Etch Queue Time Budget

# Etch Queue Time Budget

An etch queue time budget tracks the allowed elapsed interval between specified process events under specified environmental conditions. It connects dispatch decisions with the surface state that the next process requires.

A queue-time limit is qualified for a particular integration. It is not a universal number of minutes after every etch, and it is not the same as a tool's processing time or a fab's average cycle time.

## Define when the clock starts and stops

A rule might start at etch completion and stop when a subsequent clean begins. Another might apply from a completed clean to a deposition step. Name the events precisely and retain the source timestamps. Do not substitute tool arrival for process start unless that is the qualified stop event.

Record how the rule applies to individual wafers and lots. In a sequentially processed lot, wafers may finish the upstream step at different times. A lot-level timestamp can hide the longest wafer-level interval unless its use is explicitly conservative and justified.

## Count the whole defined interval

The teaching example assumes a 120 min limit and unchanged qualified storage conditions. Transfer takes 15 min, holding takes 60 min, and dispatch to the defined next-process start takes another 20 min. Total elapsed time is 95 min, leaving 25 min of nominal time margin.

The example is an accounting illustration, not an operating recommendation. Loading, preparation or tool waiting that occurs before the defined stop event must be included; the word dispatch does not exempt that time. The margin can be consumed by further delays and does not by itself prove that the wafer meets its surface requirement.

Use a common time basis and reliable event ordering. Handle missing, inconsistent or delayed timestamps explicitly rather than treating an unknown interval as zero. For a projected next-step start, distinguish a forecast from a completed event.

## Keep the environment attached to the limit

A surface can evolve during waiting through interactions with moisture, oxygen, residues or contaminants, depending on the exposed materials and prior processing. Storage hardware, atmosphere and exposure history can affect that evolution.

For example, Entegris describes queue-time defects associated with etch by-products and FOUP surfaces in 3D NAND manufacturing. That is a specific contamination pathway, not evidence that every etched surface follows the same mechanism or has the same time limit.

Record relevant container, purge and handling conditions together with the clock. A limit qualified in one environment should not be silently applied to another. Changing to a purged container does not erase the previous exposure history or automatically restart the clock.

## Separate overlapping rules

An etch-to-clean interval and a clean-to-deposition interval can have different start events, stop events, environments and acceptance criteria. A completed clean may end one qualified interval and start another, but only when the integration defines that transition.

If multiple limits apply simultaneously, evaluate each applicable rule. Do not average their elapsed times or combine distinct environmental histories into one equivalent exposure without a validated model.

## Qualify the interval using process evidence

Evaluate representative waits and environments using suitable surface, geometric, chemical or functional measurements. Include the relevant film stack, residue condition, layout and process sequence. A useful qualification identifies its sample coverage, uncertainty, acceptance rules and the scope of any extrapolation.

A lack of visible particles does not establish that every surface property is unchanged. Conversely, a longer interval is not proof of failure without reference to the applicable rule and evidence. Use the qualified decision process for exceptions rather than inventing a reset or assuming a rework step restores the original condition.

## Make dispatch decisions traceable

Retain the upstream completion event, expected and actual downstream start, elapsed time, applicable limit and storage history. When a limit is approaching, scheduling can use the remaining time margin, while accounting for uncertainty in tool availability and processing order.

For an exceeded or uncertain interval, preserve the wafer history and apply the defined engineering disposition. Any rework needs its own evidence for compatibility, restoration of the required surface state and effects on the remaining material budget.

A queue time budget is a clock tied to a qualified process state and environment. Clear event definitions make the arithmetic useful; process evidence establishes what that interval means for the wafer.

## Sources and scope

• [Entegris — 3D NAND Solutions](https://www.entegris.com/en/home/resources/industry-insights/3d-nand-solutions.html): etch by-products, FOUP conditions and queue-time contamination concerns.
• [Lam Research — Selective etch portfolio announcement, February 2022](https://newsroom.lamresearch.com/2022-02-09-Lam-Research-Introduces-Groundbreaking-Suite-of-Selective-Etch-Tools-to-Accelerate-Chipmakers-3D-Roadmaps): integrated processing and queue-time control in a specific tool context.

Original diagrams and explanatory synthesis. The 120 min limit and the 95 min interval are teaching assumptions, not a vendor specification or a qualified limit.