Post Etch Polymer Clean
The post-etch polymer clean exists because of the exact mechanism that made the extreme aspect-ratio etches earlier in this project possible in the first place: the same fluorocarbon sidewall passivation that kept the storage-node hole etch vertical and gave the TSV's Bosch process its scalloped profile leaves behind a polymer residue once the etch finishes, and that residue has to be stripped before anything else can happen to the structure. This is a genuinely different cleaning problem from the pre-clean covered in the previous round — pre-clean removes native oxide that forms passively from ambient air exposure between process steps, while this clean removes a byproduct the immediately preceding etch step actively deposited as part of its own working mechanism, which means post-etch polymer clean is a direct, inevitable consequence of how well an anisotropic etch did its job, not an unrelated contamination problem showing up from outside.
Choosing between oxygen plasma ashing and a wet chemical strip depends entirely on what material the etch just exposed, which is a different selectivity concern from the dry-chemical-versus-sputter tradeoff the pre-clean round already covered. Oxygen plasma efficiently burns off fluorocarbon polymer residue and works well wherever the exposed surface underneath is a dielectric that can tolerate oxidizing conditions without consequence; but if the etch exposed metal — copper especially — that same oxygen plasma readily oxidizes the metal surface, forming an oxide layer that degrades contact resistance exactly where a clean, low-resistance connection is needed, so process flows route metal-exposing etches toward wet amine-based strippers instead, which remove the polymer chemically without the oxidation risk plasma ashing carries on an exposed metal surface.
This clean step is scheduled immediately after specific etches rather than running generically throughout the flow the way the bevel etch does, which ties it much more tightly to individual process stages. Bevel etch runs after many different, otherwise-unrelated deposition steps because edge-film contamination is a generic risk regardless of what film was deposited; post-etch polymer clean instead follows specific anisotropic etches whose own passivation chemistry is known in advance, which lets the clean recipe be tuned precisely against the particular polymer composition that etch's specific chemistry produces, rather than needing to handle an arbitrary, unpredictable residue the way a truly generic cleaning step would.
Read through a byproduct-of-success lens, the post-etch polymer clean is a direct accounting entry for the same chemistry that made this project's hardest etches — the storage-node hole, the TSV through bulk silicon — achievable at all: every bit of sidewall protection those etches borrowed from fluorocarbon passivation has to be paid back afterward in the form of a dedicated clean step, which makes this round less a separate process concern and more the second half of a mechanism this project has already explained once, now settling its account.