A 2 mL clear glass vial from two suppliers, both 12 × 32 mm, both with a 9 mm neck, both certified to the same dimensional drawing, ran side by side on the same autosampler vial tray. Peak area RSD for the internal standard was 0.4% on one lot and 1.9% on the other across 60 injections. Nothing in the paperwork explained the gap. The difference was that one of those vials was made on a mold owned by the brand on the label, and the other was made on a mold owned by the factory that also sells the same vial to three other labels. That is the OEM versus private label distinction, and it is not visible from the certificate of analysis.

This article is about who owns the tooling, who controls the recipe, and what actually changes when the brand on the carton changes but the factory on the ground does not. It is not about price, and it is not about contract terms. Those belong to a different conversation. Here I am only concerned with the physical and documentary consequences of the ownership arrangement.

Procurement teams use “OEM” to mean “made by the company whose name is on the box” and “private label” to mean “made by someone else.” In the vial trade, that mapping is wrong most of the time. There are three arrangements that matter, and they behave differently.

Model A: brand owns the mold, brand owns the recipe

The brand commissioned the tool, holds the drawings, and specifies the glass batch, the annealing profile, the wash cycle, and the septum compound. The factory runs the tool under a manufacturing agreement. If the relationship ends, the tool leaves with the brand (or the brand has the right to move it). This is true OEM. It is the only model where the brand can actually guarantee a dimensional attribute over a five-year horizon, because the brand controls the steel.

Model B: factory owns the mold, brand owns the recipe

The brand specifies glass, dimensions, and septum, but the mold is the factory’s. The brand gets a private label with real specification control over materials and tolerances, but no control over tooling wear, cavitation count, or which press the tool runs on. This is the most common arrangement in the 9 mm vial market. It works well until the factory decides to retire the tool or move it to a different line.

Model C: factory owns the mold and the recipe, brand owns only the label

The brand picks a catalog vial from the factory’s range and prints a carton. No drawing control, no material control, no change notification beyond what the factory volunteers. This is private label in the strict sense. It is not inherently bad, but the buyer needs to know they are in it, because every guarantee they hold is a guarantee the factory chose to extend.

The question to ask a prospective supplier is not “are you OEM?” It is: who owns the tool, and can I see the tool number and the cavitation map? A supplier who cannot answer that in writing is in Model C whether they admit it or not.

What changes when the brand changes but the factory does not

This is the scenario that catches analytical labs. Two brands, one factory, same nominal vial. The purchasing department consolidates to a cheaper brand and the method starts drifting. Here is what actually moved.

The glass batch is not the same batch

Even in Model B, two brands running on the same tool may draw from different glass tubing lots. Tubing from different draws has different residual stress and slightly different inner diameter. On a 12 × 32 mm vial, a 0.1 mm ID shift changes the wetted surface area by roughly 3% and the headspace volume by a similar amount. For a headspace method that is already at the edge of the USP <922> repeatability window, that is enough to move peak area outside the acceptance criterion. The headspace vial buyer in particular needs to know which tubing lot the vials came from, not just which factory.

The wash and depyrogenation cycle is brand-specific

Two labels on the same factory line can specify different final wash protocols. One may run a DI rinse at 70 °C followed by a 250 °C depyrogenation tunnel; the other may run a lower-temperature wash and skip the tunnel entirely because the buyer did not ask for it. The result is two vials that look identical and have different surface chemistry. This is where ghost peaks come from, and it is why the vial specification has to include the wash cycle, not just the dimensions.

The septum compound is a separate decision

The vial and the septum are almost never made in the same place. The factory that presses the glass buys septa from a converter. Two brands on the same glass line may use different septum compounds: one a 55 Shore A silicone with a PTFE face, the other a 50 Shore A with a thinner PTFE skin. Puncture count and reseal behaviour diverge immediately. A septum rated for 100 punctures at 23 gauge will not survive 100 punctures at 22 gauge on the same needle lot. The brand that specifies the compound controls this. The brand that buys catalog septa does not.

The change-control obligation sits with the brand, not the factory

This is the part that matters most in an audit. Under a true OEM arrangement, the brand is contractually obliged to notify you of a tool change, a glass supplier change, or a septum compound change, because the brand owns the specification. Under private label, the factory has no obligation to tell the brand, and the brand has no obligation to tell you. The change-control log is the document that reveals which model you are actually in. If the log is thin, you are in Model C.

A side-by-side of what each model actually controls

Those tolerance bands are not marketing numbers. They are what I have seen hold across multi-year supply from each model, measured on a coordinate measuring machine at 20 °C on a 30-vial sample. A ±0.15 mm neck OD tolerance is enough to cause intermittent gripper failures on some autosampler trays, because the gripper jaw opening is set to a nominal with a fixed clearance.

Two field cases where the ownership model showed up in the data

Case 1: the RSD drift that no certificate explained

A contract lab running a GC vial method for residual solvents switched from a Model A brand to a Model C private label to save on unit cost. The first two lots passed incoming inspection. On the third lot, the internal standard peak area RSD climbed from 0.6% to 2.4% across a 40-injection sequence, and the retention time of the early-eluting analytes shifted by 0.02 min.

The measurement that identified the problem was headspace volume. We weighed 30 vials empty, filled 30 to the 1.5 mL mark with water, and weighed again. The Model A vials held 1.48 mL ± 0.01 mL. The Model C vials held 1.55 mL ± 0.04 mL. The internal geometry was different because the tool was different, and the split ratio was calibrated to the old headspace volume. The fix was to re-qualify the method against the new vial geometry, not to change the method parameters. The lab went back to the Model A supplier for that method and kept the private label for a different, less sensitive assay.

Case 2: the septum that failed at 60 punctures

A pharma QC lab running a 96-position tray with a 22-gauge needle started seeing coring debris in the vial after about 60 injections per position. The septum was a private label PTFE-faced silicone. The brand claimed a 100-puncture rating. We cut the septum in half and measured the PTFE skin thickness under a microscope: 0.05 mm on the private label versus 0.13 mm on the OEM reference septum. The thinner skin was being punched through by the needle bevel at a lower puncture count. The 100-puncture claim was true for a 23-gauge needle with a different bevel geometry, not for the 22-gauge needle in use.

The fix was a septum specification change: minimum 0.10 mm PTFE face, 55 ± 5 Shore A silicone, and a stated puncture rating qualified against the actual needle part number. That specification only exists in Model A or Model B, because in Model C the brand cannot dictate the skin thickness to the converter.

What to ask for, and what the answers should look like

If you are qualifying a new vial supplier, the ownership model is the first thing to establish, before you look at a single certificate. Here is the checklist I use.

• Tool number and cavitation map. Ask for the mold number and the number of cavities. A 16-cavity tool with no cavitation map means the supplier cannot trace a defect to a specific cavity, which means they cannot contain a problem.
• Glass tubing supplier and lot traceability. Ask which tubing supplier and whether lot numbers are recorded against each production run. A supplier who cannot trace a vial to a tubing lot cannot investigate a dimensional complaint.
• Wash cycle specification. Ask for the final rinse water resistivity (should be ≥ 18 MΩ·cm), the rinse temperature, and whether a depyrogenation tunnel is used. Get it in writing.
• Septum compound and skin thickness. Ask for the compound data sheet, the Shore A hardness, the PTFE skin thickness in mm, and the puncture rating qualified against your needle gauge and bevel.
• Change-control log. Ask to see the last 24 months of changes. If the log is empty, either nothing changed (unlikely) or nothing is being logged. Both are useful to know.
• Dimensional capability. Ask for Cpk on the critical dimensions (ID, neck OD, height) over the last 12 months. A Cpk below 1.33 on neck OD is a warning.
• Certificate of analysis content. A real CoA lists the measured value and the specification for each attribute, not just a pass/fail. If it only says “conforms,” it is a marketing document.

For reference on what a well-documented consumable specification looks like, the Agilent chromatography consumables documentation and the Thermo Fisher sample handling literature are both useful benchmarks. The Restek technical resources also publish dimensional tolerances that are worth comparing against whatever your supplier sends you.

Where the ownership model collides with regulated methods

If your method sits inside a regulated framework, the ownership model is not a commercial detail. It is a method attribute. Under U.S. EPA SW-846, the sample container is part of the analytical batch, and a change in container geometry or surface chemistry is a change to the method as written. Under USP <922> and the general chapter framework, the vial is an instrument component, which means the specification has to be under change control.

Here is the practical consequence. In Model A, a tool change triggers a notification, which triggers your change-control review, which triggers a revalidation decision. In Model C, the tool can be retired and replaced with a new one, and the first you hear about it is when your RSD drifts. The gap between those two realities is the entire value of the OEM arrangement. It is not about the vial being better made. It is about the vial being the same vial next year.

Torque and seal integrity, briefly

Crimp torque is a separate axis from ownership, but it interacts with it. A Model A vial with a matched septum and a calibrated crimper should hold a seal at 1.5 to 2.0 N·m with a 11 mm crimp. A Model C vial with a thinner septum face may need a different crimp setting to hold the same seal, and the setting may drift lot to lot. If you are crimping in-house, record the torque per lot and watch for drift. If you are buying pre-crimped, ask for the torque specification and the seal-integrity test method.

The decision rule

If the method is regulated, if the vial appears in a validated procedure, or if the analyte is volatile enough that headspace volume matters, buy from a supplier who owns the tool and the recipe, and require a written change-notification clause. If the method is development-stage, if the vial is a transfer vessel only, or if the assay is insensitive to a 3% surface-area shift, a Model B private label with a real specification and a real change-control log is a defensible choice. Model C is only acceptable when you have independently qualified the specific catalog vial against your method and you accept that the qualification expires the moment the factory changes anything. The question is not which brand is better. It is who can tell you, in writing and in advance, when the thing you qualified stops being the thing you qualified.

References & further reading

Related resources from our lab consumables network:

• Autosampler Vial at autosamplerhplc.com
• Vial at vials-filters.com
• Headspace Vial at headspacevialsupplier.com
• Gc Vial at gcvialfactory.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