What a septa actually has to do--Aijiren HPLC Vials

I have watched a septa ruin a run in three different ways in the same week: a coring event that dropped a rubber plug into a GC inlet, a silicone disc that bled enough siloxane to bury a trace sulfur peak, and a cap that somebody over-torqued until the PTFE face split along the injection point. None of these were exotic failures. They were the same four or five mistakes I see every quarter, and they all cost more in lost instrument time than the septa cost to replace. This article is what I tell new lab techs in their first month, written down so I do not have to repeat it.

If you handle sample introduction at all, whether on an HPLC vial or a GC autosampler tray, the septa on your vials and inlet are the most abused consumable in the workflow. They are punctured, heated, compressed, and bathed in solvent, often hundreds of times per week. Understanding why they fail is the fastest route to fewer ghost peaks, fewer aborted sequences, and fewer calls to the service engineer.

A septa is a small elastomer disc, usually 11 mm, 13 mm, or 20 mm in diameter, that seals a vial or an inlet against the atmosphere. In a crimp vial it sits inside a crimp cap. In a screw-thread vial it sits in a cap that engages a GPI 13-425 or GPI 8-425 thread (GPI = Glass Packaging Institute). On a GC inlet it is clamped under a reducing nut at the top of the injector body.

The job sounds simple. It is not. A good septa has to:

Hold pressure without leaking, typically up to 1.5 to 2.5 bar of carrier gas on a split/splitless inlet.
Re-seal after each needle puncture, hundreds of times, without coring or leaking.
Stay chemically inert against the sample solvent, which may be hexane, methanol, acetonitrile, dichloromethane, or something aggressive like dimethylformamide.
Survive the temperature of the injection port, which may be anywhere from 40 °C to 350 °C depending on the method.
Not bleed. Any volatile material that escapes the septa ends up as a baseline artifact or a ghost peak.

Those five requirements are in tension. A soft silicone seals beautifully but bleeds and tears. A hard PTFE-faced laminate resists solvent but does not reseal as well. The septa you buy is a compromise, and most failures come from using the wrong compromise for the job.

The seven failures I see most often

1. Coring (the needle punches a plug out of the septa)

Coring is the classic septa failure. A blunt or burred needle, or a needle that is slightly off-axis, punches a small disc out of the septa instead of slicing through it. That disc falls into the vial (contaminating the sample) or, worse, into the inlet liner, where it sits and slowly pyrolyzes for the rest of the sequence.

I have pulled cores out of inlet liners that were 2 mm across and 1 mm thick, sitting on top of the glass wool. Every injection after that core formed showed a rising baseline and a cluster of siloxane peaks at m/z 73, 147, 207, 281. That is the classic silicone fingerprint.

Root causes:

Needles that have been used past their service life. A 23-gauge needle on a CTC or Gerstel autosampler should be inspected every 500 to 1000 injections. Point 2 of the bevel should be sharp under a loupe.
Needle guide misalignment. If the guide is bent, the needle enters the septa at an angle and cuts rather than pierces.
Septa that are too thin or too soft for the needle gauge. A 0.5 mm silicone disc on a 22-gauge needle is going to core eventually.
Needle depth set too shallow. If the needle only penetrates the top 2 mm of the septa, it drags material instead of cutting cleanly through.

Prevention: use a needle with a sharp, undamaged bevel, check alignment monthly, and choose a septa with a PTFE face on the sample side. The PTFE layer is what the needle actually slides against, and it dramatically reduces coring. If you are buying septa for a vial septa application, ask for a silicone/PTFE laminate rather than straight silicone.

2. Bleed (the septa outgasses into the carrier gas)

Every septa bleeds. Silicone septa bleed siloxanes, which show up as a regular series of peaks at m/z 73, 147, 207, 281, 355, 429. At low inlet temperatures (under 200 °C) the bleed is small and mostly invisible. Above 280 °C it becomes a real problem, and above 320 °C a standard silicone septa will produce a baseline that never settles.

Bleed is worse when:

The inlet is hot and the septa is old. A septa that has been at 300 °C for 8 hours has effectively been baked. Siloxane loss accelerates.
The septa is the wrong formulation. General-purpose silicone is not the same as a low-bleed silicone or a high-temperature fluoroelastomer.
The septa purge flow is too low. The purge flow (typically 3 to 5 mL/min on a split/splitless inlet) is what sweeps septa bleed away before it reaches the column. If the purge is off or blocked, bleed goes straight onto the column.

Prevention: match the septa to the inlet temperature. For inlets running above 300 °C, use a high-temperature septa (fluoroelastomer or a premium low-bleed silicone) and keep the septa purge flow at 3 to 5 mL/min. Change the septa on a schedule, not on failure. I change septa every 100 to 200 injections on a hot inlet, and every 500 injections on a cool one.

3. Leaks at the cap or nut

A septa leak is usually not the septa itself. It is the cap, the crimp, or the nut that holds the septa in place. On a crimp vial, an under-crimped cap (crimp diameter too large) or an over-crimped cap (crimp diameter too small) will both leak. The correct crimp for a 11 mm aluminum cap is a crimp diameter of 10.9 to 11.0 mm measured across the crimp. Below 10.8 mm you crush the septa and split it. Above 11.1 mm you do not compress it enough to seal.

On a screw-thread vial, leaks come from under-torquing. A GPI 13-425 cap should be hand-tightened to roughly 0.8 to 1.2 N·m. Most people overtighten, which deforms the septa and can cause it to extrude into the neck of the vial. I have seen septa extruded so far into the vial neck that the needle hit rubber on the way down.

On a GC inlet, the septa is held by a reducing nut. The nut should be snug, not torqued to spec. Overtightening compresses the septa, closes the needle channel, and causes the needle to core. A reducing nut should be tightened by hand plus a quarter turn with a wrench, no more.

Prevention: use a calibrated crimper, check crimp diameter with a go/no-go gauge once a month, and train staff on the correct torque for screw caps. For high-throughput workflows, a torque-limited capper is worth the money.

4. Chemical attack (the septa swells, softens, or dissolves)

Silicone is the default septa material because it is chemically resistant to most solvents, but it is not resistant to everything. It swells in nonpolar solvents like hexane and toluene, and it degrades in strong acids, strong bases, and some chlorinated solvents. A swollen septa loses its compression set, which means it stops resealing after puncture.

The worst offenders I have dealt with:

Dichloromethane and chloroform. These swell silicone noticeably within hours and can cause the septa to extrude through the cap.
Dimethylformamide and dimethyl sulfoxide. These attack the silicone slowly but steadily, and the septa becomes gummy after a few days.
Trifluoroacetic acid and other strong acids. These will pit and crack a silicone septa within a single sequence.

For these solvents, use a PTFE-faced septa (PTFE on the sample side) or a PTFE/silicone laminate. The PTFE face is what contacts the sample, and it is essentially inert. The silicone behind it provides the seal. For very aggressive solvents, a pure PTFE or PTFE/silicone laminated septa is the only sensible choice.

Prevention: match the septa material to the solvent. If you are unsure, run a 24-hour swell test: weigh the septa, soak it in the solvent, weigh it again after 24 hours. Anything more than 5% weight gain is a red flag.

5. Thermal degradation (the septa hardens and cracks)

At high inlet temperatures, silicone septa harden. The elastomer cross-links further, loses elasticity, and eventually cracks. A hardened septa does not reseal after puncture, and it can crumble into the inlet. I have seen a septa that had been at 330 °C for a week turn into a brittle disc that shattered when the reducing nut was loosened.

Prevention: use a high-temperature septa rated for the inlet temperature. Standard silicone is rated to about 250 to 280 °C. Premium low-bleed silicone is rated to about 300 to 320 °C. Fluoroelastomer septa are rated to about 350 to 400 °C. Do not run a standard septa at 350 °C and expect it to last.

6. Contamination (the septa introduces artifacts)

Septa are manufactured with curing agents, plasticizers, and mold-release agents. These can leach into the sample or outgas into the carrier gas. A cheap septa can introduce phthalates, adipates, and other plasticizers into a trace analysis. If you are running EPA SW-846 methods or any method with low detection limits, a contaminated septa is a real source of false positives.

Prevention: buy septa from a supplier that specifies low extractables. For trace work, precondition new septa by heating them at 100 °C for an hour before use, or buy pre-cleaned septa. The U.S. EPA SW-846 test methods are a useful reference for what “trace” actually means in your method, and they will tell you whether a cheap septa is going to cost you a re-run.

7. Physical damage (the septa is cut, torn, or deformed)

This is the catch-all category. A septa can be cut by a needle that is too large, torn by a cap that is over-crimped, deformed by a reducing nut that is over-tightened, or damaged by a needle that is inserted at an angle. Physical damage is usually obvious on inspection: a visible slit, a chunk missing, or a distorted shape.

Prevention: inspect septa on removal. If you see a slit longer than 2 mm, a missing chunk, or a septa that is no longer flat, you have a mechanical problem upstream. Fix the needle, the crimper, or the nut before you replace the septa.

Septa selection: a decision checklist with numbers

The table below is what I use when specifying septa for a new method. It is not exhaustive, but it covers 90% of the decisions I make.

Inlet temperature under 250 °C, standard solvents: standard silicone, 11 mm for a GPI 13-425 vial, 0.9 to 1.1 mm thick. Change every 500 injections.
Inlet temperature 250 to 300 °C, standard solvents: low-bleed silicone, 11 mm, 1.0 to 1.3 mm thick. Change every 200 injections.
Inlet temperature above 300 °C: high-temperature silicone or fluoroelastomer, 11 mm, 1.0 to 1.5 mm thick. Change every 100 to 150 injections.
Aggressive solvents (DCM, DMF, TFA): PTFE-faced silicone or PTFE/silicone laminate, 11 mm, 1.0 to 1.5 mm thick. Change every 100 injections or every 24 hours, whichever comes first.
Trace analysis (low ppb): pre-cleaned low-bleed silicone, 11 mm, 1.0 to 1.3 mm thick. Change every 100 injections.
Headspace or SPME: PTFE-faced silicone, 20 mm for a GPI 20-400 vial, 1.5 to 2.0 mm thick. Change every 50 to 100 injections.

One more criterion that gets overlooked: needle gauge. A 23-gauge needle (0.64 mm OD) is fine for most septa. A 26-gauge needle (0.46 mm OD) is too fine for a soft silicone septa and will coring more easily. A 22-gauge needle (0.72 mm OD) needs a PTFE-faced septa to avoid coring.

A troubleshooting checklist you can pin to the wall

When a run goes wrong and you suspect the septa, work through this in order. It takes about 10 minutes and will save you an hour of guessing.

Check the inlet pressure. A sudden drop in pressure at the inlet is a septa leak. A sudden rise is a blocked purge line. Both are septa-adjacent.
Run a blank. A blank injection with no sample will show septa bleed as a regular series of siloxane peaks. If the peaks are there, the septa is bleeding.
Inspect the septa on removal. Look for coring, slits, extrusion, hardening, or discoloration. Photograph it. A photo log will tell you whether the failure is random or systematic.
Check the crimp diameter. Use a go/no-go gauge on a sample of 10 caps. If more than one fails, recalibrate the crimper.
Check the needle. Look at the bevel under a loupe. If it is blunt or burred, replace it.
Check the reducing nut torque. If the nut is over-tightened, the septa will be compressed and the needle will core. Loosen it and re-tighten by hand plus a quarter turn.
Check the purge flow. A purge flow below 2 mL/min is not enough to sweep septa bleed away. Set it to 3 to 5 mL/min.
Check the inlet temperature. If the inlet is above the septa rating, the septa is degrading. Either lower the temperature or change the septa material.
Check the solvent. If the sample solvent is aggressive, the septa may be swelling. Run a swell test or switch to a PTFE-faced septa.
Check the lot. If you have just opened a new box and failures started, the lot may be bad. Compare with the previous lot and, if necessary, contact the supplier.

Preventive maintenance schedule that actually works

The single biggest improvement I have made in my own lab is to stop treating septa as a “replace when it fails” item. They are a scheduled consumable, like a column or a lamp. Here is the schedule I run:

Daily: check inlet pressure at the start of the sequence. A 5% deviation from the method setpoint is a flag.
Weekly: inspect the septa on the inlet. If it is more than 0.2 mm thinner than a new one, replace it.
Every 100 to 200 injections: replace the inlet septa. Log the injection count.
Every 500 injections: replace the autosampler needle and check the needle guide alignment.
Monthly: check crimper calibration with a go/no-go gauge and check reducing nut torque on all inlets.
Quarterly: run a full bleed test on a new lot of septa before putting it into routine use.

The injection-count approach is more reliable than a calendar-based one, because septa wear is driven by punctures, not by time. A septa that sits on a shelf for a month is not the same as one that has been punctured 400 times.

Where to buy and what to specify

When you specify septa, give your supplier the following information. If they cannot answer all of it, find another supplier.

Vial thread: GPI 13-425, GPI 8-425, or GPI 20-400.
Septa diameter: 11 mm, 13 mm, or 20 mm.
Septa thickness: 0.9 to 1.5 mm for standard vials, 1.5 to 2.0 mm for headspace.
Material: silicone, PTFE-faced silicone, PTFE/silicone laminate, or fluoroelastomer.
Temperature rating: at least 20 °C above your maximum inlet temperature.
Bleed specification: siloxane bleed at the rated temperature, expressed as a baseline rise in pA at a specified detector gain.
Puncture count: the number of injections the supplier will warrant before replacement.

For a broad range of septa and caps, septa at septacaps.com is a good starting point. If you are buying for a specific instrument, the instrument vendor’s own consumables catalogue is worth checking for compatibility. Agilent chromatography consumables and columns and Thermo Fisher chromatography and sample handling both publish detailed septa specifications, including temperature ratings and bleed data, that are useful for cross-checking.

For method development, Restek technical resources for chromatography has a good set of application notes on septa bleed and inlet maintenance. And if you are specifying vials and caps for a new autosampler, the autosampler vial page at autosamplerhplc.com has a useful compatibility table.

What I would do differently

If I were setting up a lab from scratch tomorrow, I would standardize on one septa for routine work and one for high-temperature work, and I would not let anyone buy a third type without a justification. Two types cover 95% of the methods I have run in 15 years. The third type, the “just in case” septa that sits in the back of the cupboard, is usually the one that fails when someone grabs it in a hurry.

I would also put a go/no-go gauge on the crimper and a torque wrench on the screw capper, and I would train every new tech on both before they touch a vial. Most septa failures I have investigated were mechanical, not chemical. The septa was the victim, not the cause.

Takeaway

Septa failures are predictable. Coring comes from blunt needles and misalignment. Bleed comes from running a standard septa above its temperature rating. Leaks come from bad crimps and over-torqued nuts. Chemical attack comes from using silicone on solvents it cannot handle. Thermal degradation comes from ignoring the temperature rating. Contamination comes from buying on price instead of specification. Physical damage comes from all of the above. Fix the mechanical setup first, match the septa material to the solvent and the temperature, and replace on an injection count rather than on failure. That is the whole method, and it will cut your septa-related downtime to near zero.

References & further reading

Related resources from our lab consumables network:

External references and standards: