The traceability chain for a vial is seven links, and the break is almost--HPLC GC Vials Supplier

A 4 L batch of certificate-of-analysis reference material was rejected last March because the gravimetric preparation record listed a vial lot number that appeared on no purchase order, no incoming inspection sheet, and no supplier certificate. The standard was fine. The balance was calibrated. The problem was that the traceability chain for the ISO 17025 traceability for hplc vial had a broken link at the inventory-to-lot handoff, and nobody could reconstruct which physical vial held the stock solution on the day it was weighed. The assessor wrote one finding, but the same gap was hiding in 23 other preparation records from the previous 12 months.

If you are buying vials for an accredited lab, you already know the product. What you may not have is a chain that survives a third-party audit from the raw material batch to the injection on the autosampler. That chain has seven links, and in my experience running QC labs and later selling into them, six of the seven are documented by most labs. The seventh is the one that fails.

The traceability chain for a vial is seven links, and the break is almost always at link five

ISO/IEC 17025:2017 clause 6.5 requires traceability of measurement results to a reference, and clause 6.4 requires that equipment and consumables affecting the result be controlled. Most labs read that and think about balances and pipettes. The vial is a consumable that touches the sample, so it falls under 6.4.4 (consumables) and, if you are preparing calibrators in it, under 6.5 as well.

A complete chain for a single vial looks like this:

Link 1: resin lot. The polymer batch from the molder, with a certificate of analysis covering extractables and, for cyclic olefin copolymer, the residual monomer figure.
Link 2: molding lot. The specific run, with dimensional inspection data (12.00 mm ±0.10 mm body OD, 32.00 mm ±0.20 mm height, per the tolerance block on the drawing).
Link 3: septum lot. Silicone or silicone/PTFE, with durometer (typically 50 ±5 Shore A), thickness (usually 3.0 mm for a 12 x 32 mm closure), and a puncture-count qualification.
Link 4: assembly and cleaning. The wash protocol, the water grade, and the final rinse conductivity (I want to see under 1 µS/cm on the last rinse).
Link 5: incoming inspection and lot release. The dimensional check, the extractables screen, and the record that ties the physical box in the cabinet to the lot number on the certificate.
Link 6: inventory issue to the bench. Who took which box, when, and to which preparation.
Link 7: injection record. The autosampler sequence that names the vial position and, ideally, the lot.

Links 1 through 4 come from the supplier. Links 6 and 7 come from your LIMS or your sequence table. Link 5 is the one that most labs cannot produce on demand, because it is the only link that requires a human to write something down at the moment the box crosses the receiving dock. Everything else is either printed by the manufacturer or generated by software.

What an unbroken chain actually looks like on paper

An assessor who knows consumables will ask for one specific preparation and trace it backward. Here is what a defensible package contains for a single injection of, say, a 1.0 mg/mL standard prepared in a 2 mL HPLC vial:

The preparation worksheet with date, analyst, balance ID, and the vial lot number written by hand or scanned from the box label.
The supplier certificate for that lot, showing molding date, resin lot, septum lot, and the dimensional inspection summary.
Your incoming inspection record for that lot, with the actual measurements you took (not “pass”), the gauge you used, and the date.
The inventory transaction showing the box issued to that analyst on that date.
The chromatographic sequence file showing the vial position and the run.

Five documents. The first three are easy. The fourth is where it falls apart, because many labs issue vials from an open cabinet with no transaction record at all. The analyst grabs a box, the box has a lot number on the label, but the label is on the outside of a bag that was opened three months ago and the inner bags carry no lot marking. Now you have a physical vial with no identity.

Incoming inspection is the link nobody funds, and it is the one the assessor pulls

I have sat through four ISO 17025 surveillance audits in labs that buy vials by the case. In three of them, the incoming inspection record for consumables was a signed checklist with a single “received in good condition” line. That is not an inspection. It is a receipt.

What a defensible incoming inspection looks like for a 12 x 32 mm vial, done on a sampling basis (I use √n + 1, so 11 vials from a 100-piece case):

Check	Method	Acceptance criterion	Record
Body OD	Digital caliper, 0.01 mm resolution	12.00 mm ±0.10 mm	Individual readings
Height	Caliper	32.00 mm ±0.20 mm	Individual readings
Neck ID	Pin gauge or optical comparator	Per drawing, typically 6.0 mm ±0.15 mm	Go/no-go
Septum thickness	Micrometer	3.00 mm ±0.10 mm	Individual readings
Septum durometer	Shore A durometer	50 ±5	Three readings
Crimp or screw torque	Torque tester (screw) or pull test (crimp)	0.6 to 0.9 N·m for 9 mm screw; seal integrity for crimp	Per vial
Extractables screen	Blank injection, 10 µL, same gradient	No peak above 0.5 mAU at the analyte retention time	Chromatogram attached

The extractables screen is the one that catches real problems. I have seen a lot of autosampler vial that passed every dimensional check and still produced a 2 mAU ghost at 8.4 minutes on a 250 x 4.6 mm C18 column. The cause was a change in the silicone curing agent at the septum molder. The dimensions were identical. The extractables profile was not.

Case one: the lot number that lived only on the outer bag

Symptom. During a client audit, an assessor asked to trace a 2023 calibration standard back to the vial lot. The analyst produced the certificate. The certificate named lot 22-4471. The box in the cabinet was labeled 22-4471. The vials inside the box were in an unlabeled inner bag.

What the measurement showed. The lab could not demonstrate that the vial used in the preparation was from lot 22-4471 rather than from a partially used box of an earlier lot that had been consolidated into the same cabinet. The chromatographic data were fine. The traceability was not.

The fix. We required the supplier to print the lot number on the inner bag and on the base of the vial (laser marking, which most molders can do for a small premium). We also added a receiving step: the lot number is transcribed to the inventory record at the moment the outer carton is opened, not at the moment it is received. The two-step record closed the gap. Cost was about 4% on the unit price and one extra minute per case at receiving.

Case two: the septum that passed puncture count and failed under pressure

Symptom. A stability lab running a 96-injection sequence on a headspace autosampler saw increasing peak area for the internal standard over the run, about 0.8% drift across 80 injections. The RSD on the first 20 injections was 0.4%. On the last 20 it was 1.9%.

What the measurement showed. We pulled the septa after the run and measured the puncture channels with a 0.5 mm pin gauge. The channels were within the supplier’s stated puncture-count spec (they were rated for 100 punctures with a 22-gauge needle). But the headspace vial was being pressurized to 20 psi, and at that pressure the channels in the 50 Shore A silicone opened enough to leak. The supplier’s puncture count was qualified at atmospheric pressure. That is a different test.

The fix. We moved to a silicone/PTFE-faced septum for the headspace method and specified the pressure rating in the purchase spec. For the headspace vial work, the PTFE face is what carries the pressure. The lesson for ISO 17025 is that the specification on the certificate must match the operating condition, or the traceability is to the wrong test.

What the supplier certificate must contain to be usable

A certificate that says “meets specification” is worthless for traceability. The certificate has to name the lot, the test, the method, and the result. For a vial, I want to see:

Resin lot and resin supplier. Not just “PP” or “COC.” The actual lot, so you can trace a change in extractables back to the resin.
Molding date and machine. Useful when a dimensional drift appears across a run.
Septum lot and durometer. The durometer is the single best predictor of puncture behavior.
Dimensional data, not pass/fail. I want the mean and the range for the sample, so I can see whether the lot is centered or running at the edge of tolerance.
Extractables result with the method. Which solvent, which column, which gradient, and the mAU threshold.
Cleaning and packaging statement. Class of cleanroom if applicable, and whether the vials are ready to use or require a rinse.

If the supplier cannot produce this, the lab can still use the product, but the lab has to do the inspection and hold the records itself. That is a decision with a cost, and it belongs in the purchasing file, not in the analyst’s head.

The audit questions that expose a weak chain, and how to answer them

Assessors who specialize in consumables ask a predictable set of questions. Here is the checklist I use to pre-audit a lab before a surveillance visit:

Show me the certificate for the vial lot used in this preparation. If the analyst has to leave the room to find it, the chain is weak.
Show me your incoming inspection for that lot. If it is a signed checklist with no numbers, it is a receipt, not an inspection.
Show me the inventory transaction that issued that box to this analyst. This is the link that fails most often.
Show me how you qualify a new vial supplier or a new lot. There should be a written protocol with acceptance criteria, not a one-off comparison.
Show me what happens when a lot fails inspection. The nonconformance record and the disposition should exist.
Show me the septum specification for this method. Durometer, thickness, and pressure rating should be in the method or the purchasing spec.
Show me the extractables screen for the last three lots. If there is none, the lab is relying on the supplier’s certificate alone, which is a defensible position only if the certificate actually contains the data.

Six of these seven are paperwork. The second one requires a gauge and a few minutes per case. The seventh requires a blank injection, which most labs can run on an existing sequence. None of them require new equipment.

Where the standards actually put the obligation

ISO/IEC 17025:2017 clause 6.4.4 says the lab shall have a procedure for the selection and purchase of consumables that affect the validity of results, and shall verify they are used as specified. Clause 6.5 says measurement traceability applies to the result, and for a preparation made in a vial, the vial is part of the path from the reference material to the injected sample. The standard does not tell you to certify the vial. It tells you to control it.

The U.S. EPA SW-846 series treats the container as part of the method in several procedures, which is a useful cross-check: if the method names the container, the container is in scope for the quality system. USP <621> takes a similar position on the chromatographic system, though that chapter is owned by a sibling article and I will not repeat it here. The practical translation is that the vial is a controlled consumable, and controlled consumables need a lot, an inspection, and an issue record.

For labs buying through distribution, the vial supplier’s own quality system matters. Ask whether the supplier is ISO 9001 certified and whether the molder is. Ask for the last two audit reports. A distributor that cannot name the molder is a link in the chain that adds no traceability.

How to build the link that is missing, in about a day of work

If your lab cannot produce link five today, here is the sequence I would use to close it. None of this requires a new LIMS module.

Write a one-page consumable specification for each vial type in use, with the dimensional tolerances, the septum durometer, the pressure rating, and the extractables threshold. Reference the drawing.
Add a receiving inspection form with the table above. Sample √n + 1. Record the numbers, not pass/fail.
Add a lot transcription step at carton opening. The lot number goes onto the inner bag and into the inventory record. This is the step that closes the gap in case one.
Add an issue log at the cabinet. A clipboard works. The analyst writes the date, the lot, and the preparation ID when they take a box.
Run one blank injection per new lot on the method that uses the vial, and attach the chromatogram to the inspection record.
Put the lot number in the sequence table or the preparation worksheet, so link seven connects to link five without a search.

Total effort: about six hours to write the documents and one minute per case at receiving. The payoff is that the next audit question about vial traceability gets answered from a folder instead of a memory.

What to put in the purchase order so the supplier carries their end

Most traceability failures start with a purchase order that says “2 mL clear vial, 9 mm screw, qty 1000.” That PO gives the supplier no obligation to provide a lot, a certificate, or a dimensional record. A better PO line looks like this:

12 x 32 mm, 2 mL, clear glass (or COC), 9 mm screw thread, per drawing [your drawing number].
Tolerance: body OD 12.00 ±0.10 mm, height 32.00 ±0.20 mm.
Septum: silicone/PTFE, 3.00 mm ±0.10 mm, 50 ±5 Shore A, rated for 100 punctures with 22-gauge needle at 20 psi.
Certificate of analysis required per lot, including resin lot, molding date, dimensional data, and extractables screen.
Lot number marked on the outer carton, the inner bag, and the vial base.
Supplier ISO 9001 certified; molder identified on the certificate.

That PO is a traceability document. It is also a negotiating tool, because suppliers who cannot meet it will tell you so, and suppliers who can will price it in. The price difference in my experience is 3% to 8% on the unit cost, which is small against the cost of a single audit finding.

The decision rule

If you cannot trace a specific vial from the resin lot to the injection record in under ten minutes using documents on file, the chain is broken and the fix is the receiving inspection and the issue log, not a new supplier. Buy on a specification that names the lot, the dimensions, the septum durometer, and the extractables threshold, inspect a sample with a caliper and a blank injection, and write the lot number down at the moment the carton opens. That is what ISO 17025 traceability for an HPLC vial actually requires, and it is the only link in the chain that the lab, not the supplier, has to own.

References & further reading

Related resources from our lab consumables network:

External references and standards: