In 2022 a mid-size contract lab sent me a lot of 2 mL clear glass vials that had passed incoming inspection on every dimensional print: 12.0 mm outer diameter, 32.0 mm height, 6.0 mm neck ID, 11 mm crimp finish. Within four injections on their 1260 Infinity autosampler, the run died with a “vial not found” error on position 34 of a 100-vial tray. The vials were dimensionally correct. They were also 0.4 mm taller at the shoulder than the previous lot, which is enough to change how the tray gripper seats them. The brand on the box was the lab’s own private label. The factory was the same one that made the OEM product. The difference was that the OEM tooling had a shoulder-height check in the process control plan and the private-label tooling did not, because the tooling itself was a different cavity set that the private label had never asked to audit. That is what OEM versus private label for an HPLC vial actually comes down to. Not the price sheet. The tooling, the recipe, and who is on the hook when a lot drifts.
This is the part of the OEM versus private label difference that procurement decks usually skip. The chemistry is not what changes. The certificate of analysis is not what changes, at least not in form. What changes is who owns the steel, who writes the process control plan, and who carries the liability when a lot of 50,000 vials has to be quarantined because the shoulder height crept 0.4 mm north of the print.
An HPLC vial is formed on a mold. The mold is a set of cavities, usually 8 to 48 of them, cut into hardened steel, mounted on a glass forming line (typically an IS machine for tubing glass, or a vertical marble-mold line for the older borosilicate process). The cavity set is the tooling. It costs anywhere from $40,000 to $180,000 depending on cavity count, finish complexity, and whether the neck and shoulder are in the same block as the body.
In a true OEM relationship, the brand owner (the company whose name is on the box) either owns that tooling outright or has exclusive use rights written into the supply agreement. The tooling is tagged, serialized, and referenced in the drawing revision. When the brand changes a print, the tooling is reworked or remade at the brand owner’s cost. When the factory wants to run a different product on the same line, the OEM tooling comes off and goes into a controlled crib.
In a private label relationship, the tooling belongs to the factory. The private label brand rents capacity on it. This is not automatically a problem. Many private labels are made on the same cavities as the OEM product, back to back, with a label change at the end of the line. But the private label has no say in when the tooling is refurbished, when a cavity is pulled for polishing, or when the factory decides to run a lower-grade glass batch through the same mold to hit a cost target.
What the tooling controls that the drawing does not
A drawing says 32.0 mm ± 0.3 mm. The tooling says something more specific: the shoulder radius, the neck taper, the bottom push-up height, and the concentricity of the neck to the body. Those are not always on the drawing. They are often on a factory-internal process sheet that the private label never sees. When a private label vial “looks the same” but fails on a specific autosampler, the difference is almost always in one of those four unlisted dimensions.
I have seen a private label vial with a bottom push-up 0.6 mm deeper than the OEM part. Dimensionally fine on the print. On a Agilent autosampler with a bottom-sensing vial-present sensor, the deeper push-up changed the optical path enough that the sensor read “no vial” on 3% of positions. The lab spent two days chasing an autosampler firmware issue before someone measured the push-up with a depth gauge.
The recipe is the part nobody puts in the contract
Glass is not a commodity in the sense procurement usually means. A Type I borosilicate vial is a specific formulation: approximately 81% SiO₂, 13% B₂O₃, 4% Na₂O/K₂O, 2% Al₂O₃, with the exact ratios and the trace element profile varying by supplier. The hydrolysis resistance (the USP <660> test, the ISO 719 / ISO 720 powder and grain tests) depends on that formulation and on the thermal history of the forming and annealing steps.
The recipe includes the forming temperature, the annealing lehr profile, the mold release agent, and the wash chemistry after forming. A private label factory can change any of those without telling the brand, because the brand does not own the process. An OEM factory cannot, or at least has to document the change against a change control that the brand owner signed.
Why this matters for extractables
If you are running trace-level LC-MS work, the extractables profile of the vial is part of your method. A change in the annealing profile can shift the surface chemistry enough to move the sodium leach rate, which shows up as a rising baseline or as sodium adducts in positive-mode ESI. A private label that switches factories between lots (which happens, because the private label is buying capacity, not a product) can hand you a lot with a different leach profile and no paper trail explaining why. The Thermo Fisher chromatography sample handling documentation is blunt about this: the vial is part of the sample path, and the sample path has to be controlled.
I have watched a lab run a 6-month stability study on a private label vial and get a 0.8% RSD drift in the late-eluting peak area across the study. They re-ran the same samples in OEM vials from the same glass type and the drift dropped to 0.2%. The private label had changed factories at month 3. Nobody told them. That is the recipe problem in one sentence.
What actually changes when the brand changes but the factory does not
This is the question that matters most, and it is the one the industry answers least honestly. The factory is the same. The line is the same. The cavities are the same. The label is different. What is different?
Four things, and they are all in the quality system, not the product:
- The process control plan. The OEM product has a control plan that specifies which dimensions are checked, at what frequency, with what gauge, and what the reaction is when a check fails. The private label product may have a control plan, or it may have a final inspection AQL and nothing else. The difference between 100% in-process gauging and a 0.65 AQL final inspection is the difference between catching a drift at 200 vials and catching it at 20,000.
- The change notification clause. OEM agreements typically require 30 to 90 days’ written notice before a process change, a tooling refurbishment, or a glass batch supplier change. Private label agreements often have no notification clause at all, because the private label is not the customer of record for the process.
- The certificate of analysis content. An OEM CoA typically carries the actual measured values for the lot: dimensions, hydrolytic resistance class, and often a lot-specific extractables or UV transparency number. A private label CoA often carries nominal values with a tolerance, which tells you nothing about whether the lot was in the middle or at the edge of the tolerance.
- The traceability chain. When an OEM lot fails, the brand owner can pull the tooling record, the glass batch record, and the annealing log. When a private label lot fails, the private label calls the factory, and the factory decides how much to share.
None of this shows up on a datasheet. All of it shows up when a lot fails.
A comparison table that procurement can actually use
The table below is what I would put in front of a sourcing committee. It is not about price. It is about who holds which lever.
| Item | OEM (brand owns tooling) | Private label (factory owns tooling) |
|---|---|---|
| Tooling ownership | Brand, serialized, in controlled crib | Factory, shared across customers |
| Drawing revision control | Brand-controlled, change notice required | Factory-controlled, notice optional |
| Process control plan | Documented, in-process gauging, typically 100% on neck ID and shoulder | Often final AQL only (0.65 to 1.5) |
| Glass batch traceability | Lot-level, tied to hydrolytic resistance test | Lot-level if asked, often not retained |
| CoA content | Measured values per lot | Nominal values with tolerance |
| Change notification | 30 to 90 days written, contractual | None unless negotiated |
| Factory change between lots | Prohibited without requalification | Common, often undisclosed |
| Liability on a failed lot | Brand owner, with recourse to factory | Private label, with limited recourse |
| Typical dimensional capability (Cpk on neck ID) | 1.33 to 1.67 | 0.9 to 1.2 (varies by cavity) |
| Seal integrity after 3 punctures (11 mm crimp, PTFE/silicone) | Holds to 1.5 bar, <2% leak rate | Holds to 1.5 bar, 4 to 9% leak rate on some lots |
The seal integrity row is the one that gets people’s attention. A crimp seal that leaks at 5% is a 5% injection failure rate, and that shows up as a run that has to be re-prepped. The Restek technical resources on crimp and septum selection are the best public reference I know for why the seal is a system (vial finish + septum durometer + crimp torque), not a vial property alone.
Field case 1: the private label that passed the print and failed the tray
Symptom: A pharma QC lab running a 96-position tray on a Waters Alliance autosampler got intermittent “vial missing” errors on positions 60 to 96, roughly 8% of the tray, starting with a new private label lot. The previous lot, same private label, same part number, ran clean.
Measurement: The lab pulled 20 vials from the failing lot and 20 from the passing lot and measured shoulder height with a digital height gauge (0.001 mm resolution) referenced to the base. The passing lot was 28.42 mm ± 0.05 mm. The failing lot was 28.81 mm ± 0.09 mm. The print did not specify shoulder height, only overall height (32.0 mm ± 0.3 mm), which both lots met.
Fix: The lab went back to the private label with the measurement. The private label went back to the factory. The factory had switched cavity sets three months earlier, and the new cavities had a 0.4 mm shoulder offset. The private label had no process control plan that would have caught it. The lab moved the product to an OEM source with a documented shoulder-height control at 28.40 mm ± 0.10 mm and the errors stopped. The lesson is not “private label is bad.” The lesson is that the print did not control the dimension that the autosampler cared about, and only the OEM relationship gave the lab a way to get that dimension into the specification.
Field case 2: same factory, different annealing, different baseline
Symptom: A cannabis testing lab running residual solvents by headspace GC-FID saw a rising baseline and a 0.6% RSD increase in the internal standard response across a 40-vial sequence, starting with a new case of vials. Same factory as the previous case. Same private label. The vials were visually identical.
Measurement: The lab ran a blank-vial extractables screen: 20 vials from the new case and 20 from the old case, filled with LC-MS grade methanol, capped, and heated to 80 °C for 30 minutes. The new case showed a 2.3× higher total ion count in the 40 to 120 m/z range than the old case. The UV transparency at 210 nm was 0.041 AU on the new case versus 0.018 AU on the old case.
Fix: The factory confirmed it had changed the annealing lehr profile to increase line speed. The private label had not been notified, because the private label agreement had no change notification clause. The lab switched to an OEM vial with a contractual change notification and a lot-specific UV transparency value on the CoA. The baseline stabilized. The U.S. EPA SW-846 test methods framework is useful here as a reminder that method performance is tied to the sample container, and a container change is a method change.
The decision rule I actually use
If the method is validated, if the analyte list includes anything at trace level, or if the lab runs more than about 2,000 injections per month on a single method, buy OEM. The cost premium is typically 15 to 40% per vial, and it buys you a tooling record, a change notification clause, and a measured CoA. That is not a marketing argument. It is the difference between catching a drift at 200 vials and catching it at 20,000.
If the method is a screening method, if the analytes are at ppm or above, or if the lab is running fewer than 500 injections per month on a method that gets re-validated quarterly, private label is defensible, provided you do three things: demand a lot-specific CoA with measured values, run an incoming extractables check on the first case of every new lot (20 vials, methanol, 80 °C, 30 min, UV at 210 and 254 nm), and write a change notification clause into the purchase agreement even if the private label says they cannot control the factory. If they cannot sign a notification clause, they cannot be your supplier for a validated method.
Incoming check that takes 40 minutes and catches most of it
- Pull 20 vials from the case. Measure overall height, neck ID, and shoulder height with a digital gauge. Compare to the previous lot. A shift of more than 0.15 mm on any dimension is a flag.
- Fill 10 vials with LC-MS grade methanol, cap with the septum you will use in the method, heat to 80 °C for 30 minutes, cool, and read UV at 210 nm and 254 nm against a methanol blank. A reading above 0.030 AU at 210 nm on a borosilicate vial is worth a conversation with the supplier.
- Puncture the septum 3 times with the method needle, then pressure-test the seal at 1.5 bar for 30 seconds. More than 2% leak rate on a 20-vial sample is a fail.
- Run a 6-injection blank sequence on the method and check the baseline drift and the internal standard RSD. More than 0.5% RSD on the IS across 6 blanks is a flag.
That checklist does not replace a full qualification, but it catches the two failure modes that account for most of the OEM versus private label pain: a tooling change that moved a dimension the print does not control, and a recipe change that moved the extractables profile. Both are invisible on the datasheet and both show up in the first 40 minutes if you look.
What the label on the box does not tell you
The brand on the box is a claim about who is responsible for the process. It is not a claim about which factory made the glass. A private label can be made in the same factory as an OEM product, on the same line, and still be a different product, because the process control, the change notification, and the traceability are different. Conversely, an OEM product can be made in a factory you have never heard of, and still be a controlled product, because the brand owner owns the tooling and signs the change control.
If you want a broader view of how the vial sits in the consumables supply chain, the chromatography consumables and lab vials references are useful for cross-checking part numbers and finishes across brands. For a specific look at how tooling and recipe control are documented at the factory level, the Aijiren brand technical pages are one of the few public sources that show the process control plan alongside the drawing, which is exactly the document that separates an OEM product from a private label one. And the main HPLC vial reference on this site covers the dimensional and material specs you will need to write the incoming check.
The decision rule is simple. If the vial is part of a validated method, buy the tooling record, not the glass. The glass is the same. The tooling record is what you are actually paying for, and it is the only thing that changes when the brand changes but the factory does not.
References & further reading
Related resources from our lab consumables network:
- Hplc Vial at vialforhplc.com
- Chromatography Consumables at chromatographyconsumables.com
- Lab Vials at labvials.com
- Aijiren Brand at aijirenvials.com
External references and standards: