What a septum actually has to do

I have pulled more failed septa out of autosampler trays than I care to count, and the diagnosis is almost never the septum itself. It is the wrong septum for the job, a cap crimped with the wrong tool, or a vial filled too high and left sitting on a tray for three days. Over the years I have kept a running log of failure modes across our own QC lab and the customer sites we support, and the same six or seven problems account for nearly every complaint. What follows is that log turned into a working reference: what each failure looks like, why it happens, and the specific numbers and practices that prevent it.

A septum is a thin elastomer disc, usually 3 mm to 9 mm thick and 8 mm to 11 mm in diameter depending on cap format, held in a screw cap or crimp cap. In a 2 mL vial with a 9 mm opening, the exposed septum face is only about 6 mm across. That small disc has to do four jobs at once:

• Seal the vial against solvent loss and air ingress for the life of the sample, often 24 to 72 hours on a tray.
• Accept one or more needle punctures (from 1 to 20+ depending on workflow) without coring, tearing, or resealing badly.
• Stay chemically inert to the sample solvent, headspace, and any derivatization chemistry.
• Contribute essentially nothing to the chromatogram (no extractables, no particles, no siloxane peaks).

When any one of those fails, you get a symptom in the data, not in the vial. That is why septa problems get misdiagnosed as column problems, pump problems, or “bad sample prep.” The rest of this article walks through the failures I see most, in rough order of frequency.

Failure 1: coring and needle punch-through debris

Coring is the classic. The needle punches a plug of elastomer out of the septum, and that plug either falls into the vial or, worse, sits in the needle bore and gets injected. You will see it as a sudden spike of siloxane peaks in GC-MS at m/z 73, 147, 207, or as a particle that briefly plugs the column inlet and shows up as a pressure excursion.

Root causes

• Blunt or burred needles. A needle that has been used 500+ injections develops a rough bevel that cuts rather than parts the elastomer.
• Wrong needle gauge for the septum. A 22-gauge needle (0.72 mm OD) through a soft silicone septum with a thin PTFE face is fine; a 26-gauge needle (0.46 mm OD) through a hard butyl septum at 10 °C can shear the face.
• Needle guide misalignment. If the autosampler needle does not enter perpendicular to the septum, it drags the elastomer and tears a crescent instead of a clean hole.
• Over-tightened crimp caps. A cap crimped too hard compresses the septum and forces the needle to punch through a thinner, denser layer, which cores more readily.

Prevention

Match the septum hardness to the needle. Soft silicone (Shore A 40 to 50) tolerates repeated puncture best; laminated silicone/PTFE is the workhorse for HPLC because the PTFE face resists solvent attack and the silicone body reseals. For a workflow with more than 5 punctures per vial, use a pre-slit septum. Pre-slit septa eliminate coring almost entirely but are a poor choice for long tray residence times because the slit is a leak path once the needle is withdrawn. My rule of thumb: pre-slit for high-throughput short-run methods, solid for anything sitting on a tray overnight.

Replace autosampler needles on a schedule, not on failure. In our lab we swap at 1,000 injections for standard 22-gauge needles. Check needle alignment quarterly with a gauge block. And use a properly set crimper: a 11 mm crimp cap should be crimped to a consistent 1.0 to 1.2 mm cap skirt height, checked with a crimp gauge, not by feel.

Failure 2: solvent loss and sample concentration drift

This one is insidious because there is no dramatic failure. The peak area for a standard drifts down 3 to 8% over a 24-hour run, and the analyst blames the detector. The real cause is permeation and leakage through the septum.

Root causes

• Permeation through the elastomer. Silicone is highly permeable to many solvents, especially dichloromethane, ethyl acetate, and hexane. A pure silicone septum can lose measurable solvent through the disc over hours.
• Poor reseal after puncture. A hard, highly crosslinked septum that has been punctured several times does not close cleanly, leaving a microchannel.
• Loose caps. A screw cap torqued below about 0.5 N·m does not compress the septum enough to seal against the vial rim.
• Vial overfill. Fill a 2 mL vial above about 1.5 mL and the liquid contacts the septum, wicking solvent into the elastomer and accelerating both permeation and extractables.

Prevention

Use a PTFE-faced laminated septum for any volatile solvent. The PTFE barrier (typically 0.05 to 0.13 mm) cuts permeation dramatically compared with bare silicone. Keep fill volume at or below 1.5 mL in a 2 mL vial and below 0.7 mL in a 1 mL vial. Torque screw caps to the manufacturer’s spec; for most 9 mm screw caps that is roughly 0.6 to 0.9 N·m, which you can hit consistently with a torque-limited capper rather than hand tightening. If samples sit more than 24 hours, run a solvent-loss check: weigh 10 filled vials at t=0 and t=24 h and confirm mass loss is under 0.5%.

Failure 3: septa extractables and ghost peaks

You inject a blank and get peaks. Not huge, but there: a siloxane cluster, a plasticizer, an antioxidant. Septa are a common source, and the culprit is usually the wrong elastomer grade rather than a defective batch.

Root causes

• Unwashed or low-grade silicone. Inexpensive septa carry residual low-molecular-weight siloxanes and curing agents that bleed into the headspace.
• Sulfur-containing elastomers. Standard butyl or EPDM can release sulfur compounds that wreck sulfur-selective detection.
• Peroxide-cured silicone. Peroxide cure leaves volatile byproducts; platinum-cured silicone is much cleaner for trace work.
• Septum contact with solvent. Overfilled vials or inverted storage put liquid against the septum, pulling extractables into the sample.

Prevention

For trace analysis, specify platinum-cured, solvent-washed silicone with a PTFE face, and ask for a certificate of analysis with an extractables profile. For GC-MS at trace levels, precondition new septa by heating at 100 to 150 °C for 1 to 2 hours under nitrogen before use; this drives off most volatile bleed. Run a system blank with a fresh septum every time you change a lot. If you see siloxane peaks at m/z 73/147/207 in a blank, the septum or the septa in your inlet are suspect before the column is.

Failure 4: chemical attack and swelling

Not every solvent is compatible with every elastomer. Put the wrong pair together and the septum swells, softens, and either extrudes into the vial or disintegrates into the sample.

Root causes

• Silicone in nonpolar solvents. Silicone swells badly in hexane, toluene, and dichloromethane.
• Butyl in chlorinated solvents. Butyl has poor resistance to dichloromethane and chloroform.
• EPDM in aromatic hydrocarbons. EPDM is fine in water and polar solvents but swells in toluene and xylene.
• PTFE face damage. A PTFE-faced septum is only as good as its face; a needle that scratches through the PTFE exposes the silicone underneath to the solvent.

Prevention

Choose by chemistry, not by price. The table below is the compatibility guide I hand to new analysts; it reflects standard manufacturer data and our own swelling tests (48 h immersion at 25 °C, swell measured as volume change).

If you are unsure, run a 48-hour immersion test in your actual mobile phase or sample solvent and measure swell and mass change. More than 10% volume swell means the elastomer is wrong for the job.

Failure 5: temperature and pressure limits

Septa are rated, and the ratings matter. Exceed the upper temperature and the elastomer degrades, hardens, and starts to bleed. Exceed the pressure rating (relevant in pressurized sample loops and some headspace systems) and the septum extrudes or blows out.

Root causes

• Autoclave or oven exposure above rating. Standard silicone is good to about 200 °C; pushing to 220 °C shortens life and increases bleed.
• Cold-room use. Below about -20 °C many septa lose elasticity and reseal poorly after puncture.
• Headspace vial over-pressurization. A headspace vial pressurized above the septum’s burst rating can leak or extrude the disc.

Prevention

Stay within the published range with a 10 to 20 °C margin for long runs. For high-temperature GC inlets, use a septum rated to at least 350 °C (high-temperature silicone or a graphite-faced design) and change it more often. For cold rooms, use a low-temperature silicone and verify reseal by running a blank after 24 h at temperature.

Failure 6: mechanical damage from caps and handling

Half the “bad septum” complaints I investigate turn out to be cap or handling problems. The septum is fine; the way it was installed or stored is not.

Root causes

• Over-crimping. A crimp cap squeezed too hard cuts into the septum edge and creates a leak path at the rim.
• Under-crimping. A loose crimp lets the septum shift, so the needle hits off-center and tears.
• Cross-threading screw caps. A cross-threaded 9 mm screw cap does not seat the septum flat, and the seal fails at one point.
• Storage damage. Septa stored in a warm drawer or exposed to solvent vapor degrade before they are ever used.
• Reusing caps and septa. A septum is a single-use part. Reusing it doubles the puncture count and the leak risk.

Prevention

Use a calibrated crimper and check skirt height with a gauge. For screw caps, use caps with a bonded septum so there is no separate disc to misalign, and tighten with a torque-limited capper. Store septa in the original sealed bag at 15 to 25 °C, away from solvent vapors and direct light, and use within the manufacturer’s shelf life (typically 2 to 3 years unopened). Never reuse a septum; the cost of a septum is trivial next to the cost of a repeat run.

Failure 7: particulate shedding and needle blockage

Low-quality septa shed particles, especially after several punctures. Those particles end up in the vial, in the needle, or on the column frit. The symptom is intermittent needle blockage or a rising system pressure that clears when you change the vial.

Root causes

• Filler-rich or poorly compounded elastomer. Cheap silicone loaded with fillers sheds more.
• Coring debris (see failure 1) lodging in the needle.
• PTFE face flaking after repeated puncture or solvent attack.

Prevention

Buy septa with a documented low-particulate specification, and change them on a schedule tied to puncture count, not to visible damage. If you see intermittent needle blockage, inspect the needle bore under magnification and flush the autosampler with a wash cycle between injections. For high-particle methods, a pre-slit PTFE-faced septum plus a needle wash is the practical combination.

A troubleshooting checklist you can run in 15 minutes

When a septa-related symptom appears, work through this in order. It catches the majority of cases before you touch the column.

• Check the blank. Inject a fresh blank with a new septum. Siloxane peaks at m/z 73/147/207 point to septum bleed or coring.
• Inspect the septum. Remove it and look for a clean hole, a crescent tear, a plug missing, swelling, or discoloration. Photograph it and log the lot number.
• Count punctures. If the vial has been punctured more than the septum’s rating (5 to 10 for laminated, 10 to 20 for pure silicone), change it.
• Check fill volume. More than 1.5 mL in a 2 mL vial is overfilled; correct it.
• Weigh for solvent loss. 10 vials, t=0 and t=24 h; more than 0.5% mass loss means a seal problem.
• Check crimp or torque. Crimp skirt height 1.0 to 1.2 mm; screw cap torque 0.6 to 0.9 N·m.
• Check needle. Inspect for burrs; replace at 1,000 injections.
• Check chemistry. Confirm the elastomer is compatible with the sample solvent (see table).
• Check temperature. Confirm the septum rating covers the method temperature with a 10 to 20 °C margin.
• Check storage. Confirm septa are in date and stored sealed at 15 to 25 °C.

Selection criteria with numbers

If you are specifying septa for a new method, here is the decision path I use. Start with the chemistry, then the temperature, then the puncture count, then the format.

Step 1: chemistry

Aqueous and polar organic mobile phases: PTFE-faced silicone or butyl. Nonpolar solvents (hexane, toluene, DCM): PTFE-faced silicone for short contact, butyl for longer contact, never bare silicone. Trace GC-MS: platinum-cured, solvent-washed silicone, preconditioned at 100 to 150 °C.

Step 2: temperature

Standard methods at ambient to 40 °C: any compatible elastomer. Headspace or high-temperature work: silicone rated to at least 200 °C, or high-temperature silicone to 350 °C for GC inlets. Cold room: low-temperature silicone rated to -40 °C or below.

Step 3: puncture count

1 to 3 punctures: solid laminated septum. 5 to 10: pre-slit or pure silicone. More than 10: pre-slit PTFE-faced silicone, and change the vial.

Step 4: format

9 mm screw cap for most HPLC autosamplers; 11 mm crimp for GC and some HPLC; 20 mm for headspace. Confirm the cap format against your autosampler’s tray specification. GPI thread standards (GPI 8-425, GPI 13-425, GPI 15-425, GPI 18-400, GPI 24-400) define the thread; using a non-standard cap on a standard vial is a common source of leaks.

For a broader look at how septa fit into the full closure system, including caps and liners, the vial closures reference is worth bookmarking. If you are matching septa to a specific autosampler vial, the HPLC vial guide covers the common formats and dimensions.

Where to buy and what to ask for

Septa are commodity parts until they are not. The difference between a good lot and a bad one is often invisible until you run a blank. Ask suppliers for:

• A certificate of analysis with extractables and lot traceability.
• Elastomer grade (platinum-cured vs peroxide-cured for silicone).
• Hardness (Shore A) and thickness tolerance.
• Temperature rating and puncture rating.
• Shelf life and storage conditions.

Our own HPLC septa range is specified this way, and I would hold any supplier to the same standard. For method development and validation work, the consumables guidance from Agilent’s chromatography consumables and Thermo Fisher’s sample handling documentation is a useful cross-check on compatibility claims. For regulatory methods, the U.S. EPA SW-846 series specifies sample handling and holding times that directly constrain how long a septum must seal; and Restek’s technical resources have good data on septa bleed and inlet maintenance. If you need independent verification of extractables or seal performance, an analytical testing lab can run the immersion and mass-loss tests for you.

Takeaway

Most septa failures are specification and handling failures, not manufacturing defects. Match the elastomer to the solvent, the hardness and pre-slit choice to the puncture count, and the temperature rating to the method with a 10 to 20 °C margin. Keep fill volume at or below 1.5 mL in a 2 mL vial, crimp to 1.0 to 1.2 mm skirt height or torque screw caps to 0.6 to 0.9 N·m, and replace needles at 1,000 injections. Store septa sealed at 15 to 25 °C and never reuse them. Run the 15-minute checklist before you blame the column. Do those things and the septum stops being a variable in your data.

References & further reading

Related resources from our lab consumables network:

• Hplc Septa at hplcsepta.com
• Vial Closures at vialsclosures.com
• Analytical Testing at testinganalytical.com
• Hplc Vial at vialforhplc.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