Importance of photolithography in IC

Photolithography is the key step that defines the geometry of every layer of an IC. Without it no pattern could be put on the wafer, so it decides the size, density and cost of the chip.

Importance

• Pattern transfer: it copies the layout of each mask (wells, diffusions, gates, contacts, metal) onto the wafer accurately.
• Minimum feature size: the resolution of the lithography (R = k1·λ/NA) sets the smallest transistor, which sets how many transistors fit on a chip (Moore's law).
• Selective processing: the patterned resist protects some regions while others are etched, doped (diffusion or ion implantation) or coated.
• Layer alignment: an IC uses 20-40 mask layers that must overlay precisely, so lithography accuracy determines yield.
• Cost and yield: it takes about 30-35 % of the total wafer cost and is repeated for each layer; defects here directly reduce yield.
• Mass production: one mask exposes thousands of identical dies at once.

Example: making an NMOS source and drain

1. Grow SiO2 on a p-type wafer and coat it with resist.
2. Expose through a mask that has openings at the source and drain areas and develop.
3. Etch the oxide through these windows, down to silicon.
4. Strip the resist and diffuse or implant n-type dopant (phosphorus/arsenic); the oxide blocks dopant elsewhere. This forms the n+ source and drain.
5. A later photolithography step patterns the polysilicon gate, aligned between source and drain, then another patterns contacts and metal.

Thus each device region exists only because photolithography defined it.