The traceability chain for a consumable has six links, and most labs only

A lab I work with failed an ISO 17025 surveillance audit over a box of headspace vials. Not the instrument, not the calibration certificate, not the analyst. The assessor picked up a vial from an open pack, asked for the certificate of analysis, and got a one-page sheet with a lot number, a glass expansion coefficient of 3.3 × 10⁻⁶ /°C, and a “conforms” stamp. The assessor asked one question: “How do you know the glass in this vial is the glass on that sheet?” Nobody in the room could answer. That is the traceability link that breaks in most labs, and it is the one this article is about. ISO 17025 traceability for a headspace vial does not fail at the balance or the thermometer. It fails at the boundary between the manufacturer’s process and the lab’s sample, where the chain of identity has to be documented and usually is not.

I have spent fifteen years on the manufacturing and distribution side of this, and I have sat on both ends of that audit question. What follows is what an unbroken chain actually looks like for a consumable, which link most labs cannot produce, and how to close the gap without rebuilding your whole quality system.

The traceability chain for a consumable has six links, and most labs only document three

For a calibrated instrument, ISO 17025 traceability is a well-worn path: the reference standard traces to a national metrology institute, the certificate lists the uncertainty, done. For a consumable like a headspace vial, traceability is a chain of custody and identity, not a calibration chain. There is no “true value” of a vial to calibrate against. What you are tracing is the claim that the physical object in your autosampler is the object described by the manufacturer’s specification, made from the lot of material the manufacturer says it was made from, under the process controls the manufacturer documents.

I count six links in that chain. A defensible ISO 17025 traceability file for a headspace vial needs all six:

• Link 1: Material identity. The borosilicate glass type (typically USP Type I, 33 expansion, or Type I 51 expansion depending on the application), the septum elastomer compound, and the crimp seal alloy and coating. Each should trace to a material lot.
• Link 2: Material lot to finished lot. The manufacturer’s batch record that ties the raw material lots to a specific production run, with date, machine, and operator.
• Link 3: Finished lot to dimensional and performance test data. The actual measured dimensions (not catalog nominal), the septum hardness, the seal dimensions, and the leak or pressure test results for that lot.
• Link 4: Finished lot to the certificate of analysis you received. The CoA must reference the same lot number and the same test data, signed by someone with authority.
• Link 5: CoA to the physical pack in your lab. The pack label, the lot number on the box, and the receiving record that says when it arrived and who accepted it.
• Link 6: The pack to the specific sample. The lab’s own record that says which vial lot was used for which analytical run, so a result can be reconstructed if a lot is later found defective.

Most labs have links 4, 5, and 6. They have the CoA, they have the receiving record, and if they are disciplined they log the lot per run. What they cannot produce, when pressed, is links 1, 2, and 3. The CoA says “borosilicate glass, conforms.” It does not say which glass lot, which furnace run, or what the actual measured wall thickness was versus the nominal. That is the gap.

The link that breaks is the material lot to finished lot, and it breaks at the glass supplier

Here is the mechanism. A headspace vial starts as borosilicate tubing, drawn from a glass melt. A single melt can produce hundreds of kilograms of tubing, and that tubing is cut and converted into vials over days or weeks across multiple production runs. The vial manufacturer buys tubing in bulk and may hold several tubing lots in inventory. When they convert a run of vials, they pull tubing from whatever is on the rack. The finished vial lot number reflects the conversion date, not the tubing melt.

So when the assessor asks “what glass is in this vial,” the honest answer is “tubing from melt X, drawn on date Y, converted on date Z.” If the manufacturer’s certificate only carries the conversion lot, the melt is invisible. And the melt is where the properties that matter actually originate: the expansion coefficient, the hydrolytic resistance (which drives extractables and ghost peaks), and the annealing history that sets the residual stress.

I have seen this play out concretely. A pharma QC lab was getting intermittent sodium spikes in headspace GC runs on a 51-expansion vial used for a high-temperature method (oven at 200 °C, 30-minute equilibration). The CoA said Type I, conforms. The lab had run the same method for two years on the previous lot with no issue. The new lot showed sodium at roughly 3× the historical baseline. When we pulled the manufacturer’s batch record, the tubing lot had changed three months earlier. Same conversion process, same dimensions, different melt, different hydrolytic resistance class within the Type I band. The CoA never mentioned it because the CoA tracked conversion lots, not melts. The fix was not a new method. It was a supplier change: we moved that account to a manufacturer whose CoA lists the tubing melt and the hydrolytic resistance test result per melt, and the spikes stopped.

That is the link most labs cannot produce: the melt-to-conversion trace, documented on the certificate they receive. Without it, the CoA is a conformity statement, not a traceability record.

What a traceable CoA for a headspace vial has to contain, line by line

I have read several hundred of these. The ones that survive an audit share a structure. Here is the checklist I hand to procurement when they ask what to demand from a supplier:

Two of these fields are the ones suppliers push back on hardest: the tubing melt and the measured (not nominal) dimensions. If a supplier will not put the melt on the CoA, ask why. The answer is usually that they do not track it, which means they cannot trace it, which means you cannot either.

Why the certificate alone is not traceability, and the receiving record is

A CoA is a claim. Traceability is the demonstrated connection between the claim and the object. The connection happens at goods-in, and this is where labs underinvest. When a box of headspace vials arrives, the receiving record needs to capture the lot number on the box, the quantity, the date, and the receiving inspector. If the box lot and the CoA lot do not match, stop. I have seen a distributor ship a mixed pallet where two lots were interleaved in the same carton, and the lab ran six weeks of samples on a lot that had no CoA in the file at all. The receiving clerk did not check the lot against the CoA because nobody had told them the lot mattered.

The receiving record is also where you catch the failure mode that the CoA cannot: physical damage in transit. A crimp vial that has been dropped has a stress pattern that may not show until it is heated to 150 °C and cracks at the neck. Your receiving inspection should include a visual check for neck cracks and a dimensional spot check on a sample of the carton. A go/no-go check with a caliper on the crimp OD (typically 20.0 mm ± 0.2 mm for a standard 20 mm crimp vial) takes thirty seconds and catches the lots that were packed badly.

This is also where I tell labs to think about the vial as one part of a stack. The GC vial that sits in the autosampler tray, the HPLC autosampler vial on the liquid side, and the reagent bottle holding the standard all need their own traceability records if they touch the result. The headspace vial is not special in kind, only in the pressure and temperature it sees.

How to build the missing link when the supplier will not provide it

Some suppliers, especially low-cost converters, simply do not have the melt-to-lot record and will not build one for you. You have three options, and I have used all three.

Option 1: Qualify a supplier that already documents the melt

This is the cleanest fix. The large instrument manufacturers who sell their own consumables, and the major chromatography suppliers, tend to document more because their own quality systems demand it. Agilent’s GC columns and accessories catalog and Thermo Fisher’s chromatography and sample handling line both publish detailed consumable specifications and lot documentation. Restek’s technical resources are also worth reading on the glass and septum side. The trade is cost; the gain is a CoA that survives an audit. For a regulated lab, the cost difference per vial (often 15 to 30 percent) is usually smaller than the cost of a failed audit finding.

Option 2: Build a bridging record with incoming testing

If you must stay with a supplier that will not document the melt, you can create the missing link yourself with incoming testing. The bridge is a documented incoming test protocol on a sample of every lot, with the results filed against the lot number. The minimum I would run:

• Dimensional check on 10 vials per lot: crimp OD, height, and wall thickness with a calibrated caliper, against your method’s tolerances.
• Hydrolytic resistance spot check per ISO 719 on a small sample, or a surrogate extractables test (a blank headspace run at your method temperature and a check for sodium or silicon by ICP-MS if you have it).
• Leak test on 10 crimped vials per lot: pressurize to 1.5 bar and hold for 60 seconds, or run a headspace blank and check for air ingress.
• Septum durometer check on a sample of septa from the lot against the specified Shore A value.

This does not give you melt traceability. It gives you a lot-specific performance record that you can defend as the basis for use. It is weaker than Option 1, and you should say so in your quality documentation rather than pretend otherwise.

Option 3: Treat the vial as a critical reagent and manage it like one

This is the operational discipline that ties it together. Assign each vial lot an internal identifier, log it against every analytical run, and keep the CoA, the receiving record, and the incoming test data in one file per lot. When a lot is implicated in an out-of-specification result, you can reconstruct exactly which samples used it. That reconstruction is the practical value of ISO 17025 traceability, and it is what the assessor is actually testing for.

Two field cases where the missing link was the finding

Case 1: The lot that drifted by 0.3 mm. A contract lab running USP <467> residual solvents on a 20 mm crimp vial started seeing autosampler gripper errors on one lot. The symptom was intermittent: maybe one vial in forty. The measurement that identified it was a caliper check on the crimp OD across the lot, which showed a mean of 20.3 mm against a nominal 20.0 mm, with a spread that put the upper end at 20.45 mm. The autosampler gripper had a nominal 20.2 mm opening. The vials were within the manufacturer’s stated tolerance (the CoA said 20.0 ± 0.5 mm), but that tolerance was wider than the autosampler’s. The fix was twofold: tighten the purchase specification to ± 0.2 mm, and add a receiving dimensional check on every lot. The traceability lesson is that the CoA tolerance was technically true and operationally useless, and only the receiving record caught it.

Case 2: The septum that was not the septum on the certificate. A food testing lab had a run of ghost peaks at retention times matching a siloxane pattern. The CoA specified a silicone septum with a PTFE facing. When we pulled the pack and inspected the septa, the facing was there but the durometer was visibly harder than the previous lot. A Shore A check confirmed 58 against a specified 45 ± 5. The manufacturer had substituted a different compound mid-lot without changing the CoA. The traceability failure was at link 1: the material lot on the CoA did not match the material in the pack. The fix was a supplier corrective action and a switch to a supplier that lists the septum compound lot on the CoA and holds it stable across the finished lot. The ghost peaks disappeared on the next lot.

What to put in your SOP so the next audit does not find the gap

I keep this short on purpose, because a long SOP does not get followed. The traceability clause for headspace vials needs to say five things:

• Purchase specification states the glass type and expansion, the dimensional tolerances (tighter than the autosampler’s, not the manufacturer’s default), the septum compound and durometer, and the requirement that the CoA list the tubing melt and the septum lot.
• Receiving inspection checks the box lot against the CoA lot, performs the dimensional spot check, and files the record against the lot.
• Incoming testing is defined per lot with the tests, sample sizes, and acceptance criteria above.
• Lot logging ties each analytical run to the vial lot used, in the run record, not in a separate spreadsheet.
• Supplier qualification includes a documented review of the CoA content against the checklist in this article, reviewed annually.

If you are working to EPA methods, note that SW-846 does not name your vial but binds your choice through the method’s quality control requirements, and the traceability record is what demonstrates you met them. That is the one clause I will give the sibling articles on this site; the rest is here.

The decision rule

If your supplier’s CoA does not list the tubing melt and the septum material lot, you do not have ISO 17025 traceability for that headspace vial, no matter how complete the rest of the file looks. Either qualify a supplier who documents those two fields, or build the bridging record with incoming testing and label it honestly as a performance record rather than a traceability chain. The choice is a procurement decision, not a quality one, and it should be made before the next audit notice arrives.

References & further reading

Related resources from our lab consumables network:

• Headspace Vial at headspacevialsupplier.com
• Gc Vial at gcvialfactory.com
• Reagent Bottle at reagentbottleforsale.com
• Hplc Autosampler Vial at hplcautosamplervials.com

External references and standards:

• Agilent chromatography consumables and columns
• Thermo Fisher chromatography and sample handling
• U.S. EPA SW-846 test methods for waste and materials
• Restek technical resources for chromatography