A 12,000-vial run of 16 mm COD tubes came off the molding line at 2:14 AM with a 0.7 mm wall eccentricity on the shoulder radius. The shift lead caught it at the third hourly audit, not at final inspection. By then 3,400 vials were already packed. That is the difference between a factory that controls the line and one that inspects at the end. This article maps how a cod vial is manufactured and what the factory controls, station by station, parameter by parameter, with the actual numbers a buyer should ask for.
The preform oven is where the factory controls wall distribution, not just temperature
Most people assume the critical variable in cod vial manufacturing is the glass or polymer chemistry. It is not. The critical variable is wall distribution. A Cod Vial for chemical oxygen demand testing has to survive a 150°C digestion block for two hours, then cool to room temperature without cracking, while holding a 2 mL to 3 mL sample with a mercury sulfate and dichromate reagent load. The thermal stress concentrates at the shoulder and the base corner. If the wall is 0.1 mm thinner there, the vial fails at hour one, not hour two.
The preform oven controls this before the blow mold ever closes. For PET vials, the preform is injection molded as a thick-walled tube, then reheated in a multi-zone infrared oven. The oven has four to six zones, each with independent quartz heater banks and thermocouple feedback. The control loop holds each zone to ±2°C, but the real control is the temperature profile across the preform length, not the absolute setpoint. A preform that enters the blow station at 105°C at the neck, 112°C at the body, and 118°C at the base will stretch differently than one with a flat profile. The factory controls this by adjusting lamp power per zone, not by changing the oven air temperature.
I have walked lines where the oven setpoint was 115°C and the actual preform surface temperature, measured with an infrared pyrometer at the transfer point, varied from 101°C to 124°C across a single cycle. That line produced vials with a 12% wall thickness coefficient of variation. A properly controlled line holds that CV under 4%. The difference is not the oven brand. It is whether the factory treats the oven as a closed-loop system with preform surface temperature feedback, or as a box with a dial.
Blow molding closes the loop on stretch ratio and pre-blow timing
The blow molding station is where the factory controls the axial stretch ratio and the radial stretch ratio. For a 16 mm diameter cod vial with a 60 mm body length, the axial stretch ratio is typically 2.2 to 2.6, and the radial stretch ratio is 3.0 to 3.5. These numbers determine the crystallinity and the residual stress in the wall. If the axial stretch is too low, the vial wall stays amorphous in patches and the vial shrinks when it hits the digestion block. If the radial stretch is too high, the wall thins at the base corner and the vial cracks under the thermal gradient.
The pre-blow timing is the parameter that separates a controlled line from an uncontrolled one. Pre-blow is the low-pressure air pulse that pushes the preform into a bubble before the high-pressure final blow. On a reheat stretch blow machine, pre-blow pressure is usually 6 to 10 bar, and pre-blow duration is 0.08 to 0.25 seconds. The factory controls this with a proportional valve and a pressure transducer in the blow cavity, not with a timer alone. A 0.02 second shift in pre-blow timing changes the wall thickness at the shoulder by 8 to 12% on a 16 mm vial. That is enough to move a vial from passing to failing a 150°C thermal shock test.
What the factory does not control at this station: the final wall thickness at every point. That is only checked at the end, on a sampling basis. The blow mold process is closed-loop on pressure, timing, and preform temperature, but open-loop on the resulting wall distribution. A factory that claims full closed-loop control of wall thickness on a reheat stretch blow line is either running a very expensive in-mold sensor array or is not telling you the truth.
The neck finish station controls thread form and sealing surface flatness with hard gauges
The neck finish is where the cod vial interfaces with the cap and the septum. This is a dimensional control problem, not a thermal one. The thread form on a GPI 18-400 or 20-400 neck has a pitch, a flank angle, and a root radius. The factory controls these with a go/no-go thread ring gauge and a profile projector check every two hours. The sealing surface, the flat ring at the top of the neck where the septum sits, is controlled to a flatness of 0.05 mm total indicated runout. That is a tight number, and it matters because a 0.1 mm dip in the sealing surface lets the septum tilt under cap torque, and the vial leaks during the two-hour digestion.
The cap torque is set at the capping station, not at the neck finish station, but the two are linked. A cod vial factory controls the capping torque to 1.2 to 2.0 N·m for a 20-400 cap with a PTFE-lined septum. Below 1.2 N·m, the seal is not reliable under the vapor pressure build-up at 150°C. Above 2.0 N·m, the cap can crack the neck or the septum can extrude into the vial. I have seen a run where the capping torque drifted from 1.5 to 2.4 N·m over a shift because the torque driver’s clutch wore. The failure showed up as septum extrusion in 7% of the vials, caught only at final leak testing. The fix was a torque audit every 30 minutes, not every shift.
Leak testing is a 100% inline station, not a sampling station
Every cod vial factory that supplies to a regulated lab runs a 100% leak test. The method is usually vacuum decay or pressure decay. For a 16 mm vial with a 20-400 neck, the factory pulls a vacuum of -80 kPa on the sealed vial and measures the pressure rise over 3 to 5 seconds. A leak rate above 0.005 sccm fails the vial. That is the number to ask for. Some factories use a water immersion bubble test, but that is slower and does not catch micro-leaks below 0.01 sccm reliably.
The leak test station is closed-loop on the test parameters. The vacuum level, the test time, and the pass/fail threshold are set in the machine controller and locked with a password. The operator cannot change them. What the operator can change is the sealing force on the test head. If the test head presses too hard on the vial neck, it can temporarily seal a leak that would open under normal cap torque. A controlled factory verifies the test head force with a load cell at the start of every shift, and the force is set to match the actual capping torque, not to exceed it.
One field case: a lab reported intermittent leaks on cod vials after a supplier change. The vials passed incoming inspection at the lab, but leaked during digestion. The root cause was the factory’s leak test head force. It was set to 35 N, while the actual cap torque produced a sealing force of 18 N. The factory’s leak test was passing vials that would leak in the lab. The fix was to recalibrate the test head force to 18 N and revalidate the leak test. The factory lost two weeks of production on that line.
The annealing station controls residual stress, and most factories skip it
This is the station that separates a cod vial factory from a generic plastic vial factory. After blow molding, PET vials have residual stress locked into the wall. The stress comes from the rapid cooling in the mold. If the vial goes straight from the mold to the packing line, that stress releases slowly over the next weeks. The vial can shrink by 0.2 to 0.5% in the body diameter, and the neck can go out of round. For a cod vial that has to fit a 16 mm digestion block hole and seal against a septum, that is a failure.
The annealing station heats the vials to 65°C to 75°C for 20 to 40 minutes, then cools them at a controlled rate of 1°C per minute. The factory controls the oven temperature with a thermocouple array, and the cooling rate with a variable-speed conveyor. The result is a vial with residual stress below 2 MPa, measured by birefringence. A non-annealed vial typically shows 8 to 15 MPa. The difference shows up as dimensional drift in the first month after molding.
I have measured this. A lot of cod vials molded without annealing had a body diameter of 16.02 mm at day one and 15.88 mm at day 30. The digestion block hole was 16.10 mm. The vials still fit, but the wall-to-block contact changed, and the heat transfer rate dropped. The lab’s COD results drifted by 4% over that month, and nobody could explain it until we measured the vials. The fix was to specify annealing and to ask the factory for a birefringence reading on the first article inspection.
The printing and graduation station controls ink adhesion and volume accuracy
Cod vials carry a graduation mark, usually at 2 mL or 3 mL, and sometimes a fill line for the reagent. The factory controls the graduation position with a vision system that measures the distance from the neck sealing surface to the line. The tolerance is ±0.15 mm. That corresponds to a volume error of about 0.04 mL on a 16 mm diameter vial. For a COD test where the sample volume is 2.00 mL, a 0.04 mL error is a 2% bias in the result. That is outside the EPA method’s acceptance range for a standard.
The ink adhesion is controlled by a tape pull test and a solvent rub test. The factory applies a 3M 610 tape to the printed area, pulls it off at a 90° angle, and checks for ink transfer. The solvent rub test uses isopropanol or acetone, depending on the ink system, with 10 double rubs. The ink must not smear or lift. This matters because the vial sits in a 150°C block with acidic digestion fluid. If the ink softens at temperature, it contaminates the sample and the spectrophotometer reading.
A second field case: a lab switched to a lower-cost cod vial and saw a positive bias in their COD results. The bias was 6 to 8 mg/L on a 100 mg/L standard. The investigation found that the graduation line was printed with a solvent-based ink that leached trace organics into the digestion fluid. The ink passed the room-temperature solvent rub test, but failed at 150°C. The factory had never tested ink extraction at digestion temperature. The fix was to switch to a ceramic ink fired at 600°C, which costs more but does not leach.
Cleanroom packaging controls particulate load, and the spec is tighter than you think
The packaging station is where the factory controls particulate contamination. A cod vial for EPA method 410.4 or Standard Method 5220 goes into a cleanroom of ISO Class 7 or better for final packaging. The vials are washed with deionized water, blown with filtered air at 0.2 µm, and packed in sealed trays or bags. The factory controls the rinse water conductivity to below 1 µS/cm and the air particle count to below 3,520 particles per cubic meter at 0.5 µm or larger.
The particulate spec on the vial itself is usually fewer than 10 particles per vial at 10 µm or larger, measured by liquid particle counting after a rinse. That is a tight spec, and it is only checked on a sampling basis, not 100%. A factory that cannot hold this spec will produce vials that contribute background turbidity to the COD reading. The turbidity interference is small, but it matters when you are measuring a 5 mg/L COD standard and the blank is already 0.5 mg/L high.
The packaging station is also where the factory controls the seal integrity of the tray or bag. A vacuum-sealed bag with a leak lets moisture and particulates in during shipping. The factory checks the seal with a dye penetration test or a vacuum leak test on one bag per lot. The spec is no dye penetration beyond the seal edge and no pressure rise above 0.5 kPa over 30 seconds.
What the factory does not control, and where that leaves you
There are parameters that no cod vial factory controls in real time. The final dimensional stack, the full set of neck, body, and base dimensions, is checked on a coordinate measuring machine or a vision system at the end of the line, on a sampling basis. The chemical extractables profile is tested on a quarterly basis, not per lot. The long-term dimensional stability, the shrinkage over 12 months, is tested by accelerated aging at 50°C for 7 days, which approximates one year at room temperature. These are end-of-line or off-line checks, not closed-loop controls.
This is not a defect in the process. It is the reality of plastic vial manufacturing. The factory controls the process parameters that determine the final properties: preform temperature, stretch ratios, pre-blow timing, annealing temperature, capping torque, leak test parameters, and packaging environment. The final properties are verified, not controlled. A buyer who understands this distinction asks different questions. Instead of asking “what is your tolerance on the body diameter,” they ask “what is your preform temperature CV and your annealing profile.” The first question gets a spec sheet answer. The second gets a process answer.
Decision checklist: what to ask a cod vial factory before you qualify them
Here is the checklist I use when I audit a cod vial factory. It is specific to the manufacturing line, not to the QC lab.
| Parameter | What to ask for | Acceptable number | Why it matters |
|---|---|---|---|
| Preform surface temperature CV | IR pyrometer data from the transfer point, last 30 days | < 4% CV across the preform length | Wall distribution starts here |
| Pre-blow timing control | Pressure trace from the blow cavity, one cycle | Pressure rise within 0.05 s of setpoint | Shoulder wall thickness depends on this |
| Annealing profile | Oven temperature log and conveyor speed | 65-75°C for 20-40 min, cool at 1°C/min | Residual stress and long-term shrinkage |
| Capping torque | Torque audit data, last shift | 1.2-2.0 N·m, CV < 10% | Seal reliability at 150°C |
| Leak test threshold | Machine controller settings, locked | 0.005 sccm at -80 kPa, 3-5 s | Catches micro-leaks before shipping |
| Graduation position | Vision system data, last lot | ±0.15 mm from neck sealing surface | Volume accuracy, 2% bias risk |
| Ink extraction at 150°C | Test report, not just room temp rub test | No detectable organics by GC-MS | Prevents positive COD bias |
| Packaging particulate | Liquid particle count on rinsed vials | < 10 particles at 10 µm per vial | Background turbidity in the blank |
The line map: closed-loop vs end-of-line checks
A cod vial factory line has seven stations where the process is actively controlled, and four where the output is only checked at the end. The controlled stations are: preform oven, blow molding, neck finish, capping, leak testing, annealing, and packaging. The end-of-line checked parameters are: final dimensional stack, chemical extractables, long-term stability, and visual defects like scratches or bubbles.
The distinction matters because a factory can pass a final inspection and still ship vials that fail in the lab. The final inspection catches gross defects. The process controls prevent the subtle ones: the wall thin spot that cracks at hour two, the neck out-of-round that leaks under vapor pressure, the ink that leaches at digestion temperature. These are the failures that cost a lab a full day of digestion and a rerun of 20 samples.
For buyers sourcing cod vials alongside other Hplc Consumables, the same logic applies. A Vial Insert for an autosampler has its own process controls, and the same questions about closed-loop versus end-of-line apply. The 2Ml Vial that goes into an HPLC autosampler is manufactured on a similar reheat stretch blow line, with similar controls on preform temperature and annealing. The difference is the dimensional spec and the cleanliness level, but the process logic is the same.
The external references are useful for context. Agilent and Thermo Fisher publish specifications for their vial lines that show what a controlled factory can achieve. Restek has technical notes on vial contamination and extractables that explain why the ink and particulate controls matter. The U.S. EPA SW-846 methods set the performance requirements that the factory controls are designed to meet.
The decision rule
Qualify a cod vial factory on their process control data, not their spec sheet. Ask for the preform temperature CV, the pre-blow pressure trace, the annealing oven log, and the capping torque audit. If they cannot produce those, they are inspecting at the end, not controlling on the line. That is the factory that ships the 3,400 vials packed before the shift lead caught the wall eccentricity. The spec sheet will say 16 mm ±0.1 mm. The process data will tell you whether the vial holds that dimension after two hours at 150°C and two months on the shelf.
References & further reading
Related resources from our lab consumables network:
- Cod Vial at codvial.com
- Vial Insert at hplcinsert.com
- 2Ml Vial at 2mlsamplevial.com
- Hplc Consumables at hplcconsumable.com
External references and standards: