A customer’s incoming dimensional check rejected a lot of HPLC vials at 0.42 mm above the nominal crimp-neck diameter. The supplier’s certificate package could not show which resin batch produced the pellets molded that week. The certificate existed; the traceability did not. Resin-to-carton traceability for analytical testing consumables is the work of keeping that link intact through every handoff, and it is where most supplier quality systems quietly come apart. Analytical testing raw material traceability is not a document problem at the end of the line. It is a chain of identity decisions that begins at the resin silo and the glass tubing draw, and every handoff that drops a lot code is a place where a deviation becomes unattributable.
The traceability chain has to survive five handoffs, and two of them are usually undocumented
A molded vial or a polypropylene centrifuge tube passes through more identity changes than most purchasing teams assume. The chain runs: resin or glass batch, conversion (molding, drawing, or extrusion), secondary assembly, packaging, and shipment. Each step should carry forward a lot identifier that resolves to the upstream material batch. In practice, two handoffs are routinely broken.
The first is the resin-to-molding handoff. A molder receives resin in 25 kg bags or in bulk, and the incoming inspection record often captures the resin grade and the supplier lot but not the batch number the resin producer used for its own CoA. When the molder blends regrind with virgin material, the blend ratio and the regrind source are frequently absent from the production record. That single omission makes the finished part untraceable to a resin batch, even if every other record is perfect.
The second is the glass tubing draw. Glass vials begin as tubing, and tubing is drawn from a glass melt batch. The draw date, the tubing lot, and the converting line that formed the vial are three separate identifiers. Many converters record the vial lot and the converting date but not the tubing lot. When a hydrolytic resistance failure appears months later, the lab can identify the vial lot but cannot reach back to the glass batch that caused it.
The practical rule: a finished-goods lot number should resolve, in two lookups or fewer, to a resin batch number or a glass tubing lot number, plus the molding or converting parameters used on that run. If the lookup takes three phone calls, the chain is already broken. For a broader view of how these consumable decisions sit inside a testing program, see analytical testing at testinganalytical.com.
What a lot number must resolve to, and the difference between a lot and a batch
Teams often use “lot” and “batch” interchangeably. They are not the same, and the confusion is where traceability claims get inflated.
A batch is a single production run: one resin batch fed to one molding machine over one continuous window, or one glass tubing lot run on one converting line. A lot is the unit the supplier ships and the lab receives. One lot can contain several batches, and one batch can be split across several lots. That is normal. What matters is that the lot-to-batch mapping is recorded and retrievable.
A traceable lot number should resolve to the following minimum data set:
- Material identity: resin grade and producer batch, or glass tubing lot and draw date.
- Conversion record: machine or line identifier, date, and the key process parameters for that run (melt temperature, cycle time, draw speed).
- Regrind status: whether regrind was used, the percentage, and the source lot of the regrind.
- Secondary operations: any assembly, printing, or sterilization step with its own lot code.
- Packaging lot: the bag or carton identifier, which is the unit the lab actually receives.
If any of those five is missing, the lot number is a shipping reference, not a traceability record. The distinction matters because audit findings and deviation investigations both depend on it. A supplier that cannot produce the conversion record for a specific lot is telling you the chain ends at the loading dock.
The resin batch is the first link, and the regrind question decides whether it holds
For polymer consumables, the resin batch is the origin of every downstream property: extractables profile, dimensional stability, and color. The resin producer’s CoA carries a batch number and typically a set of properties (melt flow rate, density, additive package). The molder’s job is to keep that batch number attached to the parts.
Regrind is the variable most likely to sever the link. When a molder recycles sprues and rejects, the regrind carries the history of every batch it came from. A blend of regrind from three batches produces a part whose material identity is a mixture, and the lot number can only honestly point to a blend record, not a single resin batch. That is acceptable if the blend record exists and states the percentages. It is not acceptable if the regrind is unrecorded.
Ask a supplier two questions and listen for the specificity of the answer. First: what percentage of regrind is permitted, and is it capped per product family? A documented cap (for example, no more than 20% regrind, virgin-only for cell-culture and trace-analysis grades) is a sign the process is controlled. Second: can you show the regrind source lot for a given finished lot? If the answer is a policy statement rather than a record, the traceability is aspirational.
Where the resin link usually snaps
The snap point is the incoming goods record. Many molders log the resin supplier and the delivery date but not the producer batch number printed on the bag. Once the bag is emptied into a hopper, the batch identity is gone unless it was written down. A supplier that scans bag labels into the production record at the hopper has closed this gap. One that relies on the delivery note has not, because the delivery note often covers multiple batches.
Glass tubing lots are the weakest link in vial traceability
Glass vials and ampoules have a two-stage origin: the glass melt and the tubing draw. The melt batch determines the glass composition, including the boron oxide content that governs hydrolytic resistance under USP <660> and the ISO 719 / ISO 720 test regimes. The draw converts the melt into tubing, and the tubing lot is the identifier that should travel with the vial.
In practice, many converters buy tubing from a distributor and receive a tubing lot number that resolves to a distributor stock record, not a melt batch. The distributor may have consolidated tubing from more than one draw. When a delamination or a hydrolytic resistance failure appears, the investigation reaches the distributor lot and stops. The melt batch is unreachable.
Two controls close this gap. First, require that the tubing lot number on the incoming record resolves to the glass manufacturer’s melt batch and draw date, not a distributor’s internal code. Second, require that the vial lot’s conversion record names the tubing lot used on that run. A converter that can name the tubing lot for a specific vial lot has a chain that reaches the melt. One that cannot has a chain that reaches the warehouse.
The same logic applies to the dimensional side. Crimp-neck and screw-thread vials are defined by standards such as GPI 13-425 for the thread finish and the dimensional envelopes in the relevant pharmacopeial chapters. A dimensional deviation that traces to a tubing lot is diagnosable. One that traces to nothing is a recurring mystery. For dimensional and fit context on the vial side, see HPLC vial at vialforhplc.com.
Where the chain breaks in practice: four failure points worth auditing
Traceability failures cluster at the same four points across suppliers. Each has a specific test you can apply during a supplier audit or a deviation investigation.
Failure point 1: the blend record
If regrind or recycled glass cullet is used, the blend record is the only document that can reconstruct material identity. Ask for the blend record for one finished lot and check whether it names source lots and percentages. A blend record that says “regrind, standard ratio” is not a record.
Failure point 2: the line-changeover log
A molding machine or a converting line runs multiple products. If the changeover log does not record the last lot of the previous product and the first lot of the next, parts can be mixed at the boundary. Ask how the supplier prevents lot mixing at changeover. A purge-and-verify step with a documented first-article check is the control. A verbal assurance is not.
Failure point 3: the sterilization or secondary-operation lot
If the product is sterilized or printed, the secondary operation assigns its own lot, and that lot must map back to the molding lot. A sterilization record that lists quantities but not the source molding lots breaks the chain at the last step before packaging.
Failure point 4: the packaging lot
The carton is what the lab receives, and the carton label is the entry point for any investigation. If the carton label carries only a packaging lot and not the finished-goods lot, the lab cannot reach the molding record. Require that the carton label carry the finished-goods lot and that the packaging record map it to the molding lot. This is the handoff where the chain most often ends, because packaging is treated as logistics rather than as part of the quality record.
A traceability audit checklist with concrete pass criteria
The table below converts the discussion into pass/fail criteria you can apply to a supplier’s records for a single finished lot. The numbers are illustrative thresholds, not universal standards; set them against your own method requirements.
| Check | What to request | Pass criterion |
|---|---|---|
| Resin identity | Producer batch number for the resin used in the lot | Resolves to a CoA batch within two lookups |
| Regrind control | Blend record with source lots and percentage | Regrind capped (for example, ≤20%) and recorded per lot |
| Glass tubing | Tubing lot and draw date for the vial lot | Resolves to a melt batch, not a distributor code |
| Conversion record | Line ID, date, and key parameters for the run | Parameters recorded for the specific lot, not a generic spec |
| Changeover control | Log showing last lot of previous product and first-article check | Documented purge and first-article verification |
| Secondary operation | Sterilization or print lot mapped to molding lot | Mapping recorded, quantities reconcile |
| Packaging | Carton label and packaging record | Carton lot maps to finished-goods lot and molding lot |
| Retention | Retained samples per lot | Retention quantity and duration defined (for example, 2 years or per contract) |
Run this checklist against one lot per product family, not against the supplier’s quality manual. The manual describes the system. The lot record shows whether the system ran. A supplier that passes on the lot record has a working chain; one that passes only on the manual has a documented intention.
Two diagnostic examples: reading a traceability break from the bench
The following examples are hypothetical and illustrate the diagnostic logic. They are not reports of specific customer cases.
Example 1: a drift in extractables that traces to a resin batch change
Symptom: a gradient HPLC run shows a rising baseline and a new peak at around 8.2 minutes that appears only with one shipment of polypropylene tubes. The peak is absent with the previous shipment.
Useful measurement: compare the extractables profile of the suspect lot against a retained sample from the prior lot using the same solvent and the same column. If the new peak appears only in the suspect lot, the material differs. Then request the resin batch records for both lots. A change in resin batch, or a change in regrind percentage, is the likely cause.
Corrective action: quarantine the suspect lot, request the resin batch CoA and blend record, and require the supplier to confirm whether the resin batch changed. If the batch changed without notification, treat it as an unapproved material change and re-qualify. For the chromatographic side of this kind of investigation, the resources at Restek cover extractables screening and column selection.
Example 2: a dimensional deviation that traces to a tubing lot
Symptom: crimp-neck vials from one lot show a neck diameter 0.3 to 0.5 mm above nominal, and capping torque readings scatter by more than 0.4 N·m across the lot. The sealer alarms on a fraction of the lot.
Useful measurement: measure neck diameter on a sample of 30 vials across the lot and compare the distribution to the prior lot. A shift in the mean, not just the spread, points to a material or tooling change rather than random variation. Then check whether the tubing lot changed between the two lots.
Corrective action: request the tubing lot and draw date for the suspect lot. If the tubing lot changed, the dimensional shift likely originates in the tubing draw or the glass composition. Reject the lot against the drawing tolerance, and require the supplier to record tubing lot on the conversion record so the next investigation can reach the melt. For the broader consumables picture, see chromatography consumables at chromatographyconsumables.com.
What to write into the purchase specification so the chain holds
Traceability is a contract term before it is a quality practice. A specification that requires only a certificate of analysis at shipment does not require a chain. The specification should require the following, stated as deliverable records rather than as intentions:
- Lot-to-batch mapping: each shipped lot resolves to a resin batch or glass tubing lot, retrievable on request within a stated time (for example, two business days).
- Change notification: any change in resin batch, glass tubing source, regrind policy, or molding line triggers written notification before shipment.
- Record retention: traceability records retained for a stated period (for example, five years or per contract), with the same retention applied to the lot-to-batch mapping.
- Carton labeling: the carton label carries the finished-goods lot, and the packaging record maps it to the molding lot.
- Retained samples: a defined retention quantity per lot, stored under defined conditions, available for comparison during an investigation.
These terms do not duplicate the certificate requirements covered elsewhere on the site. They define what the supplier must be able to reconstruct after the certificate has been filed. For method and materials context on the regulatory side, the U.S. EPA SW-846 test methods show how method documentation depends on material identity, and the consumables catalogs from Agilent and Thermo Fisher illustrate the level of specification detail that is normal for chromatography hardware. The same level of detail belongs in a consumables purchase specification.
The decision rule: traceability is proven by one lot record, not by a policy
When you evaluate a supplier for analytical testing raw material traceability, do not start with the quality manual or the certificate template. Pick one finished lot, ask for the resin batch or glass tubing lot behind it, and time how long the answer takes. If the supplier can produce the mapping and the conversion record within the stated window, the chain is real. If the answer is a policy statement, a distributor code, or a request to “check with the plant,” the chain ends before it reaches the material. Make that one-lot test a condition of approval, repeat it at each re-qualification, and write the retrieval window into the purchase specification. The chain holds when the record exists before you ask for it.
References & further reading
Related resources from our lab consumables network:
- Analytical Testing at testinganalytical.com
- Hplc Vial at vialforhplc.com
- Chromatography Consumables at chromatographyconsumables.com
- Lab Vials at labvials.com
External references and standards: