Etch Pattern Collapse Budget

# Etch Pattern Collapse Budget

An etch pattern collapse budget defines the structural condition and clearances that must survive the etch-related process sequence. It includes the relevant strip, clean, rinse, drying and handling stages rather than assuming that the etched shape remains unchanged afterward.

Here budget means a set of geometric and integrity constraints supported by evidence. It is not a universal aspect-ratio threshold or a single force limit that applies to every material and structure.

## Separate the observed condition from its cause

Slender neighboring structures can deflect, touch or remain attached. Those are distinct observations. A final image of touching features does not by itself establish when contact occurred, whether the features fractured, or which force dominated.

Primary studies on silicon nanopillars show that liquid-related forces during drying and adhesion after contact can both matter. Their relative importance depends on the geometry, surfaces and process. Do not assume that every collapse observed after etch was created by the plasma step itself.

## Retain the complete process checkpoint

Record the structural state after the relevant operations. If inspection occurs only after cleaning and drying, the observation describes that combined interval until evidence isolates a narrower source.

A comparison before and after a suspected step needs suitable measurement methods and representative coverage. Inspection and sample preparation can themselves disturb fragile structures, so preserve those limitations in the interpretation.

## Track local clearance explicitly

The teaching example begins with a 20 nm gap. At the same height and cross-section, an 8 nm inward displacement on the left and a 7 nm displacement on the right leave 5 nm of clearance.

This subtraction is geometric accounting, not a prediction of mechanical stability. It assumes defined facing boundaries and corresponding measurements. A positive gap at one location does not rule out contact elsewhere or at an earlier point in the process.

If deflection varies with height, position or time, evaluate the region and interval governed by the requirement. Two independently reported maxima do not necessarily occur together; state whether combining them is a justified bound.

## Connect geometry with mechanical evidence

Height, width, shape, support conditions, material properties and stress can affect structural response. The initial gap and the surface interaction also matter. A single height-to-width ratio does not establish a universal collapse threshold.

Mechanical models require explicit boundary conditions and force assumptions. A simplified beam estimate may help investigate a trend but does not automatically represent a complex multilayer feature, a full array or an adhesion event.

Contact is also different from permanent adhesion. A structure may recover after a transient load or remain attached, depending on the relevant restoring and surface forces. Verify the final state and any applicable functional requirements rather than inferring recovery from nominal geometry.

## Qualify the clean and drying sequence

A process that clears residue can still change the structural condition. Evaluate cleaning effectiveness, material loss and collapse evidence together.

Published studies explore approaches such as modifying surface interactions or changing drying conditions in specific test structures. Such results do not make a treatment universally compatible with device films or downstream processing. Any change can introduce its own surface, contamination, dimensional or throughput constraints.

A lower-surface-tension liquid alone is not a universal guarantee against collapse. Use evidence for the complete stack and sequence rather than substituting one liquid property for qualification.

## Define a measurable acceptance rule

Specify whether acceptance governs minimum gap, maximum deflection, touching or clustered features, fracture, a sampled defect fraction or another defined outcome. Retain the inspected population and denominator for counted failures.

Distinguish no observed collapse in a sample from proof that every feature is stable. Include relevant pitches, dimensions, wafer locations and process histories, and document measurement uncertainty and detection capability.

Do not average away a local contact if the requirement prohibits any contact in the inspected region. Conversely, do not call every small displacement a collapsed structure without a stated definition.

## Close the budget with traceable evidence

Retain incoming geometry, etch recipe revision, material and support conditions, subsequent process steps and the measurement checkpoint. Compare proposed improvements against the full set of dimensional, chemical and functional requirements.

A collapse budget follows the structure through the process that it must survive. Geometry identifies vulnerable clearances; scoped measurements and mechanical evidence establish the accepted condition.

## Sources and scope

• [Ghosh et al. — Preventing the Capillary-Induced Collapse of Vertical Nanostructures](https://ris.utwente.nl/ws/files/294347934/Ghosh_2022_Preventing_the_capillary_induced_co.pdf): primary study of drying, adhesion and surface modification in silicon nanopillar arrays.
• [Ultra-Narrow Metallic Nano-Trenches Realized by Wet Etching and Critical Point Drying](https://pmc.ncbi.nlm.nih.gov/articles/PMC8003182/): primary study of a particular wet-etch and drying integration.

Original sketches and gap arithmetic. No universal collapse limit, mechanical model validation or qualified clean/dry recipe is implied.