I watched a lot of 37-position trays come off a line in 2021 where the crimp height ran 0.15 mm high for about four hours before anyone caught it. The inline camera was reading seal diameter, not crimp height. The leak tester downstream only sampled one cavity in twelve. Nothing failed until a customer ran the vials on an autosampler with a short needle and started bending them at the neck. That is the whole lesson of a 2ml vial factory in one shift: what the line controls in the loop and what it merely inspects at the end are two different products, and the gap between them is where your rejects come from. This piece maps the line station by station and tells you which parameters are held closed-loop and which are only checked after the fact, so you can ask your supplier the right questions about how a 2ml vial is manufactured and what the factory actually controls.

A 2ml vial line is not one process. It is a chain of about nine stations, and the control philosophy is different at every one. Some parameters are held by a servo with a feedback loop that corrects every cycle. Others are set once at changeover and then trusted until the next audit. Knowing which is which tells you where the risk lives.

The tube forming station is the most tightly controlled part of the line. Glass tubing (typically borosilicate 33 expansion, sometimes 51 for higher thermal duty) is fed into a rotary machine where burners soften it and a mandrel forms the shoulder and neck. The parameters held closed-loop here are:

• Tube outer diameter (OD) and wall thickness at the infeed, measured by a laser micrometer every tube and used to trim the burner gas flow and forming time. A 0.05 mm drift in wall thickness changes how the shoulder forms, so this loop matters.
• Bottom forming time and blow pressure, held to a few percent so the base thickness lands inside its window.
• Neck inner diameter (ID), typically held to ±0.10 mm on a nominal around 6.0 mm for a 11 mm neck finish, because the septum and cap geometry depend on it.

What is not closed-loop at forming: the annealing profile and the residual stress pattern. Those are set by a lehr temperature curve and checked by polarimetry on a sample pull, not on every vial. If a lehr zone drifts, you get vials that pass dimensional checks and crack at the neck three weeks later. That is a downstream inspection problem, not a forming problem, and it is one of the reasons I tell buyers to ask for the lehr audit log separately from the dimensional report.

Where the annealing lehr sits in the control hierarchy

The lehr runs a temperature curve that peaks around 560 to 580 °C for borosilicate 33 and cools slowly enough that residual stress stays below the threshold where a vial survives a 1.5 m drop onto a hard floor. The setpoint is controlled. The actual stress in the glass is not measured inline. A polarimeter check on a sample every shift is the only feedback, and it is a spot check. I have seen a lehr with a failed thermocouple run 40 °C cold for a full day; the vials were dimensionally perfect and shattered in the customer’s hands. Nothing on the line caught it because nothing on the line was measuring stress.

The dimensional inspection gate is a filter, not a control

After forming and annealing, vials pass a dimensional gate. This is where a lot of people assume the factory is “controlling” the part. It is not. It is filtering it. The gate typically includes:

• Laser or camera measurement of neck ID, neck OD, and body OD.
• Height check at the shoulder and at the base.
• Concentricity of the neck to the body, usually held to 0.15 mm TIR or better.

These are 100% inspected on a modern line, but the measurement is a pass/fail gate, not a correction. If a vial is out, it is rejected. The line does not adjust itself because one vial drifted. The only closed-loop correction happens at forming, upstream. So when a buyer asks “how do you control neck ID,” the honest answer is: we hold it at forming with a servo and we verify it at the gate. The gate is the proof, not the control.

This distinction matters for procurement because it tells you where to push. If your application is sensitive to neck ID (say, a crimp cap that needs a consistent seal), the control you want is a forming-station SPC chart, not a final inspection report. The final report tells you what shipped. The SPC chart tells you whether the process was in control while it shipped.

Crimp geometry is the parameter most factories get wrong, and here is the number

Crimp height. I will die on this hill. The crimp on an 11 mm aluminum seal has a target height that most reputable suppliers hold to 6.95 to 7.05 mm for a standard 2ml vial with a 11 mm finish. The tolerance is tight because the seal has to compress the septum enough to hold pressure but not so much that it cores the septum or splits the aluminum. A crimp at 6.80 mm leaks. A crimp at 7.20 mm cores the septum on the first needle insertion.

Here is the problem: on many lines, crimp height is not closed-loop. The crimping head is set mechanically at changeover, and the height is checked by a gauge on a sample every 30 minutes. Between checks, the head can drift as the collet wears. A collet that has run 200,000 cycles can drift 0.08 mm, which is enough to move you from a good crimp to a marginal one. The factory that controls this has a load cell on the crimp head and a closed-loop servo that holds crimp force, not just height. The factory that only checks height at the end is trusting a mechanical setpoint to hold for 30 minutes of production.

Ask your supplier which one they have. If they cannot tell you, assume it is the second.

The septum compression ratio nobody puts on the datasheet

The septum is compressed between the crimp and the vial neck. The compression ratio (compressed thickness divided by free thickness) should land between 0.65 and 0.75 for a standard PTFE-faced silicone septum. Below 0.65, you get leaks under autosampler pressure. Above 0.75, you get coring and septum push-through. The factory controls this indirectly through crimp height and septum thickness, but neither is usually closed-loop on a standard line. This is why coring shows up as a “septum defect” when it is really a crimp geometry problem, a point the sibling article on coring covers in more depth.

Cleaning and depyrogenation: the station where closed-loop control actually exists

The wash and depyrogenation tunnel is the one part of a 2ml vial line where I see genuine closed-loop control on a routine basis, because the regulatory pressure forces it. A typical tunnel runs:

• Wash zones with recirculated filtered water at controlled temperature and conductivity.
• Rinse zones with water for injection (WFI) quality water, conductivity held below 1.3 µS/cm at 25 °C per USP <645>.
• Depyrogenation zone at 250 °C or higher for enough time to achieve a 3-log endotoxin reduction, validated per the cycle.

The tunnel temperature is closed-loop. The belt speed is closed-loop. The endotoxin reduction is validated once and then trusted. What is not closed-loop is the actual cleanliness of each vial; that is checked by sampling. A vial that comes out of a tunnel with a partially blocked nozzle can be visually clean and still carry particles. The factory controls the tunnel, not the vial. This is a subtle but important distinction for anyone specifying vials for HPLC consumables or trace analysis, where a single particle can ruin a run.

If you are buying vials for EPA SW-846 methods or any trace work, ask for the tunnel validation report and the particulate count on the finished vial, not just the tunnel setpoint. The U.S. EPA SW-846 test methods do not specify vial cleanliness, but your blank injections will tell you the truth.

What the leak and dimensional end-of-line test actually catches

End-of-line testing is where the factory finds out what it missed. A typical 2ml vial line runs:

• 100% leak test by pressure decay or vacuum decay, holding 0.5 bar for 3 seconds with a reject threshold around 0.02 mbar/s decay.
• Camera inspection for cosmetic defects (chips, cracks, stones in the glass).
• Sample-based dimensional audit on a pull of 5 to 30 vials per shift.

Here is the honest limitation: the leak test catches a bad crimp or a cracked neck. It does not catch a marginal crimp that will leak after the customer’s autosampler heats the vial to 40 °C. It does not catch a septum with a slow compression set. It does not catch a vial with residual stress that will crack in transit. Those are the failure modes that pass end-of-line and fail in the field, and they are exactly the ones that a closed-loop forming and crimping station would have prevented.

Two field cases where the line control gap caused the failure

Case 1: the crimp height drift that passed every leak test

Symptom: a customer running a 2ml vial on a GC autosampler reported that about 3% of vials leaked at the septum after two hours at 40 °C. The vials passed the supplier’s 100% leak test at ambient.

Measurement that identified it: we pulled 50 vials from three lots and measured crimp height with a calibrated gauge. The mean was 7.14 mm with a standard deviation of 0.06 mm. The upper spec was 7.05 mm. The line’s 30-minute crimp check had been reading 7.02 mm because the operator was checking the first vial off the head, which ran colder and crimped slightly lower than the steady-state vials. The leak tester passed them because a marginal crimp holds at ambient and fails at 40 °C.

Fix: the factory added a load cell to the crimp head and moved to closed-loop crimp force control, and changed the SPC sampling to pull vials from the middle of the run rather than the first off the head. Reject rate at the customer dropped to zero over the next six lots.

Case 2: the lehr thermocouple that ran cold for a day

Symptom: a lot of 2ml vials passed dimensional and leak testing, then arrived at the customer with a 6% breakage rate at the neck during unpacking. The vials looked fine. They broke when the customer’s techs snapped them into a tray.

Measurement that identified it: polarimetric stress measurement on 20 vials from the lot showed stress values roughly double the historical mean. The lehr log showed zone 3 running 40 °C below setpoint for about 18 hours. The thermocouple had drifted, and the lehr controller was compensating for a false reading.

Fix: the factory added a second thermocouple per zone with a cross-check alarm, and moved to a daily polarimeter check on a sample pull from each lehr zone rather than one check per shift on a single sample. The breakage rate went back to baseline within two lots.

Both cases share a pattern: the parameter that failed was not the one being controlled in the loop. The line was controlling something adjacent and inspecting the rest. This is the structural weakness of any 2ml vial line, and it is why the good factories invest in closing the loop on crimp force and lehr temperature rather than adding more end-of-line inspection.

What to ask a factory before you trust its line

You do not need to audit the whole plant. You need to know which parameters are closed-loop and which are checked at the end. A short list of questions gets you there:

• Which parameters have a feedback loop? Expect forming (neck ID, wall thickness) and the wash tunnel. If they say crimp height is closed-loop, ask for the load cell calibration record.
• What is the crimp height spec and how often is it checked? Expect 6.95 to 7.05 mm and a check every 30 minutes at minimum. Ask whether the check pulls from the middle of the run.
• How is residual stress measured, and how often? Polarimetry, daily per zone, is the answer you want. “Visual inspection” is not an answer.
• What does the leak test hold and for how long? Expect at least 0.5 bar for 3 seconds, with a documented reject threshold.
• Can I see the SPC chart for the last three lots? If they cannot produce one, the process is not in statistical control, and the inspection report is just a snapshot.

For OEM work, the questions get sharper. If you are buying under your own label through an OEM vial program, you want the line control data tied to your lot numbers, not a generic certificate. The 2ml vial you receive should be traceable to a forming run, a lehr zone, and a crimp head, and the factory should be able to show you the closed-loop records for each. If they cannot, you are buying inspection, not control.

The decision rule for line control

When you evaluate a 2ml vial supplier, separate the line into two buckets: parameters the factory holds with a feedback loop, and parameters it only checks at the end. The first bucket is where you get consistency. The second bucket is where you get rejects that pass the certificate and fail in your lab. A factory that can show you closed-loop records for forming, crimping, and lehr temperature, and that pulls its SPC samples from the middle of the run rather than the first vial off the head, is telling you the truth about its process. A factory that hands you a dimensional report and calls it control is telling you it inspected the part, not that it controlled it. Buy from the first, and specify your crimp height and residual stress limits in the purchase order so the line has a number to hold.

For a broader look at how these vials fit into sample handling and instrument consumables, the resources at HPLC consumables and the technical libraries at Agilent, Thermo Fisher, and Restek are worth reading alongside the line data, because the instrument side tells you which of these parameters actually bites in your method. If you are sourcing across a range of formats, a chromatography supplier that can show you the closed-loop records for each format is a better bet than one that only shows you the catalog.

References & further reading

Related resources from our lab consumables network:

• 2Ml Vial at 2mlsamplevial.com
• Hplc Consumables at hplcconsumable.com
• Chromatography Supplier at chromatographysupplier.com
• Oem Vial at oemsamplevials.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