Photoresist Development Dissolution Resist Stripping Solvent Developer
Development is the step where the chemical pattern finally becomes a physical one: the wafer is flooded with a developer solution — almost universally 0.26N TMAH (tetramethylammonium hydroxide) for modern positive resists — that dissolves the deprotected (exposed) polymer while leaving the unexposed, still-protected polymer largely intact. After a rinse and spin-dry, the resist standing on the wafer is the first real, three-dimensional copy of the mask pattern in the entire process — everything before this step was light and chemistry; everything after this step is topography that etch can finally see.
Dissolution isn't gradual — it's deliberately engineered to be a near-switch. A good positive resist is formulated so its dissolution rate stays extremely low below a threshold exposure dose and extremely high just above it, which is exactly what the contrast curve shows: a narrow dose window separates "barely dissolves" from "dissolves almost instantly." That steepness — the resist's contrast, γ — directly determines sidewall angle. A low-contrast resist develops gradually through partially-exposed regions, leaving sloped, rounded sidewalls whose width shifts noticeably if dose or focus drifts even slightly. A high-contrast resist snaps from fully intact to fully dissolved over a much narrower dose range, producing near-vertical sidewalls whose critical dimension barely moves even when dose varies — exactly the dose latitude a production process needs to stay in spec across an entire wafer, lot after lot.
TMAH development replaced older solvent-based developers for a specific, practical reason: it's a pure, metal-ion-free base. Early resist processes used organic solvent developers that could leave trace contamination behind; 0.26N aqueous TMAH dissolves deprotected resist cleanly via acid-base chemistry (converting the exposed polymer's carboxylic acid groups to a water-soluble salt) without introducing sodium, potassium, or other mobile ions that would poison transistor gates. That single standardized developer concentration — 0.26 normal — became an industry-wide constant precisely so that resist chemistry, not developer chemistry, is the variable every fab tunes.
This is the last stop before the pattern leaves the "optical" world entirely. Everything from here — etch, implant, deposition, metrology — operates on the physical resist topography that development just created, not on light or latent chemistry anymore. A defect introduced anywhere upstream, from the light source's wavelength stability all the way through PEB diffusion length, shows up here as the final arbiter: either the developed pattern matches the design, or it doesn't.