I have watched a lot of septa fail in ways that had nothing to do with the septum itself. A lab calls and says the septa are bad because the GC is showing ghost peaks, or the HPLC is losing pressure, or the autosampler is coring needles. Nine times out of ten, the septum is doing exactly what the wrong septum does in that application. The rest of the time it is damaged on installation, over-torqued, run past its temperature rating, or punched far more times than the material can survive. This article is what I tell those labs: the failure modes I see most, what actually causes them, and the numbers to check before you order another box.
A septum is a soft, compressible barrier that has to seal a pressurized or heated headspace while a needle passes through it repeatedly. Those two jobs fight each other. Softness gives you a good seal at low torque and easy needle penetration. Hardness gives you puncture resistance and low coring, but needs more compression to seal. Every septum failure comes from picking the wrong point on that trade-off, or from pushing a material past its rated conditions.
The common failure modes are:
- Coring and fragmentation: a plug of septum is punched into the vial or injector port.
- Ghost peaks and carryover: bleed from the septum shows up as extraneous peaks or elevated baseline.
- Leaks and pressure loss: the seal fails at the injection port or the vial headspace.
- Needle damage and bending: a hardened or misaligned septum deflects the needle.
- Adhesion and sticking: the septum welds to the vial or the injector body.
- Swelling and chemical attack: the elastomer absorbs solvent and loses its shape.
- Septum push-through: the disc is forced into the vial neck during capping or injection.
Each one has a root cause you can fix. Let me go through them the way I would on a plant floor.
Coring: the failure everyone blames first
Coring is when the needle punches a small disc out of the septum instead of parting the elastomer. The disc either falls into the vial (contaminating the sample and jamming the needle) or stays lodged in the injector port, where it eventually blocks flow or gets pushed into the column.
What causes it
The biggest cause is a needle that is too large for the septum, or a septum that is too hard for the needle. A standard 23-gauge autosampler needle has an outside diameter around 0.64 mm. That is fine for most silicone and silicone/PTFE septa. Step up to a 21-gauge needle (about 0.81 mm) and the puncture stress on a hard rubber septum rises enough that coring becomes routine after a few hundred injections.
The second cause is a dull or burred needle. A needle that has been used for thousands of injections develops a slightly flattened tip. It no longer cuts; it tears. I have seen labs replace a $40 box of septa three times before someone thought to look at the needle under a loupe.
The third cause is a needle that does not travel straight. Bent needle guides on an autosampler, or a vial that sits off-center in the tray, mean the needle enters at an angle and slices the septum instead of piercing it.
How to prevent it
- Match needle gauge to septum hardness. For 21-gauge needles, use a softer silicone septum or a pre-slit design.
- Inspect needles every 2,000 to 5,000 injections depending on sample matrix. Replace at the first sign of a burr.
- Use pre-slit septa for high-throughput methods. A slit septum parts around the needle and cuts coring dramatically.
- Check vial seating. A vial that rocks in the tray will core even a good septum.
If you are running a high-count autosampler method, the autosampler vial geometry matters as much as the septum. A vial with a true 11 mm crimp neck and a flat sealing surface seats the septum squarely and reduces angled punctures.
Ghost peaks and septum bleed
This is the failure that wastes the most instrument time, because people chase it through the column and the detector before they suspect the septum. Septum bleed shows up as a rising baseline during a temperature ramp, or as repeatable extraneous peaks at specific retention times.
What causes it
Every elastomer outgasses something when heated. Silicone septa are the worst offenders above about 250 °C; they release low-molecular-weight siloxanes that land on the column and elute as peaks. General-purpose silicone with no barrier layer will bleed noticeably even at 200 °C if you hold there long enough.
The second cause is a septum that has been punctured many times. Each puncture creates fresh surface area, and fresh surface area bleeds. A septum that was clean at injection 1 can be a bleed source by injection 500.
The third cause is a septum that was never conditioned. Many silicone septa carry residual volatiles from manufacturing. If you install them straight from the box into a hot injector port, those volatiles come off into your chromatogram.
How to prevent it
- Use a PTFE-faced septum for high-temperature work. The PTFE barrier faces the sample and blocks siloxane migration into the vial headspace.
- Condition new septa before use: heat them at 10 to 20 °C above your maximum operating temperature for 1 to 2 hours in a clean oven, then cool in a desiccator.
- Replace injector-port septa on a schedule tied to injection count, not to calendar time. For high-temperature GC, that may be every 50 to 100 injections.
- Run a blank injection after any septum change to confirm the bleed is gone before you trust the next sample.
For method development, the Restek technical resources have good guidance on bleed profiles by septum type. Their data lines up with what I have measured: PTFE-faced silicone bleeds far less than plain silicone above 250 °C.
Leaks at the injector port
A leaking injector-port septum shows up as a dropping inlet pressure, erratic retention times, or a split ratio that will not hold. On a split/splitless inlet, a small leak changes the actual split ratio and quietly shifts your results.
What causes it
Under-torquing is the usual culprit. The septum has to be compressed enough to seal against the inlet body. Too little compression and carrier gas escapes past the septum. Too much compression and you crush the septum, which causes coring and shortens its life. The window is narrower than people think.
On a typical 1/4-inch injector-port nut, the correct torque is often in the 10 to 15 in-lb range for a fresh septum. That is hand-tight plus a small extra turn with a torque wrench, not a full gorilla turn. I keep a small 0-50 in-lb torque wrench on the bench for exactly this reason. It ends the argument.
The second cause is a septum that has taken a compression set. Silicone under heat and pressure slowly stops springing back. A septum that has been in place for months may look fine and still leak because it no longer presses against the inlet body.
How to prevent it
- Torque injector-port septa to the manufacturer’s spec. For most inlets that is 10 to 15 in-lb; confirm against your instrument manual.
- Replace on injection count. For a busy inlet, that is often every 100 to 200 injections.
- Use a septum retainer nut with a defined stop so you cannot over-tighten.
- Leak-check the inlet after every septum change, not just when results look wrong.
The same logic applies to vial headspace. A crimp cap that is not crimped to the right dimension leaves the septum loose, and the vial leaks. For 11 mm crimp necks, the cap should be crimped so the seal sits flat and the septum is compressed evenly. A crimper that is out of adjustment will produce a run of leaking vials that looks exactly like a bad septum batch.
Needle damage and bending
A bent needle is expensive and it is almost always a septum or alignment problem, not a needle problem. When a needle hits a hard septum at an angle, the shaft flexes. Do that a few thousand times and the needle bends or snaps.
What causes it
Hard septa are the main cause. A septum with a high durometer reading (harder rubber) resists penetration. If the autosampler pushes the needle through with a fixed force, the needle takes the load. Silicone septa are soft and forgiving; some rubber and laminated septa are not.
Misalignment is the second cause. If the vial is not centered under the needle, the needle enters off-axis and the septum deflects it sideways.
How to prevent it
- Use soft silicone or pre-slit septa for autosampler work, especially with fine needles.
- Verify vial centering in the tray. A quick check with a straight edge across the tray catches most misalignment.
- Check needle straightness with a flat surface every few thousand injections.
- Do not mix septum types within a run. A hard septum in a tray of soft ones will be the one that bends a needle.
Adhesion and sticking
Sometimes the septum welds itself to the vial or the injector body. You go to change it and it tears, leaving a ring of elastomer stuck to the metal.
What causes it
Heat and time. Silicone in contact with a hot metal surface for long periods can bond slightly. The problem is worse with certain sample matrices and with septa that have no barrier layer.
The second cause is chemical attack. Some solvents soften the elastomer until it flows into surface imperfections on the metal and then sets there.
How to prevent it
- Use a PTFE-faced septum so the PTFE contacts the sample and the metal, not the raw silicone.
- Do not leave septa in a hot inlet for weeks. Replace on schedule.
- Clean the inlet body when you change septa. A residue ring left behind will grab the next septum too.
Swelling and chemical attack
Not every septum failure is mechanical. Some are chemical. A septum that swells in a solvent loses its dimensions, seals poorly, and cores easily.
What causes it
Silicone is broadly resistant but swells in nonpolar solvents like toluene, hexane, and some chlorinated solvents. The swelling is reversible if the exposure is short, but a septum sitting in solvent vapor for days will take a permanent set.
The second cause is an incompatible elastomer. If you are running a strongly nonpolar matrix, a silicone septum may not be the right choice. A PTFE-faced or a fluoroelastomer septum handles those matrices better.
How to prevent it
- Match elastomer to matrix. Silicone for aqueous and polar; PTFE-faced for broad solvent compatibility; fluoroelastomer for aggressive nonpolar work.
- Do not store vials capped for long periods with aggressive solvents. Cap and run.
- Check septa for swelling during method development. A swollen septum is a leak waiting to happen.
If you want the full material selection matrix, the septa range at septacaps.com lists durometer, temperature rating, and recommended matrices for each type. That is the table I use when a lab sends me a solvent list and asks what to buy.
Septum push-through
Push-through is when the septum disc is forced into the vial neck during capping or injection. It shows up as a vial that will not seal, or a needle that hits the septum at the wrong depth.
What causes it
Under-sized septa in an over-sized cap, or a capping tool that is out of adjustment. If the crimp or screw cap does not hold the septum against the vial lip, the needle pushes the septum inward instead of piercing it.
How to prevent it
- Use septa sized to the cap. For 11 mm crimp caps, the septum diameter should match the cap seat, typically 11 mm.
- Check the capping tool against the manufacturer’s setting. A crimper that is too loose lets the septum move.
- For screw caps, do not over-tighten. Over-tightening distorts the septum and can push it off-center.
A troubleshooting checklist you can run in 10 minutes
When a lab calls with a septum problem, this is the sequence I run. It catches most failures before anyone orders new parts.
Step 1: Confirm it is the septum
- Run a blank injection. If the ghost peaks persist with a fresh septum, the problem is upstream (carrier gas, liner, column).
- Swap one septum and re-run. If the symptom moves with the septum, you have your answer.
- Check inlet pressure. A drop points to a leak; a rise points to a blockage.
Step 2: Check the installation
- Torque the injector-port nut to spec (10 to 15 in-lb for most 1/4-inch inlets).
- Verify the septum sits flat and centered.
- Leak-check the inlet after the change.
Step 3: Check the needle
- Inspect the tip under a loupe for burrs.
- Confirm needle gauge matches septum hardness.
- Check needle straightness on a flat surface.
Step 4: Check the vial and cap
- Confirm vial centering in the tray.
- Check crimp or screw-cap torque and seat.
- Look for swollen or distorted septa.
Step 5: Check the operating conditions
- Confirm the septum temperature rating exceeds your maximum inlet temperature by at least 20 °C.
- Confirm the elastomer is compatible with your sample matrix.
- Confirm the septum is being replaced on injection count, not calendar time.
Selection guide: matching septum to application
Here is the decision table I use. The numbers are the ones I check first.
| Application | Recommended septum | Key numbers to check |
|---|---|---|
| Routine HPLC, aqueous/polar mobile phase | Silicone, 45 to 55 Shore A | Temperature rating ≥ 80 °C; puncture life ≥ 500 at 23-gauge |
| High-temperature GC, inlet up to 300 °C | PTFE-faced silicone | Temperature rating ≥ 300 °C; bleed below 1 pA at 300 °C hold |
| High-throughput autosampler, > 1,000 injections/day | Pre-slit silicone | Puncture life ≥ 1,000; needle gauge 23; coring resistance rated for pre-slit |
| Aggressive nonpolar solvents (toluene, hexane) | PTFE-faced or fluoroelastomer | Swelling < 10% after 24 h immersion; temperature rating ≥ 200 °C |
| Headspace GC | PTFE-faced silicone, 20 mm for 20 mm crimp | Temperature rating ≥ 250 °C; leak-tight at 1.5 bar headspace pressure |
The vial septa catalog at vialsepta.com has these specs per part number, so you can match the table to a SKU without guessing. For the HPLC side, the HPLC vial range at hplcgc.com covers the 9 mm and 11 mm formats most labs run.
Operational precautions that actually extend septum life
These are the habits that separate labs with low septum consumption from labs that buy a box a month.
- Condition before use. Heat new septa at 10 to 20 °C above max operating temperature for 1 to 2 hours, then cool in a desiccator. This removes manufacturing volatiles before they reach your column.
- Replace on injection count. Track injections per septum. For high-temperature GC that may be 50 to 100; for routine HPLC it may be 500 to 1,000.
- Torque to spec. Keep a torque wrench at the bench. 10 to 15 in-lb for most injector-port nuts.
- Do not reuse septa. A septum that has been punctured and removed has taken a compression set and will not reseal reliably.
- Store sealed and dry. Silicone absorbs moisture and airborne contaminants. Keep the box closed and store away from solvent vapors.
- Match needle gauge to septum. 23-gauge is the safe default for autosamplers; step to pre-slit if you need 21-gauge or larger.
- Check the capping tool. A crimper out of adjustment produces leaking vials that look like bad septa.
For method-level guidance on sample handling and consumable compatibility, the Thermo Fisher chromatography and sample handling documentation and the Agilent GC columns and accessories pages both cover inlet and septum specifications in detail. For regulated environmental work, the U.S. EPA SW-846 test methods set the QC expectations that drive a lot of septum replacement schedules in contract labs.
Standards and dimensions worth knowing
A few numbers come up again and again. Keep these on a card at the bench.
- GPI 9 mm and GPI 11 mm are the two common screw-thread vial finishes in HPLC. The septum diameter must match the cap seat.
- 11 mm crimp necks use an 11 mm septum and a 11 mm aluminum seal. Crimp height and diameter are set by the crimper, not by hand.
- 20 mm crimp necks are standard for headspace vials and use a 20 mm septum.
- ISO 4793 covers laboratory glassware tolerances; USP <1058> covers analytical instrument qualification and is a good reference for why a leak-free inlet matters.
- 23-gauge needle OD is about 0.64 mm; 21-gauge is about 0.81 mm.
What I recommend
If you are setting up a method or fixing a recurring septum problem, here is the concrete recommendation I give.
For routine HPLC and autosampler work, use silicone septa at 45 to 55 Shore A, pre-slit if you run more than 500 injections per vial position, sized to your 9 mm or 11 mm cap. Torque injector-port nuts to 10 to 15 in-lb. Replace on injection count, not calendar time. For high-temperature GC above 250 °C, use PTFE-faced silicone rated to at least 300 °C and condition new septa at 270 to 300 °C for 1 to 2 hours before installation. For aggressive nonpolar matrices, move to a PTFE-faced or fluoroelastomer septum and check swelling after 24 hours immersion before committing to a production run.
None of these numbers are exotic. They are the ones that separate a septum that lasts 1,000 injections from one that fails at 50. The septum is a consumable, but it is not a random one. Match the material to the application, install it to the right torque, and replace it on a count. Do that and most of the failures in this article stop happening.
The takeaway: a septum failure is almost never a bad batch. It is a mismatch between material, needle, torque, and temperature, and every one of those is a number you can check before the next run.
References & further reading
Related resources from our lab consumables network:
- Septa at septacaps.com
- Vial Septa at vialsepta.com
- Hplc Vial at hplcgc.com
- Autosampler Vial at autosamplerhplc.com
External references and standards:
