light passing a fine pattern spreads out. The finer the pattern, the wider the spread. The lens catches only a cone of it, and what it misses is detail lost. Lord Rayleigh put it in one line:
The surface is the light intensity across a row of lines. The flat plane is the resist's threshold. Where a ridge rises through it, a line prints. Shrink the lines and watch the ridges sink.
Simplified model. Scanners light the mask from an angle so the lens catches two diffraction beams instead of three, which is what lets them work near k₁ = 0.25. Below it, the first beam misses the lens and no pattern is left in the light.
05 Numerical aperture
A film of water buys 45%.
NA = n · sin θ. In air n = 1, so NA can never reach 1, and in practice it stops near 0.93. The steepest rays leaving the last lens hit the glass surface and bounce back inside.
Fill the gap with ultrapure water and n becomes 1.44 at 193 nm. Those steep rays now reach the wafer and NA climbs to 1.35.
The catch is mechanical. The wafer races under the lens while the water has to stay put, with no bubbles and no droplets left behind on a surface that will be etched next.
06 The scanner
Step, scan, repeat.
The scanner does not flash the whole field at once. A narrow slit of light stays still. The reticle and the wafer slide through it in opposite directions, the reticle four times faster. One field done, the wafer steps to the next and scans back the other way.
On the right sits a second wafer stage. While one wafer is exposed, the next is measured: its height map and its alignment marks. Each new layer must land on the one below to within a couple of nanometres.
At 1× this runs roughly eight times slower than a production DUV immersion scanner, which finishes up to about 300 wafers an hour.
07 Extreme ultraviolet
Making light that air swallows.
At 13.5 nm no lamp or ordinary laser will do, and air, glass and water all absorb it. So the machine makes its own light, in a vacuum, from tin.
A tin droplet about 25 µm across falls through the focus. A first laser pulse flattens it into a pancake. A second, far stronger pulse turns it into plasma at about 220,000 °C. The plasma glows at 13.5 nm, and a curved mirror, the collector, gathers that glow into one point.
Real rate: 50,000 droplets a second, each hit twice. The collector sits a hand's width from a tin explosion, so hydrogen gas sweeps the debris away from its surface.
08 Optics without glass
Mirrors all the way down.
No lens passes 13.5 nm, so every optic is a mirror, the mask included. Each one is a stack of about 40 pairs of molybdenum and silicon layers, each pair about 7 nm thick, so the tiny reflections add up in step. Even then, the best mirror returns only about 70%.
Smoothness is the other half. ZEISS, which makes these mirrors, compares it to a mirror the size of Germany whose highest bump is 0.1 mm.
09 Beyond one exposure
When one pass is not enough.
Below k₁ = 0.25 physics says no. So before EUV the industry split the pattern: print every other line, etch, then print the rest in the gaps. Two masks, two exposures, and the second must land within a few nanometres of the first.
Self-aligned variants grow spacers on the sides of printed lines to double or quadruple density from one exposure. EUV exists largely to delete those extra steps, and each step it deletes is a deposition, an etch and a chance for error.
10 Putting it together
From disc to atom.
Slide from the wafer down to a single silicon atom. Everything in between is what one scanner pass has to control.
Every chip in your pocket passed through a scanner several dozen times. Each pass: coat, measure, align, scan, develop, etch, strip. The rest is engineering at the edge of what vibration, heat and materials allow.
Next in this series:
- Sculpting in atoms: what etch and deposition do with the stencil
- The impossible machine: why a scanner is fast and exact at once
- Anatomy of a fab: the building the scanner stands in
- Chips on chips: HBM and advanced packaging
- Wrapping the gate: how the transistor went 3D
Textbook physics and published specifications, rounded. Further reading: