Product manufacturing uses a wide range of processing centers

A 40 mL volatile organic analysis (VOA) vial lot was rejected last month because the cap liner compressed 11% under a 22 in-lb torque setting, and the auditor wanted the raw extrusion data, not the certificate. That is the reality of EPA method requirements for vial closures and auditor expectations: the method text gives you a container spec, but the audit gives you a paper trail problem. Most labs pass the chemistry and fail the documentation. This article covers what SW-846 and related methods actually say about sample containers, what the auditor pulls first, and where closure selection quietly decides whether a data set survives review.

SW-846 is a compendium, not a single method. Chapter Two and Chapter Four set the container rules that apply across volatile, semivolatile, metals, and pesticide fractions. The specific method, say EPA 8260D or 8270E, inherits those rules. What the method text does not do is name a closure brand, a septum formulation, or a torque setting. It says “glass, with PTFE-lined septum” or “amber glass, screw cap, PTFE-faced silicone septum.” That is a material requirement, not an engineering spec.

Where auditors get traction is the data quality objective (DQO) process. If your laboratory’s quality manual says you will hold 8260D samples at 4 ± 2 °C for 14 days, the auditor will ask how you verified the closure maintains that hold time. The method says “analyze as soon as possible” but your SOP committed to a number. The closure is the variable that makes or breaks the number. I have seen a 14-day hold time survive with a 1.3 mm thick PTFE/silicone septum and fail in 9 days with a 0.9 mm septum from a different mold, same nominal material. The method never changed. The part did.

Auditors trained on the U.S. EPA SW-846 test methods for waste and materials framework will not accept “the vendor said it was compliant.” They want the link between the closure lot, the method requirement, and your in-house verification. That link is what this article maps out.

The three closure properties an EPA auditor can actually test against

An auditor cannot test everything. They focus on properties that are observable in your documentation and defensible against the method. Three stand out every time.

1. Septum integrity after puncture, not before

Most certificates of analysis (CoAs) report virgin septum properties: durometer, thickness, extractables profile. The auditor cares about the septum after 3 punctures, because that is the real sample condition. An autosampler needle at 23 gauge, 51 mm length, bevel tip, will core or tear a septum if the closure geometry is wrong. The failure shows up as a split in the PTFE facing layer after 5 punctures on a 9 mm short thread closure. The sample then breathes through the split. Benzene and toluene losses at 4 °C run 15 to 30% over 72 hours once the facing layer is compromised.

Ask your closure supplier for a puncture count rating at a specific needle gauge and entry angle. A competent molded septum should survive 5 punctures at 23 gauge without facing layer breakthrough. I have rejected lots that failed at 3. The auditor will ask for your puncture test record if you claim resealability in your SOP. If you do not have one, do not write “resealable” anywhere in your container procedure.

2. Closure torque and the compression set problem

EPA methods do not specify torque. They specify a leak-free seal. The auditor will ask how you know your cap was tight enough. The answer is a torque value, and the number that matters is the compression set of the liner at that torque. A silicone liner with a 25 Shore A durometer will compress 8 to 12% at 18 in-lb on a 24-414 closure. A PTFE-faced silicone liner with a harder core will compress less but may not conform to the glass rim at the same torque.

The failure mode I see most often in environmental labs is over-torquing. A technician cranks a cap to 30 in-lb because “tighter is safer.” The liner takes a permanent set. When the autosampler needle pierces, the septum has no recovery left. The puncture hole stays open. The sample off-gasses. The auditor sees a hold time study that passed at 18 in-lb and a standard operating procedure that says “tighten firmly.” That is a finding. Write the number. Verify it with a torque wrench. Keep the log.

3. Leachables that interfere with the target analyte list

SW-846 methods have target compound lists and tuning compounds. A closure that leaches siloxanes will wreck an 8270E semivolatile run. The classic offender is a low-durometer silicone septum without a full PTFE barrier. The siloxane peaks appear at m/z 73, 147, 207, and 281. Your 8270E tuning check for DFTPP will pass, but your BFB tune for 8260D will fail if the siloxane background is high enough. The auditor does not need to see the chromatogram. They need to see that you selected a closure with a documented leachables profile and that the profile covers the method’s target list.

This is where a vial closures specification sheet earns its cost. A proper sheet lists the facing material, core material, Shore A durometer, thickness tolerance, and an extractables test method, typically a reflux in methylene chloride or a headspace scan. If the sheet only says “PTFE/silicone,” you are buying a commodity, and the auditor will treat it as one.

What the auditor asks for first: the documentation stack

I have sat through enough EPA and state-level audits to know the sequence. The auditor does not start with the closure. They start with the sample. But the closure documentation is the second or third request, and labs that are not ready lose time and credibility. Here is the stack an auditor will pull, in order.

• Container and closure certification for the lot in use. This is the vendor CoA plus your incoming inspection record. The CoA must state the material grade, the facing layer, and the lot number. Your incoming record must show you checked something measurable: thickness, durometer, visual defects, or a seal test.
• Hold time study data. If your SOP claims a 14-day hold for 8260D, the auditor wants the study that proves it. The study must use the same closure lot and the same torque setting as routine operation. A hold time study run with a different closure is not defensible.
• Blank and matrix spike records that show closure background. The auditor will look at your method blanks over the last 12 months. A closure that leaches phthalates or siloxanes will show up as a rising blank trend. You need to be able to explain any trend as something other than the closure, or you need to show the corrective action.
• Puncture and reseal verification, if claimed. If your procedure says the closure is resealable, the auditor asks for the test. No test, no claim. Remove the word from the SOP.
• Supplier audit or qualification record. For a state or federal lab, this is standard. For a commercial lab, it is increasingly common. The auditor wants to know you did not buy closures from a random distributor with no traceability.

The weakest link in that stack is almost always the hold time study. Most labs buy a closure, run one study five years ago, and never re-verify when the vendor changes the mold or the material. The auditor knows this. When a vendor changes a mold cavity, the septum thickness can shift by 0.1 mm, and the compression set changes enough to break a 14-day hold. Your CoA will not flag it. Your incoming inspection will, if you measure thickness.

Hold time failures trace back to closure geometry more often than chemistry

Here is a field case from a commercial lab running 8260D on groundwater samples. The lab switched from a 24-414 closure with a 1.27 mm thick PTFE/silicone septum to a lower-cost 24-414 closure with a 1.02 mm septum. Same nominal material, same Shore A. Within two months, their 14-day hold time verification samples started failing for vinyl chloride and 1,1-dichloroethene at day 10. The lab blamed the purge-and-trap system. The real problem was the thinner septum. At 18 in-lb torque, the 1.02 mm septum compressed 14%, and the PTFE facing layer wrinkled. The wrinkle created a microchannel along the glass rim. The volatile compounds escaped slowly, and the two lightest analytes dropped below the acceptance limit first.

The fix was not a new purge-and-trap. It was returning to a 1.27 mm septum and adding a thickness check to incoming inspection. The lab now rejects any lot with a mean thickness below 1.22 mm. That is a number an auditor can see and verify. It is also the kind of number that separates a real closure specification from a catalog description.

Another case involved a metals lab running EPA 6020B on acidified water samples. The lab used a closure with a natural rubber core and a thin PTFE facing layer. Acid vapors from the preserved samples attacked the rubber through pinholes in the PTFE. The rubber leached zinc and iron into the samples. The lab’s method blanks were clean, because blanks were not acidified the same way. The matrix spikes were high, but the lab attributed it to the matrix. The auditor found the problem by comparing acidified blank results across three closure lots. One lot leached 4.2 µg/L of zinc into a 2% nitric acid blank after 7 days. The other two lots leached less than 0.3 µg/L. The difference was the PTFE facing layer thickness: 0.13 mm on the bad lot versus 0.25 mm on the good lots. The auditor cited the lab for not qualifying the closure for acidified matrices. The fix was a 0.25 mm minimum PTFE facing layer and a lot-specific acidified blank check.

Reading a closure CoA like an auditor reads it

Most people read a CoA top to bottom and check that the values fall within the listed ranges. An auditor reads it differently. They look for what is missing. Here is what I check on every closure CoA before it goes into a lab file.

• Lot number traceability to the shipping carton and the lab’s inventory log. If the lot number on the CoA does not match the box on the shelf, the CoA is worthless. This is a simple check that fails more often than it should.
• Facing layer material and thickness. “PTFE” is not enough. The CoA should state the facing layer thickness in millimeters or mils. For environmental methods, I want a minimum of 0.20 mm PTFE facing for semivolatile work and 0.25 mm for acidified metals work.
• Core material durometer and compression set. The durometer tells you how the septum will behave under torque. A 40 Shore A core will compress more than a 50 Shore A core. The compression set value, reported as a percentage after 22 hours at 70 °C, tells you how much recovery the septum has after compression. I reject anything above 15% compression set for a closure that will be punctured more than once.
• Extractables test method and result. The CoA should name the method, not just “pass.” A proper test is a 24-hour reflux in a specified solvent with GC-MS or LC-MS analysis. The result should be a total extractables number, typically below 0.5 mg per closure for environmental work.
• Sterility or cleanliness claim, if applicable. For 6020B trace metals, a clean closure matters. The CoA should state the cleaning process and the residual metals specification. If it does not, assume the closure is not clean enough for trace metals without a rinse.

When the CoA is thin, the auditor will ask for the raw data. If the vendor cannot supply raw data, the auditor will treat the CoA as a marketing document. I have seen a state auditor reject an entire closure lot because the vendor’s CoA said “meets EPA requirements” without naming a single method or number. That phrase is a red flag. It means the vendor has not done the work.

Closure selection decision table for EPA methods

The table below is the decision logic I use when specifying closures for environmental work. It is not exhaustive, but it covers the method families that generate the most audit findings.

The torque values are starting points. Your lab should verify them with a seal test on the actual glass vial and closure combination. A 24-414 closure from one manufacturer will not seat the same as a 24-414 from another. The thread pitch is nominally the same, but the liner thickness and the glass finish vary. The auditor will accept a torque value that is verified. They will not accept a torque value that is copied from a vendor catalog.

Incoming inspection: the part most labs skip and auditors notice

You cannot audit-proof a closure supply chain by reading CoAs. You have to measure the parts. The minimum incoming inspection for environmental closures is three checks, and each one takes less than five minutes per lot.

Thickness check. Use a calibrated micrometer or a drop gauge. Measure 10 septa per lot. Record the mean and range. Reject the lot if the mean is more than 0.05 mm below the specification or if any single septum is more than 0.10 mm below. This catches mold drift and mixed lots.

Durometer check. Use a Shore A durometer with a calibrated stand. Measure 5 septa per lot. Reject if the mean is more than 5 points outside the specification. A soft septum will over-compress and fail the hold time. A hard septum will not seal at the specified torque.

Visual check. Look for pinholes in the PTFE facing layer, delamination at the edge, and flash on the cap threads. A 10x loupe is enough. Pinholes in the facing layer are the most common defect I find, and they are invisible to the naked eye until the sample fails.

These checks do not replace a full qualification. They are the ongoing verification that the lot you received matches the lot that was qualified. When an auditor asks “how do you know this lot is the same as the one in your hold time study,” the answer is these records. Without them, the answer is “we trust the vendor,” and that is a finding.

Where the closure decision actually gets made: before the method, not after

Most labs pick a closure after the method is written. That is backwards. The closure should be part of the method development, because the closure determines the blank background, the hold time, and the puncture behavior. If you write the method with a generic “PTFE-lined cap” and then buy whatever is cheapest, you have created an audit exposure that no amount of documentation will fix.

The sequence that works is this. Pick the method and the target analyte list. Identify the closure properties that affect those analytes: leachables, puncture recovery, compression set. Select three closure candidates. Run a blank and a hold time study with each. Pick the one that passes with margin. Write the torque value, the septum thickness, the durometer, and the puncture limit into the SOP. Then buy that closure, and only that closure, for that method.

This is the same logic that applies to analytical testing consumables in general. The method and the part are one system. When you split them, the auditor finds the seam. For HPLC work, the same principle applies to HPLC vial selection: the closure is part of the injection precision, not an accessory to it. And across chromatography consumables, the parts that touch the sample are the method, whether the method text says so or not.

For GC and GC-MS methods, the closure interacts with the inlet and the column. A septum that sheds particles will contaminate the inlet liner and degrade the column. The Agilent chromatography consumables and columns documentation is a useful reference for inlet maintenance intervals, and the same logic applies to the sample closure upstream. A closure that fails under repeated puncture will send septum particles into the inlet, and the resulting ghost peaks will be blamed on the column. The Thermo Fisher chromatography and sample handling resources cover the sample path from vial to detector, and the closure is the first component in that path. For method-specific closure notes, Restek technical resources for chromatography publishes application notes that include closure recommendations for specific EPA methods. None of these replace your own verification, but they give you a starting point that is defensible in an audit.

Troubleshooting checklist for closure-related audit findings

When an auditor flags a closure issue, the response has to be specific and fast. Here is the checklist I use to diagnose and fix the problem before it becomes a formal finding.

• Rising blank background for siloxanes or phthalates. Check the closure lot number against the blank trend. If the trend started with a new lot, pull 10 septa and run a methylene chloride reflux extract. Compare total extractables to the qualified lot. Replace the lot if the number is more than 20% higher.
• Hold time failure on light volatiles. Measure septum thickness on the current lot. Check torque settings on the capping station. Run a leak test with a helium sniffer or a pressure decay tester. If the septum is thinner than the qualified lot, reject it.
• Puncture site leaking after autosampler injection. Check the needle gauge and the entry angle. A bent needle or a misaligned autosampler will tear the septum. Verify the puncture count against the vendor rating. If the septum is rated for 5 punctures and you are at 8, reduce the punctures per vial or switch to a pre-slit septum.
• Acidified metals blank contamination. Check the PTFE facing layer thickness. Run an acidified blank with the current lot and a retained sample from the qualified lot. If the current lot leaches, reject it and tighten the incoming inspection for facing layer thickness.
• Cap loosening during storage or shipment. Check the closure torque after 24 hours. A liner that takes a compression set will lose torque. If the torque drops more than 15% from the initial setting, the liner is failing. Switch to a higher-durometer core or a thicker liner.

Each of these fixes produces a record. The record is what the auditor wants to see. A fix without a record is a rumor. A fix with a thickness measurement, a torque log, and a new blank run is a corrective action.

The decision rule is simple. If your closure does not have a measured thickness, a verified torque, a puncture count, and a lot-specific CoA that names the extractables method, you do not have a compliant closure. You have a part that has not failed yet. For EPA methods, the auditor is not asking whether the closure works. They are asking whether you can prove it works, lot by lot, number by number. The lab that can answer that question with a micrometer and a torque wrench will pass. The lab that answers with a vendor’s marketing claim will not.

References & further reading

Related resources from our lab consumables network:

• Vial Closures at vialsclosures.com
• Analytical Testing at testinganalytical.com
• Hplc Vial at vialforhplc.com
• Chromatography Consumables at chromatographyconsumables.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