Tooling ownership is the first fork in the road, and it is written on the

A buyer sent me two 4 L amber borosilicate bottles from the same Chinese factory last March. Same wall thickness, same GL45 thread, same 0.6 mm wall tolerance callout on the drawing. One carried a distributor’s private label, the other an OEM brand. The private-label bottle arrived with a certificate of analysis that listed the glass batch and the anneal cycle. The OEM bottle arrived with a spec sheet that did not name the batch, and when I asked the factory which drawing the OEM bottle was built to, the answer was “the customer’s drawing, which we do not have permission to share.” That single sentence is the whole OEM versus private label reagent bottle difference in miniature. The tooling, the recipe, the drawing, and the right to change any of them sit in different hands depending on which label is on the bottle.

Most procurement teams treat OEM and private label as synonyms with different artwork. They are not. They are two different ownership structures wrapped around the same molding machine, and the structure decides who controls the tooling, who controls the glass recipe, and what happens when the brand changes but the factory does not. This article is about that structure, not about price. If you want the pricing and contract terms, that is a separate conversation. Here I want to walk through the three things that actually move: tooling ownership, recipe control, and liability allocation, plus the field evidence that tells you which one you are standing in.

Tooling ownership is the first fork in the road, and it is written on the mold, not the PO

Every reagent bottle comes off a mold, and the mold has an owner. In a true OEM arrangement, the buyer commissions the mold or pays for the tooling amortization and, in the contract, takes title to it. The mold carries a cavity number, a steel grade (usually P20 or S136 for blow molds, hardened H13 for the neck ring and blow pin), and a tool life measured in hundreds of thousands of shots. When you own the tool, you can move it. You can pull it from one factory and drop it into another, and the bottle geometry does not change because the cavity did not change.

In a private-label arrangement, the factory owns the mold. You buy bottles off a stock cavity, and the only thing that is yours is the label and the carton. That sounds like a small distinction until the factory decides to retire the cavity, re-cut the neck ring, or move production to a sister plant. I have watched a private-label buyer discover that the GL45 thread on their “standard” bottle had drifted 0.3 mm in pitch diameter because the factory swapped a worn neck ring between lots. The caps still threaded on. The torque-to-seal curve did not, and the leak test at −80 kPa started failing on roughly one bottle in forty.

Here is the practical test. Ask the factory, in writing, three questions: who holds title to the mold, what is the cavity identification number, and what is the documented tool life remaining. A real OEM supplier answers all three. A private-label supplier answers the first with “we do” and the other two with silence or a deflection. If you cannot get the cavity number, you do not own the tooling, and you do not control the geometry of your own bottle.

What changes when the brand changes but the factory does not

This is the case that trips up almost everyone. You switch from a private-label supplier to an OEM supplier, or the reverse, and the bottles still come off the same line at the same plant. Nothing about the glass changed. What changed is who is allowed to sign the deviation. Under OEM, a dimensional drift on the neck finish triggers a deviation request to you before shipment. Under private label, the factory ships to its own internal AQL and you find out at incoming inspection, or worse, at the fill line. The bottle is identical. The information flow is not.

I have a field case for this. A pharma customer ran a private-label 250 mL amber bottle for a light-sensitive intermediate. Three lots in a row passed dimensional inspection. On the fourth lot, the fill line started seeing cap torque scatter: target 1.8 N·m, and the lot was running 1.4 to 2.4 N·m on a calibrated torque tester. The measurement that identified it was not torque, it was the neck finish. We pulled twenty bottles and measured the T-dimension (the thread root diameter) with a go/no-go gauge. Nine of twenty were at the low limit. The factory had re-cut the blow pin without telling the private-label buyer, because the buyer was not on the deviation list. The fix was to move the part to an OEM drawing with a named cavity and a deviation clause. Same factory. Different information rights. The torque scatter disappeared on the next lot.

Recipe control is where the glass stops being a commodity

A borosilicate bottle is not one material. It is a recipe: the SiO₂ content, the B₂O₃ fraction, the Na₂O and Al₂O₃ balance, the cullet ratio, and the anneal schedule. Type I borosilicate per USP <660> and ASTM E438 has a hydrolytic resistance class, and that class is a property of the recipe and the anneal, not of the label. Under OEM, the buyer often specifies the recipe down to the boron fraction and the anneal lehr profile, and the factory runs a batch record against it. Under private label, the buyer gets whatever recipe the factory is running that week, and the batch record is the factory’s property.

This matters most for two failure modes. The first is extractables. If you are running trace analysis on a bottle and the glass recipe shifts, the sodium and boron leach rate shifts with it, and your blank starts showing peaks you did not put there. The second is anneal. A bottle that is under-annealed carries residual stress, and residual stress shows up as a delayed crack at the shoulder after a few thermal cycles. I have seen a private-label lot fail a 121 °C autoclave cycle test at 15 psi with radial cracks at the neck, and the factory’s response was “the customer’s process is out of spec.” Under an OEM recipe with a specified anneal, that response is not available, because the lehr profile is on the batch record and the record says whether the bottle was annealed to the drawing.

Ask for the batch record. Under OEM you should get the glass batch, the cullet percentage, the lehr temperature profile, and the anneal point. Under private label you will get a certificate of analysis that states “meets Type I” and nothing else. That certificate is a claim, not a record. If your application is regulated under U.S. EPA SW-846 or a USP method, the difference between a claim and a record is the difference between a defensible audit trail and a finding.

The drawing is the contract, and only one party usually has it

In OEM, there is a controlled drawing with a revision letter, a tolerance stack on every critical dimension, and a change control procedure. The drawing is the contract. When the factory wants to change the parison weight or the blow time, it goes through a revision and you sign it. In private label, the “drawing” is often a catalog page and a sample bottle. There is no revision letter, and there is no procedure for changing it, because there is nothing to change.

I keep a one-page comparison for buyers, and it looks like this:

That table is the whole decision in one page. If any row matters to your application, you are an OEM buyer whether or not you have said so out loud.

Liability follows the drawing, and the label does not move it

When a bottle fails in the field, the question is who owns the failure. Under OEM, the answer is usually the party that owns the drawing, because the drawing defines what “conforming” means and the batch record proves whether the bottle conformed. Under private label, the answer defaults to the buyer, because the buyer accepted a sample and the sample is the only specification. The factory can always say “it matches the sample you approved,” and unless you have a measured drawing, you cannot rebut it.

I have watched this play out on a leachables complaint. A customer’s private-label 1 L bottle shed particulate into a solvent, and the particulate was traced to a flash line at the mold parting line. The factory’s position was that the sample bottle had the same flash line and the buyer approved it. The buyer had no drawing with a flash tolerance, so the complaint went nowhere. Under an OEM drawing with a stated flash limit (say 0.10 mm max at the parting line), the same bottle would have been a nonconformance with a clear owner.

This is not a legal argument, it is a specification argument. The label on the bottle does not change who is liable. The drawing does. If you want the factory to carry the risk, you have to give it a drawing to carry it against. That is the OEM structure, and it is the only structure where the liability question has a clean answer.

What actually changes on the floor when the brand changes but the factory does not

Here is the part most buyers miss. When you move from private label to OEM at the same factory, the bottles look identical and the price moves, and the floor does not notice for a while. Then three things change, in this order.

First, the incoming inspection plan changes. Under OEM you have a drawing, so you can write a real sampling plan against critical dimensions: neck finish, body diameter, height, wall thickness, and the cap torque-to-seal curve. Under private label you are inspecting against a sample, which means you are inspecting for “looks like the one we approved,” which is not a measurement. I have seen buyers run AQL 2.5 on a private-label lot and pass it, because the only defect class they could define was cosmetic.

Second, the deviation flow changes. Under OEM, a dimensional drift on the neck finish triggers a deviation request before the lot ships. Under private label, the lot ships and you find the drift at the fill line. The difference in detection time is the difference between a rework and a line stoppage.

Third, the change control changes. Under OEM, the factory cannot change the parison weight or the blow time without a drawing revision and your signature. Under private label, the factory changes whatever it wants to keep the line running, and you find out when the bottle stops fitting. This is the mechanism behind almost every “the bottle changed and nobody told us” complaint I have fielded in fifteen years.

A second field case: the cap that fit until it did not

A contract lab ran a private-label 500 mL narrow-mouth bottle for a GC prep method on a HPLC autosampler vial workflow that fed from the bottle. The caps sealed fine for eight months. Then a new lot arrived and the septa started coring on the reagent bottle neck, leaving silicone fragments in the solvent. The measurement that caught it was a puncture count on the septum: the old lot held 40+ punctures before coring, the new lot cored at 12. The factory had changed the septum supplier to a thinner gauge without a drawing revision, because there was no drawing. The fix was to move to an OEM cap with a specified septum thickness (3.0 mm nominal), a stated puncture rating, and a change control clause. Same factory, same bottle, different cap specification, and the coring stopped.

That case is why I tell buyers that the OEM versus private label reagent bottle difference is not a marketing distinction. It is the difference between a specification you can enforce and a sample you can only hope about.

How to tell which structure you are actually in, in under an hour

You do not need a factory audit to answer this. You need four documents and a caliper. Here is the checklist I run.

• Ask for the drawing. A real OEM supplier sends a controlled drawing with a revision letter and a tolerance on every critical dimension. If you get a catalog page, you are in private label.
• Ask for the cavity number. A real OEM supplier names the cavity and the tool life remaining. If the answer is “we will check,” you are in private label.
• Ask for the batch record. A real OEM supplier sends the glass batch, the cullet percentage, and the lehr profile. If you get a COA that says “meets Type I,” you are in private label.
• Measure the neck finish. Pull ten bottles and measure the T-dimension and the E-dimension on the GL45 thread with a go/no-go gauge. If the spread is wider than 0.2 mm, the cavity is worn and nobody is watching it.
• Run a torque-to-seal curve. Cap ten bottles to 1.8 N·m and leak-test at −80 kPa. If more than one in twenty fails, the neck finish or the cap liner is out of control.
• Ask who signs the deviation. If the answer is “we ship to our own AQL,” you are in private label, and the deviation is yours to find.

That checklist takes an hour and it tells you more about your supply than a factory tour. I have run it on suppliers that claimed OEM status and found private-label structure underneath, and I have run it on private-label suppliers that turned out to have a controlled drawing the buyer had never asked for. The label does not tell you. The documents do.

When private label is the right answer, and when it is a trap

I am not going to tell you OEM is always correct. It is not. If you are buying a 1 L bottle for a non-critical rinse solvent, running it through a sample vial prep, and the worst failure is a slow leak you catch at the bench, private label is fine. The bottle is a container, the specification is “holds liquid,” and the drawing would be overhead you do not need.

Private label becomes a trap the moment one of three conditions is true. First, the bottle touches a regulated method, where you need a batch record and a traceable drawing. Second, the bottle feeds an instrument, where a 0.3 mm neck finish drift stops the autosampler from picking up. Third, the bottle carries a solvent or a standard that would be expensive or dangerous to lose, where a leak is not a nuisance but an incident. In any of those three, the private-label structure leaves you holding a risk you cannot measure, and the OEM structure is the only one that gives you the measurement.

There is a middle path that some buyers use, and it is worth naming. You can buy private label and still impose an OEM-grade specification by writing a purchase specification with a controlled drawing, a named cavity, and a deviation clause, and requiring the factory to sign it. That converts a private-label part into an OEM part without moving the tooling. It works, but only if the factory will sign, and many will not, because signing means they own the deviation. When a factory refuses to sign a drawing, that refusal is your answer about which structure you are in.

The decision rule

If the bottle touches a regulated method, feeds an instrument, or carries a solvent you cannot afford to lose, buy OEM: own or amortize the tooling, get the drawing with a revision letter, get the cavity number and the batch record, and put a deviation clause in the contract. If the bottle is a rinse container and the worst failure is a slow leak at the bench, private label is fine and the drawing is overhead. The test is not the label on the bottle. The test is whether you can produce a controlled drawing with a revision letter and a batch record on request. If you can, you are in OEM structure regardless of what the supplier calls it. If you cannot, you are in private label regardless of what the supplier calls it, and the liability for every dimensional drift on that bottle is yours.

For buyers who want to go deeper on the consumables side, the Agilent chromatography consumables and Thermo Fisher sample handling documentation both publish dimensional tolerances on their OEM consumables that are worth comparing against your supplier’s drawing, and the Restek technical resources library has method notes on container and septum selection that show what a controlled specification looks like in practice. If you are sourcing the full workflow, the chromatography products catalog is a reasonable place to see how OEM and private-label parts are listed side by side, and the difference in the documentation is visible in the listing itself.

References & further reading

Related resources from our lab consumables network:

• Reagent Bottle at reagentbottleforsale.com
• Hplc Autosampler Vial at hplcautosamplervials.com
• Sample Vial at samplevialprice.com
• Chromatography Products at chromatographyselling.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