Match the polymer to the solvent class before anything else

A 96-position HPLC run failed at injection 41 with a peak-area RSD of 6.8% across six replicate standards. The method used acetonitrile/water 60:40 at 40 °C, a PTFE-faced silicone septum, and a cap torqued to 1.6 in·lb. The septum had swollen 0.4 mm and extruded into the headspace over 14 hours of tray dwell. The fix was not a new column. It was a change of septum material. That decision, which vial material to use for a given solvent and temperature, is the one I get asked about most, and it is the one where most labs are quietly guessing.

This guide is about polymer and laminate selection against the method, not about vial dimensions, cap fit, or price. Those are separate problems. Here is what actually governs the choice: the solvent contact, the pH, the dwell temperature, the puncture count, and the failure mode you can tolerate.

Every septum and liner you buy is a compromise. PTFE is chemically inert but has no elasticity. Silicone is elastic but swells in nonpolar solvents. EPDM handles aqueous and polar systems well but degrades against ketones and aromatic hydrocarbons. Laminates try to combine the best of two materials, and they succeed or fail at the bond line. Get the solvent class wrong and no amount of torque or technique saves the run.

PTFE: the inert barrier, never the seal

PTFE is the reference material for chemical resistance. It resists almost every HPLC solvent, from acetonitrile to dimethyl sulfoxide, and holds from roughly -200 °C to +260 °C in continuous service. Against pH it is effectively inert across the whole 0 to 14 range. What it does not do is seal. A pure PTFE septum has no memory, so once the needle withdraws, the hole stays open. That is why you rarely see solid PTFE as a working septum; you see it as the sample-facing face of a laminate, typically 0.05 to 0.13 mm thick.

The failure when PTFE is used as the only barrier is straightforward: coring and permanent puncture channels. I have counted eight to twelve usable punctures on a thin unsupported PTFE disc before leakage becomes visible at the injection port, versus 50 to 100 on a proper silicone-backed laminate. If your autosampler uses a single septum across a full tray, that puncture budget is the number that decides your cap change interval.

Silicone: elasticity with a solvent ceiling

Silicone gives the reseal. It is the layer that closes the puncture, and it tolerates a wide temperature band, roughly -60 °C to +200 °C. Its weakness is swelling in nonpolar and aromatic solvents. Toluene, xylene, hexane, and dichloromethane will take up into the silicone and cause it to swell, soften, and extrude under cap compression. The pH window is narrower than PTFE as well; strong bases above pH 10 will attack it over time.

The classic failure is the swollen septum that pushes into the headspace and either touches the sample surface or blocks the needle path. That is the case from the opening. The measurement that confirmed it was a caliper check on a used septum against a fresh one: 0.4 mm of swell after 14 hours in 60:40 acetonitrile/water. Acetonitrile alone would have been fine. The water fraction drove hydrolysis of the silicone backbone. The fix was a PTFE/silicone laminate with a thicker PTFE face, which dropped the swell below 0.1 mm and returned RSD to 0.9%.

EPDM: aqueous and polar, not for ketones

EPDM is the workhorse for aqueous, polar, and buffer-heavy methods. It holds pH from about 1 to 12 and temperature to roughly +130 °C in continuous contact. It is cheaper than silicone and reseals well. But EPDM has a specific failure profile: ketones (acetone, methyl ethyl ketone), aromatic hydrocarbons, and chlorinated solvents will swell and eventually crumble it. If your method uses acetone anywhere in the mobile phase or as a needle wash, EPDM is the wrong choice and you will find out in the form of black particulate in the vial and a fouled needle.

EPDM also ages under repeated autoclave or steam exposure. If the vial goes through sterilization before analysis, count the cycles. I have seen EPDM liners fail at cycle 12 to 15 under 121 °C steam, showing surface cracking that does not appear until you look under 10x magnification.

Laminate: the bond line is the failure point

Most working septa today are laminates: PTFE on the sample side, silicone or EPDM on the cap side. The PTFE gives chemical resistance, the elastomer gives reseal. When a laminate fails, it usually fails at the interface. Delamination shows up as a lifted PTFE face, which then coring-punches and drops a disc into the vial. That disc will be aspirated by the needle and will appear in the chromatogram as a broad, late-eluting artifact or as a blocked needle. The bond quality is not visible on a spec sheet, so this is one place where supplier qualification matters more than the data table.

Temperature turns a good choice into a bad one

Temperature is the second axis, and it interacts with solvent in ways the spec sheet does not spell out. A silicone septum rated to 200 °C is fine dry. Put it in contact with acetonitrile at 60 °C for six hours and the effective ceiling drops. Thermal expansion, solvent uptake, and compression set compound. I use a simple rule: derate the published temperature ceiling by 30 to 40 °C whenever the septum sees a swelling solvent in continuous contact.

For headspace work, the temperature question is sharper. A headspace vial runs at 60 to 150 °C with the septum under pressure, and the septum must hold both pressure and reseal after the needle. PTFE/silicone laminates dominate here for good reason: the PTFE face resists the sample matrix, the silicone reseals, and the whole assembly holds to about 200 °C. But if the sample contains a strong base above pH 10, the silicone core is at risk and you should be looking at a PTFE-faced butyl or a different cap architecture entirely.

Low-temperature work, such as a chilled autosampler tray at 4 to 10 °C, is the opposite problem. Most polymers stiffen and lose reseal capability as they cool. Silicone stays flexible to -60 °C. EPDM stiffens below -20 °C and can fail to reseal at 4 °C after several punctures. If your method runs cold, silicone or a silicone-based laminate is the safe pick.

pH narrows the field faster than solvent does

Acidic methods below pH 2 will attack the silicone backbone and the EPDM crosslinks over time. Basic methods above pH 10 will do the same to silicone. PTFE is the only face that survives both extremes without change, which is why the PTFE-faced laminate is the default whenever the method pH leaves the 3 to 10 band. The failure mode is subtle: the septum does not visibly fail, it just loses reseal strength, and you start seeing small volume losses in the vial after multiple injections. A gravimetric check on a capped vial before and after 24 hours will show the loss; anything above 0.5% mass loss in that window means the seal is not holding.

If the method uses a buffer, remember the buffer is the aggressive part, not the organic modifier. Phosphate buffers at pH 7 are gentle. Ammonium bicarbonate at pH 9 is not. Citrate buffers at pH 3 will slowly hydrolyze silicone. Check the buffer, not just the mobile phase organic fraction.

The failure each wrong choice produces, and how to read it

You can usually diagnose the wrong material from the chromatogram before you touch the vial. Here is the field guide I use.

• Rising RSD with tray dwell time. Swell or extrusion. Measure a used septum against a fresh one. If it has grown more than 0.2 mm, the material is wrong for the solvent.
• Late-eluting broad peaks with no source. A cored plug or a delaminated PTFE disc is in the vial. Inspect the septum face under 10x; look for a missing disc.
• Black or grey particulate in the vial. EPDM against a ketone or aromatic. Switch to a PTFE-faced laminate.
• Volume loss over 24 hours. Reseal failure from pH attack or thermal compression set. Weigh capped vials at t=0 and t=24 h.
• Needle blockage or bent needle. Extruded septum into the headspace, or a swollen septum in the needle path. Check cap torque as well; over-torque above about 2.0 in·lb on a 13 mm cap will deform the septum and cause this.
• Ghost peaks at the retention time of a plasticizer. Silicone or EPDM breakdown products. Confirm by running a blank vial with the suspect septum and no sample.

For GC methods, the same logic applies but the stakes are higher because the inlet is hot. A GC vial septum sees the inlet temperature directly, and a silicone septum that would survive at 40 °C in HPLC will bleed siloxanes at 250 °C in a split/splitless inlet. That bleed shows up as a baseline rise and as siloxane peaks at m/z 73, 147, 207, and 281 in MS detection. If you see those, the septum is the problem, not the column. Restek publishes useful guidance on inlet and septum artifacts that I point new analysts to when they are chasing this.

Two field cases that turned on material choice

Case 1: the headspace method that drifted at 90 °C

A lab running residual solvents by headspace at 90 °C for 30 minutes started seeing a slow drift in response for the late-eluting analytes, about 4 to 6% over an 8-hour sequence. The symptom looked like a detector drift. It was not. A blank headspace run with the same septa showed a growing siloxane envelope. The septa were silicone with a thin PTFE face, and the PTFE face was 0.05 mm, thin enough that at 90 °C the silicone bled through the face over the run. The fix was to move to a 0.13 mm PTFE-faced laminate and to lower the headspace oven to 80 °C with a longer equilibration. The drift dropped to under 1% and the siloxane envelope disappeared. The lesson: face thickness matters at temperature, and a thin face is not the same product as a thick one.

Case 2: the pH 11 method that lost volume

A lab running a basic method at pH 11 with a silicone/PTFE laminate found that replicate injections of the same vial gave decreasing peak areas, about 2% per injection over five injections. The vial was not leaking visibly. A gravimetric check on capped vials held at room temperature for 24 hours showed 0.8% mass loss on the silicone-sealed vials versus 0.1% on a PTFE/EPDM set. The base was attacking the silicone core through the needle channel and degrading the reseal. Switching to a PTFE-faced EPDM laminate, which holds pH to 12, eliminated the loss. This is the case where the spec sheet says “PTFE-faced” and the reader assumes it is fine; the face is fine, the core is not.

A decision rule you can apply in the method review

Here is the checklist I run when a method is being validated or transferred. It takes about ten minutes and it prevents most of the failures above.

1. List the solvents. Identify any nonpolar, aromatic, or chlorinated solvent. If present, silicone is out and PTFE-faced laminate is in.
2. List the pH. If below 3 or above 10, silicone is out and the core must be EPDM (to pH 12) or the face must carry the load with a compatible core.
3. Note the maximum temperature. Derate the published ceiling by 30 to 40 °C if a swelling solvent is present. If the method exceeds the derated ceiling, change material.
4. Count the punctures per septum. If the autosampler uses one septum across more than 50 injections, verify the puncture budget and set a cap-change interval accordingly.
5. Check the cap torque. Target 1.2 to 1.6 in·lb on a 13 mm cap, 1.5 to 2.0 in·lb on an 18 mm cap. Above 2.0 in·lb, the septum deforms and the failure modes above get worse.
6. Run a blank and a gravimetric check. A blank vial with the candidate septum, plus a 24-hour mass-loss check, will catch reseal and bleed problems before they reach the method.
7. Confirm the standard. For regulated work, the container closure system is part of the method under USP <660> and the extractables question under USP <1663>/<1664>. The material choice has to be documented, not just chosen.

If you are sourcing across a range of methods, the vial selection should be driven by the most aggressive method in the lab, not the average one. A single PTFE/silicone laminate will cover most HPLC and headspace work; a PTFE/EPDM laminate covers the basic end; solid PTFE is a liner, not a septum. For GC inlet work, check the septum against the inlet temperature and the injection count, and treat siloxane bleed as a material signal rather than a column problem. Agilent and Thermo Fisher both publish septum selection tables that align with the numbers above, and the EPA SW-846 methods set the container performance requirements for environmental work if that is your regulatory frame.

For method transfer documentation, keep the material grade on the method record next to the column and the mobile phase. If the grade changes, the method has changed. That is the decision rule: solvent class picks the face, pH picks the core, temperature derates the ceiling, and puncture count sets the change interval. Choose the material that survives the worst combination in the method, not the one that works on the day of validation.

References & further reading

Related resources from our lab consumables network:

• Vial at vials-filters.com
• Headspace Vial at headspacevialsupplier.com
• Gc Vial at gcvialfactory.com
• Reagent Bottle at reagentbottleforsale.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