Etch Taper Transfer Budget

# Etch Taper Transfer Budget
An etch taper transfer budget defines how much of a feature's sidewall slope comes from the mask rather than the plasma, and the evidence needed to separate the two before changing a recipe. It treats the mask as an active part of the profile, because its incoming slope, its erosion and its faceting during the etch are copied into the material below.
This budget does not restate the symmetric straight-wall angle calculation, which is covered separately. It concentrates on attribution: deciding whether an observed taper originates in the mask or in the etch, since the corrective action differs.
## See the mask as part of the wall

A mask edge is rarely a perfect vertical step. If it enters the etch already sloped, or if it erodes and facets while the etch proceeds, the opening it presents to the plasma changes with time, and the film beneath inherits part of that changing shape.
The sketch shows a sloped mask edge and the tapered feature below it. It is not a profile simulation and assigns no angle to any real process. The actual slope depends on the mask material, the chemistry and the film, and must be established for the specific stack.
A plasma etch can also set the slope directly. Reactive-ion etching produces strongly directional profiles because ions arrive mostly vertically, in contrast to the rounded profiles of isotropic wet etching. Both a directional plasma and the mask act on the same wall, so the observed taper is usually a combination.
## Separate the candidate origins
| Origin | What to look for | What the feature image alone does not show |
|---|---|---|
| Incoming mask taper | Mask sidewall slope before the main etch | Whether it changed during the etch |
| Mask erosion and faceting | Mask height and edge loss after the etch | How fast the edge moved relative to the etch |
| Directional plasma profile | Ion-driven slope on an ideal vertical mask | The mask's separate contribution |
| Sidewall passivation balance | Deposit distribution along the wall | The exact deposition-to-etch ratio locally |
More than one origin usually acts at once. Assigning a dominant one needs mask measurements before and after the etch, not a single cross-section of the finished feature.
## Define what the budget measures

Report top CD, bottom CD and depth together, with the angle convention stated, at a named checkpoint: after etch, after strip or after clean. A mask-induced taper and a plasma-induced taper can read as the same single angle while responding differently to a recipe change.
Pair the feature measurement with a mask measurement. Incoming mask taper and post-etch mask erosion are the evidence that tells whether a recipe adjustment should target the mask step or the etch step.
Sample representative pitches and mask types. A slope that is acceptable on an isolated line can differ in a dense array, and mask faceting depends on the mask material and thickness as well as the etch.
## Act on the cause, not the symptom
If the taper is mask-origin, changing plasma parameters may not fix it and can disturb other results. A controlled slope is sometimes the goal: a reported process adjusted trench slope across a narrow range near vertical while keeping high selectivity, which shows slope can be a tuned output rather than a defect.
The tolerable taper therefore comes from the next step. A liner or fill may need a specific slope for coverage; a contact may need a usable bottom CD. There is no universal angle that suits every integration, and no single knob removes taper without side effects on rate, selectivity, mask survival and profile.
## Close the budget with traceable evidence
Specify top CD, bottom CD, depth, the angle convention, the mask measurements and the checkpoint. Retain mask material, recipe revision and chamber history so a later attribution can be compared with the earlier one.
A taper transfer budget assigns a feature's slope to the mask or to the plasma before anyone tunes a recipe. Measurement, attribution and acceptance are separate questions, and each needs its own evidence.
## Sources and scope
- [Wikipedia — Reactive-ion etching](https://en.wikipedia.org/wiki/Reactive-ion_etching): notes that RIE produces anisotropic profiles from mostly vertical ion delivery, unlike isotropic wet etching.
- [AVS 2000 — Undercut and Bowing Characterisation in High Aspect Ratio Trenches (Boufnichel et al.)](http://www2.avs.org/symposium2000/ProgramBooks/ProgramBook_Session_PS2-WeA.pdf): abstract reporting a cryogenic SF6/O2 process whose trench slope is adjustable near vertical at high selectivity.
Original diagrams and explanatory synthesis. No universal taper limit, origin attribution or qualified recipe is implied.