A lab I work with failed an ISO 17025 surveillance audit over nine boxes of 2 mL clear glass vials. The assessor did not question the calibration certificate for the balance. He asked one question: “Show me the lot number on the vials used in the chromium run on 14 March, and show me the certificate that ties that lot to the dimensional test you say you perform.” The lab had a supplier certificate of analysis. It had a purchase order. It had a receiving log with a date and a quantity. It could not connect the three into a single unbroken chain, and the corrective action cost them eleven weeks and a re-run of a 40-sample batch. That is the gap this article is about: ISO 17025 traceability for sample vials, and specifically the one link in that chain that most labs cannot produce when an assessor asks for it in the room.
Traceability for a consumable is not the same problem as traceability for a balance or a pipette. A balance has a serial number and a calibration certificate with an unbroken chain to a national standard. A vial has a lot number, a molding date, a resin batch, and a set of dimensional measurements that may or may not have been taken on that specific lot. ISO 17025 clause 6.4.4 requires that equipment and, by extension, reference materials and consumables affecting the validity of results be traceable to a reference standard. The practical reading for a vial is narrower than people assume: you need to prove that the physical article in the autosampler tray is the article whose properties you verified. Everything else is documentation hygiene.
I have written this chain out for more than a dozen labs, and it always has the same five links. When a lab fails, it is almost never link one or link five. It is link four.
- Link 1, the PO. A purchase order that names the product code, the lot, and the specification. Most labs have this.
- Link 2, the receiving record. A receiving log that captures the lot number as printed on the carton, the date, the quantity, and the condition. Most labs have this, though the lot number is often transcribed wrong by one digit.
- Link 3, the supplier certificate. A CoA from the manufacturer that states the lot and the tested properties. Most labs have this, filed in a shared drive nobody can search.
- Link 4, the link between the supplier’s lot and the lab’s own incoming verification. This is where labs fail. The CoA says the supplier tested dimensional conformance on lot A. The lab’s own incoming check, if it exists, was performed on three vials pulled from a box in the middle of a nine-box shipment, and the result was recorded against a date, not a lot. There is no record that the three vials came from lot A.
- Link 5, the record of use. Which lot went into which analytical run. Fewer labs have this than admit it.
An assessor who understands consumables will walk links one through three quickly. The interview happens at link four. If your incoming verification record cannot be tied to a specific lot by number, the CoA is decoration. The chain is broken at the exact point where the supplier’s claims become your responsibility.
Why link four breaks so reliably
Because incoming verification for a vial is treated as a formality. A technician opens a carton, checks that the vials are not chipped, checks that the caps are present, signs a form. That form records a date and a count. It does not record a lot number, because the lot number is on the outside of the carton and the form is inside the lab. The fix is not more testing. The fix is a sampling protocol that pulls the verification sample from an identified lot and records the lot on the same form as the measurement. That single change closes link four for every future shipment.
What ISO 17025 actually asks of a consumable, clause by clause
Clause 6.4 covers equipment. Clause 6.5 covers metrological traceability. Clause 7.1.5 covers the review of requirements, and it is the clause that reaches the vial most directly, because the vial is part of the method’s specified consumables. When a method says “use a 2 mL clear glass vial with a PTFE-lined silicone septum,” the vial is a specified item, and clause 7.1.5.2 requires you to select and verify that the item meets the method’s requirements.
The clause most assessors use to test a lab is 6.5.2: metrological traceability shall be to a reference standard. For a balance, that is a weight set. For a vial, the reference is dimensional and material: the crimp neck dimensions defined in the manufacturer’s drawing and, where applicable, the neck finish defined by the dimensional standards the industry uses (the 11 mm and 13 mm crimp finishes and the 8-425 and 9 mm screw finishes are the ones you will meet most). If your verification measures the neck diameter with a calibrated micrometer, that measurement is traceable because the micrometer is traceable. The vial’s properties are then traceable by measurement. If your verification is a visual check, it is not traceable to anything, and no amount of filing will change that.
The other clause that matters is 7.7, ensuring the validity of results. A vial that leaches, adsorbs, or fails to seal is a source of invalid results. The traceability chain is how you prove you controlled that source. It is not a paperwork exercise; it is the evidence that the result was produced under controlled conditions.
The incoming verification protocol that survives an audit
I recommend a three-tier protocol scaled to the analytical risk. It is deliberately cheap at the low end, because most labs do not need to measure every carton.
| Tier | When it applies | What you measure | Acceptance criterion | Record |
|---|---|---|---|---|
| Tier 1, identity | Every lot, every product | Lot number matches carton, CoA, and PO; visual for chips, cracks, missing caps | Zero visible defects in a 30-vial sample | Lot, date, initials, sample size |
| Tier 2, dimensional | New supplier, new mold, or every 10th lot | Neck OD and height with a calibrated micrometer (0.001 mm resolution) | Within the manufacturer’s drawing tolerance, typically ±0.15 mm on neck OD for a 11 mm crimp | Individual measurements, micrometer ID, calibration due date |
| Tier 3, functional | Trace analysis, methods at or below 1 ppb, or any method with a documented carryover issue | Seal integrity by weight loss after 24 h at 4 °C; blank run for extractables | Weight loss <0.5 mg per vial; no peak above 3× baseline noise in the blank | Weights, blank chromatogram, lot |
The tier 2 measurement is the one that closes link four. A micrometer reading of 11.02 mm on a neck whose nominal is 11.00 mm with a ±0.15 mm tolerance tells the assessor that you measured the article, on the lot, with a traceable instrument. That is what traceability for a consumable looks like in practice. The sample vial itself is a commodity; the measurement record is what makes it a controlled item.
Two field cases where the chain failed on the measurement, not the paperwork
Case one: the crimp that sealed on the bench and leaked in the tray
A lab running EPA method 8260 on groundwater had intermittent low recoveries on the volatile analytes, always in the last third of a 40-vial tray. The CoA for the lot was clean. The crimper was set to the manufacturer’s nominal 1.5 in-lb torque. The measurement that identified the problem was seal integrity by weight loss: vials crimped at the nominal torque lost an average of 1.8 mg over 24 hours, against 0.3 mg for a control lot. The cause was a neck OD that ran at the low end of tolerance, 10.88 mm against an 11.00 mm nominal. The crimp sealed initially but relaxed as the aluminum deformed against an undersized glass neck. The fix was a tier 2 check at receiving: any lot whose sampled neck OD fell below 10.92 mm was rejected before it reached the bench. The recovery problem disappeared on the next lot, and the lab added a 0.5 mg weight-loss criterion to its tier 3 protocol. This is a traceability problem because without the neck measurement, the lab could not prove the vials caused the failure, and without that proof, the supplier would not accept the return.
Case two: the lot number that did not match
A different lab had a clean chain on paper and a broken one in fact. The receiving log recorded lot 22-0417. The carton said 22-0417. The CoA said 22-0417. But the vials in the autosampler tray, pulled for a re-run after a failed sequence, had a different lot printed on the bag label inside the carton. The supplier had packed a partial bag from an earlier lot into a later carton. The measurement that caught it was a routine tier 2 neck check: the sampled vials measured 11.09 mm, and the CoA for lot 22-0417 stated a mean of 11.01 mm. That discrepancy triggered a bag-level check, which found the mismatched bag. The fix was procedural: the receiving protocol now requires the technician to record the lot on the inner bag label, not just the carton, and to reject any carton where the two disagree. The lesson is that the chain is only as strong as its weakest physical label, and the carton is not the article.
These cases share a structure. The paperwork was fine. The measurement, or the absence of one, was the failure point. That is the pattern an assessor is trained to find.
What the supplier certificate can and cannot carry
A CoA from a competent manufacturer is a real document. It states the lot, the tested properties, the test methods, and the acceptance criteria. What it cannot do is transfer traceability to your lab automatically. Traceability is a property of your measurement chain, not the supplier’s. The CoA supports link three of the chain. It does not close link four.
When you evaluate a supplier for traceability, ask for the quality manual index, not the manual. The index tells you which procedures exist and which lot-level records are retained. Then ask two specific questions: how long are lot-level dimensional records retained, and can you retrieve the record for a specific lot from 18 months ago on 48 hours’ notice. A supplier who can answer both with a procedure reference is a supplier whose CoA you can rely on. A supplier who answers with a marketing brochure is a supplier whose CoA is a template. For reference, the major instrument vendors publish their consumable specifications openly; Agilent’s chromatography consumables documentation and Thermo Fisher’s sample handling literature are useful benchmarks for what a complete specification sheet contains. Restek’s technical resources are worth reading for the dimensional tolerances they publish on their vial lines, because those numbers are the ones you will end up measuring against.
For labs sourcing through distribution, the same question applies one layer down. A distributor who cannot produce the manufacturer’s lot record is a distributor whose CoA is a copy. If you buy through a channel like vials with caps, confirm that the lot on the carton traces to the manufacturer’s record, not to the distributor’s repack. A manufacturer with a documented molding and inspection process, such as Aijiren, will provide lot-level dimensional data on request. That is the difference between a supplier and a source.
The traceability record set: what to keep, where, and for how long
The record set is small. It does not need a new database. It needs four documents per lot, filed together, retrievable by lot number.
- The PO line with product code and lot, if the lot was specified at order.
- The receiving record with lot, date, quantity, inner-bag lot, and condition.
- The supplier CoA for that lot, with the manufacturer’s test methods and acceptance criteria visible.
- The incoming verification record with the tier, the sample size, the individual measurements, the instrument ID, and the calibration due date.
Retention: align with your method’s record retention requirement, which for most accredited methods is five years, and never less than the retention period for the analytical records the vials support. If you keep raw data for five years, keep the lot records for five years. A lot record that outlives its analytical data is harmless. Analytical data that outlives its lot record is an audit finding.
The retrieval test is the real test. Pick a lot from two years ago, and time how long it takes to produce all four documents. If it takes more than 15 minutes, the chain exists on paper and fails in practice. An assessor will ask for exactly this, and he will not give you a day to find it.
The decision rule for ISO 17025 traceability of sample vials
Treat the vial as a specified consumable with a measurement-based traceability chain, not as a commodity with a certificate. That means: one lot number recorded at receiving from the inner bag, one tier 2 dimensional check on the neck with a calibrated micrometer at least every tenth lot, and one filed record set per lot that you can retrieve in 15 minutes. The link that breaks is the one between the supplier’s lot and your own verification. Close it with a measurement, on the lot, with a traceable instrument, and the rest of the chain holds. If you cannot measure the article, you cannot trace it, and a CoA alone will not carry you through an audit. The chromatography products you buy are only as controlled as the records that connect them to your results.
References & further reading
Related resources from our lab consumables network:
- Sample Vial at samplevialprice.com
- Chromatography Products at chromatographyselling.com
- Aijiren at aijirentech.com
- Vials With Caps at vialswithcaps.com
External references and standards: