What

I have sat on both sides of this problem. For eleven years I ran a contract analytical lab where I bought closures by the case, and for the last four I have helped a manufacturer sell them. The single most expensive mistake I see is not a bad batch or a supplier switch. It is a method team specifying a closure material because it is what the method was written on, without asking whether the chemistry of the analyte and the chemistry of the elastomer are compatible. When they are not, you do not get a clean failure. You get a slow drift in the calibration curve, a ghost peak at 4.2 minutes, or a blank that looks fine on Monday and fails on Friday.

This guide is about picking the material grade on purpose. I will give you the specific grades, the temperature limits, the solvent compatibility, the puncture counts, and the decision rules I actually use. I will also tell you where I think the industry over-specifies and where it under-specifies, because both cost you money.

A closure is two parts bonded together: the cap (usually aluminum, sometimes polypropylene) and the septum (an elastomer laminate). When people say “material grade,” they almost always mean the septum. That is the part in contact with your sample and the part the needle punches. But the cap matters too, and I will cover both.

Septums are not single materials. A typical HPLC septum is a laminate: a thin PTFE film on the sample side, bonded to a silicone or EPDM core. The PTFE is there for chemical inertness. The core is there for resealability. Understanding that sandwich is the whole game. If you pick a septum by its core material alone, you will get the inertness wrong. If you pick it by the PTFE face alone, you will get the reseal wrong.

The three grades you will actually encounter in a working lab are silicone/PTFE, EPDM/PTFE, and butyl (often butyl/PTFE). There are also molded one-piece septums (pure silicone, pure PTFE-faced silicone) used in specific niches. I will take each in turn and tell you exactly where it belongs.

Silicone (with PTFE face): the default, and rightly so

Silicone is the workhorse. It has the best resealability of the common elastomers, the widest temperature range, and it is the most forgiving if your needle is slightly off-center or dull. A 1 mm thick silicone core in a 11 mm closure will typically survive 50 to 100 punctures from a 22-gauge needle before coring becomes a problem, and a good 0.003 inch PTFE face keeps the silicone out of your sample.

Temperature rating for a standard silicone/PTFE septum is roughly -40 °C to +200 °C. That covers essentially every HPLC method you will run and most GC methods up to the point where you should be using a different injection technique anyway.

Where silicone belongs

  • Reversed-phase HPLC with aqueous and acetonitrile/methanol mobile phases. This is the majority of methods. Silicone/PTFE is the correct default and I see no reason to deviate.
  • Autosampler trays where the same vial is punctured repeatedly (a 96-position tray run overnight). Resealability matters more than anything else here.
  • Headspace GC, but only the PTFE-faced grade, and only with a proper crimp or screw torque. Silicone is permeable to small molecules, which is why headspace needs a thicker PTFE barrier or a different construction. More on that below.
  • Cold storage down to -40 °C. Silicone stays flexible. EPDM gets stiff. Butyl is fine but offers no advantage.

Where silicone does not belong

Silicone is a poor choice for trace analysis of volatile organics in headspace, and it is a poor choice for anything where you are chasing low-ppb detection of small, permeable molecules. Silicone absorbs and transmits. If you are running EPA method 8260 by headspace, the silicone core will let analytes migrate through the PTFE face over a long tray run. The fix is either a thicker PTFE barrier or a butyl-based septum.

I also would not put silicone in a method that uses dichloromethane or chloroform as the sample solvent for extended periods. The PTFE face protects the silicone from the liquid, but if the face is scratched by a needle, the solvent reaches the core and the silicone swells. Swollen silicone in a crimp cap means the seal is compromised and you will see it as a pressure error on the instrument before you see it as a data problem.

EPDM (with PTFE face): the underused middle ground

EPDM gets a bad reputation because cheap EPDM septums from the 1990s were terrible. Modern EPDM/PTFE laminates are a different material. EPDM has excellent resistance to aqueous solutions, acids, and bases, and it is far less permeable to water vapor than silicone. Its temperature range is narrower, roughly -40 °C to +120 °C, which is the main reason it does not get used more.

Where EPDM belongs

  • Aqueous and ion chromatography methods where you are worried about silicone extractables showing up as a baseline rise. EPDM extractables are different and often lower in the relevant mass range.
  • Methods with high-salt buffers (phosphate, citrate, ammonium acetate). EPDM handles these without the swelling you can get at the PTFE/silicone interface over long runs.
  • Basic pH mobile phases above pH 9, where silicone can degrade over time.
  • Short-duration sample storage where water loss is the concern. EPDM is the better barrier.

Where EPDM does not belong

Do not use EPDM above 120 °C. It will harden and lose resealability, and you will not notice until your autosampler starts throwing needle errors. Do not use it for repeated punctures beyond about 30 to 40 injections per septum; the reseal is worse than silicone. And do not use it for aggressive nonpolar solvents (toluene, hexane) because EPDM swells badly in those.

I think EPDM is genuinely under-specified in ion chromatography labs. If you are running a suppressed conductivity method with a carbonate/bicarbonate eluent, EPDM/PTFE is the better closure and almost nobody reaches for it.

Butyl (with PTFE face): the headspace and long-storage specialist

Butyl rubber is the lowest-permeability common septum material. It is the right choice when the analyte is small, volatile, and you care about it migrating out of (or into) the vial over hours. Temperature range is roughly -40 °C to +120 °C, similar to EPDM, and resealability is decent but not silicone-grade.

Where butyl belongs

  • Headspace GC, full stop. If you are running a static headspace method for residual solvents or volatiles, butyl/PTFE is the closure I would specify. The lower permeability keeps the headspace composition stable across a long tray.
  • Long-term sample storage (weeks to months) where you are tracking a volatile or semi-volatile analyte and cannot afford loss through the septum.
  • Pharmaceutical stability studies where the method is validated against a specific closure and you need the lowest permeability option that still meets USP <661> requirements.

Where butyl does not belong

Butyl is a poor choice for repeated punctures. It cores more readily than silicone, and the fragments are a real problem for a needle. I would not push a butyl septum past 20 to 30 punctures, and I would inspect the needle after every tray. Butyl is also not great for very cold storage; it stiffens below about -30 °C and the reseal suffers.

The comparison table I actually use

Here is the decision matrix I have on my wall. These are the numbers I would defend in a method review.

  • Silicone/PTFE: temp -40 to +200 °C, punctures 50-100, best for RP-HPLC and autosampler trays, poor for headspace volatiles, moderate cost.
  • EPDM/PTFE: temp -40 to +120 °C, punctures 30-40, best for IC, high-salt, and basic mobile phases, poor for nonpolar solvents, moderate cost.
  • Butyl/PTFE: temp -40 to +120 °C, punctures 20-30, best for headspace and long storage, poor for repeated injection, highest cost of the three.

If you want the one-line rule: silicone for injection, EPDM for salts and bases, butyl for headspace and storage. That covers 90% of methods and it is the rule I give new method developers.

The cap: aluminum, polypropylene, and the torque question

Most of the material-grade conversation is about the septum, but the cap grade matters for two reasons: chemical compatibility with your autosampler and the seal you actually achieve.

Aluminum crimp caps are the standard for HPLC and GC. They are dimensioned to the 11 mm crimp standard and, for larger formats, the 20 mm crimp standard. The critical spec is the crimp height and the seal. A properly crimped 11 mm aluminum cap on a 12 x 32 mm vial should be crimped to a height of about 5.5 to 6.0 mm with a crimper set to the correct jaw. Under-crimp and you get leaks. Over-crimp and you deform the septum and lose resealability. I have seen labs run 200 samples with a mis-set crimper and wonder why their RSD jumped from 0.8% to 3.5%. The crimper, not the septum, was the problem.

Polypropylene screw caps are the practical choice for autosamplers that cannot crimp, and for methods where you need to open and reseal the vial. They mate to the GPI 13-425 thread for 4 mL vials and GPI 8-425 for 2 mL vials (the GPI thread standards are defined by the Glass Packaging Institute). The torque spec I use is 1.5 to 2.0 inch-pounds for a 13-425 PP cap with a PTFE-faced septum. Below 1.0 inch-pound you get leaks under headspace pressure. Above 2.5 inch-pounds you start to deform the septum and the reseal fails. A torque screwdriver is cheap. Buy one.

One more cap note: for headspace, use a magnetic cap or a crimp cap with a thicker septum, and check that the cap is rated for the headspace pressure your instrument uses (typically 10 to 20 psi). A standard screw cap with a thin septum will leak at the top of that range. This is the kind of thing that shows up as an intermittent autosampler error, not as a data failure, so it is easy to miss.

What the standards actually require

If you are buying for a regulated method, the closure has to meet the dimensional and extractables requirements of the method and the pharmacopeia. The relevant ones I work against:

  • USP <661> covers plastic packaging systems and their materials of construction, including extractables testing for elastomeric closures. If you are doing a stability study, the closure has to be qualified against this.
  • USP <381> covers elastomeric closures for injection. Relevant if your vial is a parenteral container.
  • ISO 8362-2 defines the dimensions of injection vial closures. If you are sourcing 13 mm or 20 mm injection closures, this is the dimensional reference.
  • EPA SW-846 is the umbrella for the waste-testing methods (including 8260 for volatiles by headspace or purge-and-trap). The closure choice for 8260 is butyl/PTFE, and the method’s QA/QC requirements are the reason.

For method development outside a regulated environment, you can be less formal, but I would still keep a closure qualification record. A one-page document that says “we tested three grades, here are the blank chromatograms, here is the decision” will save you a week when someone asks why you are not using the cheap silicone septum.

The extractables question, in plain terms

Every septum sheds something. The question is whether what it sheds lands in your chromatogram. Silicone sheds low-molecular-weight siloxanes. EPDM sheds antioxidants and curing agents. Butyl sheds a different set of additives. The PTFE face is there to keep those out of the sample, and it works as long as the face is intact and the sample does not sit against the cut edge of the laminate.

The practical test is a blank injection series. Fill ten vials with your mobile phase (or your sample matrix minus analyte), cap them with the candidate closure, and run them through your method. If you see a peak that grows over the tray run, the closure is contributing. If the peak is present in vial one and gone by vial ten, it was a surface contaminant and a pre-rinse fixes it. If it grows, change the grade.

I have run this test on maybe forty closure lots over the years. The failures cluster in two places: cheap silicone with a thin or poorly bonded PTFE face, and butyl that has been stored past its shelf life and is off-gassing. Both are avoidable with a supplier who will give you a lot-specific extractables report.

Temperature, pressure, and the numbers that matter

Two operational parameters decide whether a closure survives your method: the maximum temperature the septum sees and the pressure differential across it.

For HPLC, the septum sees ambient temperature and essentially no pressure differential. Any of the three grades works, and you should pick on extractables and resealability. Silicone wins.

For GC, the inlet is hot (250 to 300 °C) but the septum is in the inlet, not the vial. The vial closure sees only the autosampler temperature, usually 4 to 40 °C. So the temperature rating of the vial septum is almost never the limiting factor in GC. What matters is permeability and coring.

For headspace, the vial is pressurized to 10 to 20 psi and heated to 40 to 90 °C. Now both parameters bite. A silicone septum at 80 °C and 15 psi will pass volatiles and may creep. Butyl/PTFE is the correct grade, and I would not argue about it.

For cold storage at -80 °C (some biobanking methods), only silicone stays flexible. EPDM and butyl both stiffen and the reseal after thaw is poor. If you are storing at -80 °C and then injecting, use silicone/PTFE and accept the permeability trade-off, or use a sealed ampoule instead.

A troubleshooting checklist for closure-related method problems

When a method misbehaves and you suspect the closure, work through this in order. It is the list I use.

  • Ghost peaks that grow across a tray run: extractables or permeability. Run the blank series above. Switch grade if the peak grows.
  • RSD above 2% on a method that used to be under 1%: check the crimper or the torque first, not the septum. A 0.5 mm crimp height error will do this.
  • Intermittent autosampler pressure errors: cap not sealing. Check torque (1.5 to 2.0 inch-pounds for 13-425), check crimp height, check for a swollen septum.
  • Coring fragments on the needle: too many punctures or the wrong grade. Count your punctures. If you are above 50 on silicone or 30 on butyl, change the septum.
  • Baseline drift on IC: silicone extractables. Switch to EPDM/PTFE.
  • Loss of volatile analyte over storage: permeability. Switch to butyl/PTFE and check the PTFE face thickness.
  • Septum stuck to the vial after storage: the PTFE face has bonded to the glass, usually from over-torque or heat. Reduce torque and check the storage temperature.

If you work through that list and the problem persists, the closure is probably not the cause and you should look at the HPLC vial itself (dimensional tolerance, glass grade) or the instrument. I have seen more “closure problems” turn out to be vial problems than the other way around.

Where to buy, and what to ask for

The closure supply chain has three tiers: the elastomer manufacturer, the laminator and die-cutter, and the distributor. The grade you get is set at the laminator. Ask your supplier for the following, in writing, for each lot:

  • The septum construction (core material, face material, face thickness in inches or mm).
  • The temperature rating and the recommended puncture count.
  • A lot-specific extractables report or a certificate of analysis that references the test method.
  • The dimensional spec against the relevant standard (11 mm crimp, 13-425, 20 mm crimp, ISO 8362-2).

If a supplier cannot give you the face thickness, walk away. That number is the difference between a septum that protects your sample and one that does not. A 0.003 inch PTFE face is standard. A 0.001 inch face is a cost reduction that shows up in your blank chromatograms.

For sourcing, I keep a short list. The major instrument vendors publish closure compatibility against their autosamplers, which is useful when you are matching a closure to a specific tray: Agilent’s GC columns and accessories catalog and Thermo Fisher’s chromatography and sample handling range both list the recommended closure for each vial format. For method-specific guidance, Restek’s technical resources are good on headspace and volatiles, and the EPA SW-846 methods are the reference for the waste-testing work. For the broader consumables picture, chromatography consumables and analytical testing resources are worth bookmarking, and the full closure range is at vialsclosures.com.

My recommendation, by method

If you want the short version, here it is, with the numbers attached.

Reversed-phase HPLC, ambient autosampler, 22-gauge needle, under 50 injections per vial: silicone/PTFE, 11 mm crimp, 1 mm core, 0.003 inch PTFE face, crimped to 5.5 to 6.0 mm. This is the default and it is correct.

Ion chromatography or high-salt mobile phase: EPDM/PTFE, same dimensions, and check the blank series for extractables before you commit.

Headspace GC, 40 to 90 °C, 10 to 20 psi: butyl/PTFE, 20 mm crimp for 20 mL vials, magnetic or crimp cap rated for the pressure, and do not reuse the septum. One puncture per vial if you can arrange it.

Long-term storage, weeks to months, volatile analyte: butyl/PTFE, and store upright at the lowest temperature the method allows.

Cold storage at -80 °C: silicone/PTFE, accept the permeability, and thaw in a sealed secondary container.

The one place I would push back on common practice is the habit of using silicone everywhere because it is cheap and available. It is the right grade for most HPLC work, but it is the wrong grade for headspace and for long storage, and those are exactly the methods where a closure failure costs you a re-run of a stability study. Match the grade to the method. Write down why. Keep the blank chromatograms. That is the whole discipline.

The material grade is not a purchasing decision. It is a method parameter, and it belongs in your method development record next to the column and the mobile phase. Treat it that way and the closure stops being the variable you cannot explain.

References & further reading

Related resources from our lab consumables network:

External references and standards: