When a lapped wafer shows a different thickness trend near its edge, begin by checking whether the measurement and sample history support the pattern. Reconstruct the wafer identity, point map, orientation and method before attributing the result to a machine setting. A wafer lapping edge trend can arise in an investigation with several interacting conditions, and the first useful step is to preserve evidence that lets the team distinguish them.
Hisemi lists wafer lapping and polishing equipment, including a separate precision lapping and polishing configuration. Confirm the machine actually used for the trial rather than transferring a result between those listings. Its machine information provides the equipment context, while the actual substrate, fixture, route and inspection method define the trial. The following workflow organizes a review after a sample run. Any process adjustment should be made by authorized personnel using the applicable equipment instructions and the project's approved plan.
Check identity, comparable measurement locations and trial history before assigning a cause.

Hisemi wafer lapping equipment · Hisemi precision lapping and polishing configuration
Verify that the edge readings belong to comparable locations
Find the original incoming and final measurement maps. Check sample identity, orientation and the coordinate reference. A point described as near the edge on two sheets may be at a different physical offset or measured after the wafer was rotated. Record the actual location and any project-defined exclusion zone. Keep missing or invalid points identified.
Confirm what was measured. Local thickness, a thickness-variation summary and surface appearance answer different questions. Retain raw point or scan data and identify the tool, units and applicable method. A summary number may show that a requirement was missed while hiding the location and extent of the pattern. The map is needed to investigate where the change appeared.
Ask the measurement owner to review validity at the affected positions. Edge access, support or instrument behavior can matter under the actual method. Follow the approved procedure for verification or repeat measurement. A new reading should retain its own date and conditions rather than overwrite the first result without explanation.
Reconstruct the sample through the trial
Gather the incoming condition, process route, carrier or fixture identity and handling record. Identify the machine and setup revision used. Note any approved intervention, interruption or change during the run. Include the sample's actual orientation where it matters to the process and compare it with the final map.
If several wafers were evaluated, keep each one's route and results separate. Determine whether the pattern repeats at corresponding wafer locations, process positions or measurement positions. That comparison can help organize a follow-up without establishing cause prematurely. Preserve units that did not show the trend as well as those that did.
Review the scope of the initial trial. A demonstration on one substrate and support arrangement gives limited evidence for another combination. If the project changed material or fixture after the plan was written, identify the affected samples and retrieve the applicable instructions or authorization. The revised condition belongs in the investigation.
Separate observations from proposed explanations
| Evidence | Question to resolve first | Possible next action under the approved plan |
|---|---|---|
| Incoming map | Was a comparable trend already present? | Compare valid corresponding locations |
| Final raw data | Are the affected readings repeatable and method-valid? | Have metrology review or repeat designated points |
| Orientation record | Does the pattern follow a known coordinate or position? | Preserve orientation in the next comparison |
| Process record | What actually changed between trials? | Review one documented variable or approved experiment |
| Fixture and handling | Was the stated arrangement used throughout? | Verify actual hardware and permitted contact conditions |
| Disposition | Which acceptance criterion is affected? | Hold, investigate or repeat according to project authority |
Use neutral language in the issue report. "Three valid points in the designated outer region are lower than corresponding central readings" is an observation if the data support it. "The machine caused edge removal" is an explanation that requires further evidence. Keeping those statements distinct makes supplier and customer review more productive.
Include representative images where they help locate the condition. A photo of a wafer can show orientation or a visible feature, while it may not establish the thickness result. Label each image with sample ID and date, and cite the measurement record separately. Avoid using an unscaled photograph as the only evidence for a dimensional claim.
Plan a comparison that can answer one question
Agree the follow-up objective with process engineering and metrology. If the first question concerns measurement repeatability, keep the sample and applicable method under the approved repeat procedure. If the objective concerns a process difference, define which samples and conditions make the comparison meaningful. Identify what remains controlled and what changes.
For an illustrative investigation, two wafers show different outer-region maps after nominally similar runs. The records reveal different carrier revisions. That is a documented difference to investigate. It does not prove the carrier caused the result. The next authorized trial can compare defined arrangements with matched material and measurement conditions, using the agreed acceptance and analysis plan.
Preserve unsuccessful and uncertain outcomes. A follow-up that cannot reproduce the trend still contributes evidence about the conditions tested. Report the complete comparison and identify limits in the sample size or setup. A supplier can then help review an actual pattern instead of a selection of the most dramatic screenshots.
Close the issue with a disposition and a reproducible record
The review should state which results are valid, which requirement is affected and what action is authorized. If a sample is accepted with a documented exception, keep the approval and scope. If another run is required, identify the new sample plan and unresolved question. Avoid leaving a generic process issue open after the specific sample has already moved.
Update the approved setup or inspection method only through the relevant change process. Link the new revision to the evidence that supports it, and retain the previous state for earlier samples. Production operators need the released instruction, while engineering retains the investigation details and comparison files.
When discussing the trial with Hisemi, provide the actual equipment and sample identity, incoming and final maps, methods, fixture or carrier information and process timeline. Specify which observations have been confirmed and which explanation is still being considered. This gives the supplier a practical basis for questions and a proposed follow-up.
A concise evidence index can save considerable review time. Put the sample identifiers, relevant map revisions and process records on one page, and mark the particular points under discussion. The supplier should be able to find the same location the customer is describing. If multiple coordinate conventions appear in the file, add an approved cross-reference and preserve the original labels.
Check the proposed production consequence separately from the investigation. One sample outside a requirement may affect a pilot release, while the project may require more evidence to decide the scope of a broader hold. Have the authorized quality owner make that decision using the actual acceptance plan. The issue log should say which units and operations are affected so unrelated material is not swept into an ambiguous statement.
At the next review, begin with the question the previous trial was meant to answer. Compare its new evidence against that objective, then decide whether another question remains. This prevents a series of setup changes from producing many results without a clear chain of reasoning. Keep the final disposition tied to the required property and the conditions actually demonstrated.
Bring metrology and process engineering together for the first comparison. One team may recognize an instrument access limitation while the other recognizes a changed support arrangement. Record each observation against the evidence that supports it. Agree which question the next check will answer and what result would require further review. The plan becomes easier to follow when the intended comparison is written before another sample is processed.
If retained samples are available, identify their storage and handling condition before requesting additional measurements. Use the applicable procedure and note any intervening process. A wafer returned from another laboratory may no longer represent the original post-lapping state. Keep that history visible so the team does not confuse a new condition with a repeat of the first measurement.
Record the customer-facing explanation only after the review distinguishes confirmed observations from unresolved possibilities. A concise report can say which sample and region were affected, what evidence was valid and what follow-up is authorized. Keep internal hypotheses in the investigation record with their supporting data. That gives purchasing an accurate description for the supplier while allowing engineering to continue the technical work. When the next trial closes the issue, append its actual result and decision rather than rewriting the first observation.
Questions from the wafer review
Should the first response be to change a machine setting?
Start by preserving and reviewing sample, method and process evidence. Have authorized engineering personnel decide which follow-up is appropriate. A setting change without a defined question can make later comparison harder.
What does a thickness-variation summary leave out?
It may omit where the highest and lowest valid readings occurred and how the spatial pattern looks. Retain the map or scan and applicable point validity information alongside the summary.
How should rotated samples be compared?
Use the documented orientation and coordinate reference to map comparable locations. If orientation was not retained, state that limit and agree on the follow-up. Avoid assuming point names identify the same physical site.
Does a repeated outer-region pattern identify the cause?
It supports a repeatable observation under the recorded conditions. Review measurement validity and process variables before assigning cause. Use an authorized comparison to evaluate specific explanations.
Can a photo replace the measurement file?
Use photographs for relevant appearance, identity and orientation context. Keep dimensional data, units and methods separately available. A picture alone usually cannot establish the stated thickness result.
What should accompany a request for supplier review?
Send machine and sample IDs, incoming and final maps, measurement basis, fixture references and trial sequence. Identify the affected criterion and the exact question the next review should resolve.
Discuss the confirmed edge trend with Hisemi
Send Hisemi the labeled measurement maps, substrate and sample details, actual fixture and process references and the proposed acceptance criteria. Ask for a review and a defined follow-up under the applicable equipment procedure.

