Post-Etch Clean Loss Budget

# Post-etch Clean Loss Budget

A post-etch clean loss budget checks that residue removal leaves enough of every intended exposed film. It includes the material already lost during etching and any additional loss during strip, clean or other relevant exposure steps. A clean-looking surface is not sufficient evidence that the protected structure survived.

The general post-etch clean entry describes the operation. This guide focuses on the cumulative thickness allowance and the separate evidence needed for residue clearance and material compatibility.

## Identify what must leave and what must remain

Unwanted residue may contain organic material, reaction products or material from the etched stack. Intended oxide, nitride, conductor, liner and landing surfaces may be exposed at the same time. A removal method that works on one residue/material combination is not automatically compatible with another.

The illustration shows residue disappearing while an intended film remains. The two film thicknesses belong to the full arithmetic example below; they do not imply that the clean alone consumes the entire difference. The schematic also does not establish a specific residue composition or removal chemistry.

Distinguish intentional removal of a temporary layer from incidental loss of a layer that must survive. If a process deliberately opens a stop film, its complete removal is a separate integration objective, and the newly exposed landing needs its own allowance.

## Add losses for the same physical film

Assume an intended film starts at 10 nm and must remain at least 7 nm thick at the specified location. Its total allowable thickness loss is 3 nm. In a deliberately simplified example, the etch consumes 1.5 nm, the ash consumes 0.3 nm, and the wet clean consumes 0.7 nm.

Total modeled loss is 1.5 + 0.3 + 0.7 = 2.5 nm. Remaining thickness is 10 - 2.5 = 7.5 nm, leaving 0.5 nm of unallocated allowance above the assumed 7 nm minimum. That does not establish a qualified production margin: variation and measurement uncertainty still need to be addressed.

All three example losses refer to the same protected film and the same location. Do not add an oxide loss measured in one test to a metal loss measured in another as if they were losses of one layer. Build a separate row of the budget for each material and limiting location.

The ash loss is only an assumed contribution for this teaching calculation. Ashing can change surfaces in ways other than simple thinning, and some exposed films are incompatible with a chosen ash process. A negative thickness change, apparent growth or changed optical response requires an appropriate interpretation rather than blindly adding a removal rate.

## Use exposure intervals carefully

For a constant-rate approximation, each thickness loss equals the relevant removal rate multiplied by that film's exposure time in the step. A buried film is not necessarily exposed throughout the entire etch. The clean and its transitions may expose different surfaces for different intervals.

Avoid counting a clean loss twice if a measured pre/post full-sequence change already includes it. Conversely, do not omit rinse, dry or intervening handling effects when they materially change the surface or compatibility result. Chemistry condition, temperature, process history and incoming residues can alter the outcome, so a constant-rate estimate needs validation.

## Check compatibility beyond thickness

Residue clearance, corrosion, roughness, undercut, interfacial damage and changes to electrical behavior can each be limiting. A film may meet its remaining-thickness requirement and still fail its intended function. A local pit can also matter more than an average thickness measurement.

Qualify on the intended filmstack and representative patterned features. Measure residue removal with a suitable method and measure protected-material loss independently. Use electrical results where they are relevant to the device or connection, rather than treating appearance as the sole acceptance test.

## Choose a complete process window

A shorter clean may preserve a film but leave unacceptable residue. A longer exposure may remove residue but exceed a compatibility or loss limit. Seek a window that satisfies both requirements over the defined incoming variation. Changing chemistry or the preceding etch may be necessary when time adjustment alone cannot provide that window.

The useful result is a clean, functional structure with sufficient protected material remaining. Keep residue-removal success and film-survival success as separate checks within the same sequence.

## Sources and scope

• [Entegris — Post Etch Cleaning Solutions](https://www.entegris.com/en/home/products/chemistries/specialty-chemicals/post-etch-cleaning-solutions.html): residue and hard-mask removal applications.
• [Entegris — Evaluation of post-etch residue cleaning solutions for TiN removal](https://www.entegris.com/content/dam/web/resources/poster/poster-post-etch-tin-removal-ucpss-2016.pdf): material compatibility in a specific interconnect integration.
• [Aluminum Post-Etch Residue Removers](https://www.dupont.com/electronics-industrial/aluminum-perr.html): examples of formulations with material-specific compatibility considerations.

Original illustrations and assumed loss values. No chemical mixture, processing time or product is prescribed as a universal clean recipe.