The auditor pulled one line from our extraction log: batch 2407-118, 12 of 20 vials, zero recovery of the 2,4-DNT surrogate. Not low. Zero. The vial was intact, the cap was on, the septum had already been punctured 14 times by the autosampler. That single line cost us a week of re-extraction and produced the first finding on a 40-item EPA method requirements for 2ml vial and auditor expectations checklist. What follows is what SW-846 and the related methods actually require of the container, and what an auditor asks for before they ever open the chromatograms.

I want to be precise here because most of the confusion in audits comes from people quoting the wrong document. SW-846 is not a single method. It is a compendium, and the container requirements live in a handful of places: SW-846 Chapter Three (inorganic analytes), Chapter Four (organic analytes), and then the method-specific tables in 8260, 8270, 8081, 8151 and the rest. The general guidance says amber glass for volatiles and semivolatiles, a minimum 2 mL volume for the small-volume headspace and purge-and-trap configurations, and a PTFE-lined septum. That is the entire extent of what most methods say about the vial itself.

What SW-846 does not say is the part that bites you. It does not specify septum thickness, durometer, or the number of punctures a septum tolerates. It does not specify cap torque. It does not tell you whether a screw-thread or a crimp closure is acceptable for a given method. Those decisions are yours, and the auditor knows it, which is why they ask for your documentation rather than the method text.

The three container lines that actually carry weight

Strip away the boilerplate and three requirements in the SW-846 text do real work in an audit:

• Preservation and hold time. 8260B and its successors set a 14-day hold from collection to analysis when the sample is preserved to pH < 2 with HCl and stored at 4 °C. The vial is the vessel that holds that condition. If your closure leaks, the pH drifts and the hold time is void.
• Headspace. For volatile methods, zero headspace is the requirement for the aqueous sample itself. A 2 mL vial filled to 2 mL with a 40 µL headspace is compliant. A 2 mL vial filled to 1.6 mL is not, and the auditor can see it in the fill photograph or the gravimetric fill check.
• Container cleanliness. The method assumes the vial contributes nothing. That assumption is testable, and auditors increasingly test it by asking for a method blank run against a fresh lot of vials.

Everything else, the septum durometer, the cap torque, the closure geometry, is a control you own. That is the whole game.

The paperwork an auditor asks for first, before the data

In my experience the auditor does not start with your chromatograms. They start with the container control file, because a data package is only as good as the vessel that produced it. Here is the sequence I have seen, in order, on every EPA-focused audit I have sat through.

1. The certificate of analysis, matched to the lot

They want the CoA for the specific lot number on the box in your stockroom, not a generic spec sheet. For a 2ml vial that means the CoA should state glass type (typically Type I borosilicate), the dimensional tolerance on the 12 × 32 mm envelope, the septum material and durometer, and the lot’s method-blank result. If the CoA is a photocopy with no lot number, that is finding number one.

2. The method blank against a fresh lot

This is the single document that most often fails. The auditor wants to see a method blank run with vials from the current lot, extracted through the full method, with no analytes above the reporting limit. If you have never run one, you are asking the auditor to trust the manufacturer. They will not.

3. The septum puncture and re-seal log

For any method where the autosampler re-samples the same vial (headspace, or a mid-sequence re-injection), the auditor asks how many punctures the septum is rated for and whether you track it. A standard PTFE-faced silicone septum in a screw-thread cap is good for roughly 10 to 15 punctures before the reseal degrades and volatiles start to escape. If your sequence re-injects a calibration standard 18 times, you have a documented failure mode and no log to show you caught it.

4. The storage and temperature record

4 °C ± 2 °C, logged. The vial is not the only thing in the chain, but the auditor will tie the storage record to the hold time. A gap in the log voids the hold time for every sample in the gap.

The septum is where the EPA method requirements and reality diverge

Here is my opinion, stated plainly: the septum is the weakest link in the entire 2 mL container, and it is the component SW-846 says the least about. The method tells you “PTFE-lined septum.” It does not tell you the durometer, the thickness, or the geometry of the cap that compresses it.

A standard 2 mL screw-thread vial uses a 9 mm septum, typically a bonded PTFE film over silicone, about 1.0 mm thick, with a silicone durometer in the 45 to 55 Shore A range. The PTFE face is what protects the sample from the silicone; the silicone is what provides the reseal after the needle withdraws. When you over-torque the cap, you compress the silicone past its elastic limit and the reseal is gone. When you under-torque, you get a leak path at the cap thread. The window is narrow, and it is measured in the cap, not the vial.

What torque actually does to the seal

For a 9 mm screw-thread closure, the practical torque window is roughly 0.4 to 0.6 N·m. Below 0.4 N·m the cap does not seat against the septum and you get a leak path. Above 0.6 N·m the septum deforms and the cap can back off as the silicone relaxes. If you are hand-tightening, you have no torque control and no way to prove it in an audit. A torque driver on the fill line is cheap; a failed batch is not. The same logic applies to the HPLC consumables side of the house, where a bad ferrule or a cross-threaded fitting produces the same class of leak with the same class of audit finding.

The crimp closure is not a loophole

Some labs switch to crimp caps thinking the closure is more reliable. It is, for long-term storage, because the crimp holds constant compression. But the crimp tool has its own torque equivalent: the crimp diameter. A 11 mm crimp cap wants a finished crimp diameter of about 7.6 to 7.8 mm. Outside that, the seal is either loose (leak) or crushed (septum damage). The auditor will ask for your crimp tool calibration just as readily as your torque driver.

The two field cases that fit this topic and nothing else

Both of these are from the volatile organics bench, because that is where container control is most visible to an auditor.

Case 1: the 14-puncture standard that read 60% low on the last injection

Symptom: a 20-injection calibration sequence using a single 2 mL standard vial. Injections 1 through 14 were within 5% RSD. Injection 15 dropped 18%, injection 16 dropped 40%, and by injection 20 the response was 60% low. The auditor flagged it as a possible integration error. It was not.

Measurement that identified it: we weighed the vial before and after the sequence. A 2 mL aqueous standard at 4 °C lost 0.9 mg over the 20 injections. That is small, but for a volatile analyte at the headspace partition, it is enough to move the response. The septum had been punctured 20 times, well past the 10 to 15 rating of the standard PTFE/silicone septum.

Fix: we split long calibration sequences across multiple vials, capped the puncture count at 10 per vial, and moved to a higher-durometer septum (55 Shore A) for the standard. The RSD across the 20 injections fell from 6.8% to 1.9%. The auditor closed the finding once we produced the puncture log and the pre/post weights.

Case 2: the lot that passed the CoA and failed the method blank

Symptom: a fresh lot of vials arrived with a clean CoA. The first method blank on the new lot showed 0.8 µg/L of methylene chloride against a reporting limit of 0.5 µg/L. The previous lot had run clean at <0.1 µg/L. The auditor saw the blank excursion and asked for the lot’s manufacturing documentation.

Measurement that identified it: we ran a solvent rinse on 10 vials from the lot and injected the rinse on the GC. The excursion was consistent across vials from the same carton and absent from a retained carton of the prior lot. That pointed to the carton, not the method. The cause was a packaging change: the vials had been packed in a carton lined with a solvent-based adhesive.

Fix: we added a lot-release method blank to our incoming QC, and we now require the supplier to state the packaging adhesive class on the CoA. The chromatography supplier we moved to supplies vials in adhesive-free cartons as standard. The auditor accepted the corrective action because we could show the retained-lot comparison.

The audit readiness checklist, with numbers

This is the checklist I would hand a lab manager the week before an EPA audit. Every line has a number attached, because “we follow the method” is not an answer an auditor accepts.

Two of those lines, the septum rating and the cap torque, are the ones I see missed most often. They are also the two that map directly to the failure modes the auditor is trying to rule out. If you want the vial itself specified correctly before you get to the audit, the OEM vial route lets you write the septum durometer and the cap geometry into the purchase spec rather than accepting whatever the catalog ships.

Where the auditor will push back, and how to answer

The pushback is predictable. It comes in three forms, and each has a defensible answer if you have done the work.

“Show me the container does not contribute.”

Answer with the method blank on the current lot, plus the retained-lot comparison if you have had an excursion. If you have never run one, run it before the audit and document the date. The Restek technical resources library has published method-blank protocols for volatile containers that you can adapt without reinventing the extraction.

“Show me the closure was intact at analysis.”

Answer with the puncture log and the pre/post weights for any vial that was re-sampled. A 2 mL aqueous vial should not lose more than about 0.5 mg over a 10-injection sequence. If it does, the septum is past its rating and the data is suspect.

“Show me the method you actually ran, not the method you cite.”

This is the one that catches labs that have drifted. The method on the report header says 8260D, but the sequence used a 15-puncture septum and a 1.6 mL fill. The auditor will reconstruct the run from the sequence file. Keep the sequence file, the fill check, and the torque log together in the same batch folder. The Thermo Fisher chromatography and sample handling documentation and the Agilent GC columns and accessories reference material both give you the autosampler parameters to cross-check against, which is faster than arguing from memory.

The decision rule I use for container control

If a 2 mL vial is going to be punctured more than 10 times, weighed before and after the sequence, and its lot released only after a method blank clears the reporting limit, then it is audit-ready. If any one of those three is missing, the vial is a finding waiting for an auditor to find it. The EPA method requirements for the 2ml vial are thin on purpose: the method tells you what the container must not do, and the audit asks you to prove it did not. Build the log, run the blank, cap the punctures, and the auditor’s first question answers itself.

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