Etch Scallop Budget

# Etch Scallop Budget
An etch scallop budget defines how large and how uniform the periodic sidewall ripples from a time-multiplexed etch may be, and the evidence needed to measure them and judge their effect. It applies to the crest-and-trough texture that a cycled deposition-and-etch process leaves along a deep feature wall.
This budget does not redefine the scalloping phenomenon or the Bosch process, which are covered separately, and it is narrower than the general sidewall-roughness budget because the texture here is periodic and tied to the process cycle. It concentrates on bounding scallop size and its distribution, and on attributing a consequence.
## See one cycle as one scallop

Etch Scallop Budgettime-multiplexed cycling can leave periodic ripples along a deep sidewallSiliconSiliconOpen trenchscallop depthpitchOne cycle leaves one scallop; its depth and pitch are the quantities to bound.Schematic section · not to scale · scallop size and uniformity depend on the actual recipe and depth.

In a time-multiplexed deep etch, each cycle alternates a passivation step with an etch step. The lateral etch during each cycle produces one scallop, so the wall carries a periodic profile whose pitch follows the cycle count and whose depth follows how far each cycle etches sideways.
The sketch shows a trench with scalloped walls and marks the two primary quantities: the peak-to-valley scallop depth and the pitch between scallops. It is not to scale and assigns no size to any real process; the actual values depend on the recipe and vary with depth.
A reported process aimed at a scallop depth target below ten nanometres, which shows scallop depth is treated as a specified, measurable quantity rather than an unavoidable constant.
## Define size and distribution separately
A single number is not enough. Report the peak-to-valley depth and the pitch, and state where along the feature they were measured, because the two describe different things and respond differently to recipe changes.
Scallop size is not necessarily constant down the wall. A study of an adapted Bosch sequence notes that scallop size decreases with depth due to aspect-ratio-dependent etching, and that linearly ramping the etch-step duration can produce a more uniform distribution. A budget written only on a near-surface scallop can therefore misstate the deeper part of the feature.
## Build the budget around measured scallops

Budget The Scallop, Not Just The Wallmeasure size and distribution, then judge the next stepMeasure depth and pitchPeak-to-valley size and cycle spacing, with the measurement scaleCheck distribution with depthScallop size can vary along the feature, not stay constantQualify the consequenceSmoothing, fill, optical or mechanical requirement after etchA single near-surface scallop may not represent the whole sidewall.Conceptual evidence sequence · acceptance limits depend on the actual process, feature and downstream step.

Measure depth and pitch at more than one position along the sidewall, and state the metrology scale and the section location. A cross-section captures the scallops in one plane; features around a three-dimensional wall may differ.
Separate the periodic scallop profile from any overall sidewall taper or bow. The scallop budget governs the cycle-driven ripple; the larger-scale shape belongs to the profile and angle budgets, and mixing them hides which one is out of specification.
Record the process checkpoint: as-etched, after a smoothing step or after a liner. The texture that matters is the one present at the step that depends on it.
## Qualify the consequence, not only the texture
Whether a given scallop size is acceptable depends on what follows. A smoothing or removal step is sometimes applied: one reported approach oxidises the silicon and strips the oxide to reduce scallops, and another uses a post-etch plasma step, each with its own cost to dimensions and throughput.
Some applications tolerate significant scalloping. A study of metasurface fabrication reports that a non-optimised Bosch process with significant sidewall roughness still yielded reasonable optical efficiency, so the requirement comes from the device, not from a rule that scallops must always be minimised.
No single cycle-time or chemistry change removes scalloping without side effects. Shorter cycles can reduce scallop depth but may change rate, profile and mask load, and smoothing steps consume material and time; qualify any change on the actual stack.
## Close the budget with traceable evidence
Specify the depth limit, the pitch, the distribution requirement along the feature, the metrology scale and the checkpoint. Retain recipe revision, cycle parameters and chamber history so a later measurement can be compared with the earlier one.
A scallop budget bounds the periodic ripple from each cycle and follows it to the step that depends on it. Size, distribution and consequence are separate questions, and each needs its own evidence.
## Sources and scope
- [DREM: Infinite etch selectivity and optimized scallop size distribution ... in an adapted multiplexed Bosch DRIE process (DTU)](https://labadviser.nanolab.dtu.dk/images/5/5b/DREM1.pdf): reports that scallop size decreases with depth from RIE lag and that ramping the etch-step duration gives a more uniform scallop distribution.
- [Low-temperature smoothing method of scalloped DRIE trench by post-dry etching process based on SF6 plasma (Springer Open)](https://link.springer.com/article/10.1186/s40486-020-00116-x): notes that scalloping from the alternating Bosch process can cause problems in subsequent processes and degrade device performance.
Original diagrams and explanatory synthesis. No universal scallop limit, distribution target or qualified smoothing recipe is implied.

Take etch scallop budget further

Ask the copilot about this term, or have our engineers assess it against your process.

Ask CFSGPT about this → Request a technical assessment → Open the CFS app