Etch Chemistry Naming Fluorocarbon Halogen Oxygen
An etch chemistry is named by the element family that does the actual bond-breaking, not by the gas canister label on the fab floor — fluorocarbon chemistry, halogen chemistry, and oxygen chemistry are the three names that organize nearly every plasma etch recipe, and each name predicts which material family the chemistry attacks before you ever read a flow rate. Fluorine-based gases cut oxides and nitrides, chlorine- and bromine-based gases (the halogens other than fluorine) cut silicon and metals, and oxygen chemistry strips carbon-based films like photoresist. The periodic-table family a gas belongs to is, in practice, the same thing as the job title it gets on a process flow.
A few things explain why these three family names carry so much predictive weight:
Fluorocarbon chemistry is named for pairing fluorine with carbon, and that pairing is the whole trick. Bare fluorine radicals etch silicon and its oxide almost equally, which would make selective oxide etching impossible; adding carbon to the gas (as in CF4 or C4F8) lets a carbon-fluorine polymer deposit on exposed silicon while the oxide keeps etching, buying the selectivity that makes it possible to stop cleanly on a silicon substrate after punching through an oxide layer above it. The carbon isn't incidental to the name — it's the reason the chemistry works as an oxide etch at all.
Halogen chemistry is named for the periodic-table column, chlorine and bromine specifically, because fluorine is deliberately excluded from this family's job. Fluorine etches silicon too fast and too spontaneously (even without plasma) to give good profile control, so silicon and metal gate etches favor chlorine (Cl2) or bromine (HBr) instead, which react with silicon or aluminum only under ion bombardment — giving the anisotropic, directional etch a transistor gate needs rather than the isotropic attack fluorine would cause.
Oxygen chemistry is named for a single element because its job is narrow and almost universal: oxidize carbon into something that leaves as a gas. Photoresist and other organic masking layers are built from carbon-hydrogen chains; an oxygen plasma converts that carbon to volatile carbon dioxide and water vapor, which pump away and take the mask with them. Because nearly every inorganic film in a fab (oxide, nitride, metal, silicon) resists oxidation far better than carbon does, oxygen chemistry earns a reputation for being almost perfectly selective to resist — which is exactly why it is the default choice for post-etch ashing regardless of what sits underneath the resist being stripped.
Reading a recipe's chemistry name tells you its place in the flow before you read a single gas flow number. A real multi-step etch almost always chains these families in the same order: fluorocarbon chemistry opens a dielectric layer, halogen chemistry etches the functional silicon or metal underneath, and oxygen chemistry strips the resist mask that protected both steps. Seeing "fluorocarbon" on a recipe name tells you it's an early, oxide-opening step; seeing "halogen" tells you it's cutting the device layer itself; seeing "oxygen" or "ash" tells you the pattern transfer is already done and cleanup has begun.
Read an etch recipe by its chemistry family name rather than by its exact gas mixture, and the recipe's role in the flow is already implied: fluorocarbon chemistry is a dielectric-opening step built around a carbon-fluorine passivation trick, halogen chemistry is a silicon-or-metal cutting step that depends on ion bombardment rather than spontaneous reaction, and oxygen chemistry is a cleanup step that works by converting carbon to gas — three element families, three jobs, and almost the entire etch module of a modern fab built from chaining them in that order.