What a septum actually does, and why it fails--Aijiren HPLC Vials

I still keep a shoebox of failed septa under my desk. Every one of them cost somebody a run: a double-injection from a coring problem on a 96-well plate, a ghost peak at 11.2 minutes traced to a bleed-heavy silicone disc, a vial that vented solvent into a tray because the cap was torqued to 28 in-lb on a 9 mm thread. That box is the reason I now treat septa as a specification, not a commodity line item. The part costs cents; the data it protects costs thousands. This guide is what I wish somebody had handed me the first time I had to specify septa across a lab.

A septum is a compressible elastomeric barrier that lets a needle pass through, seals behind it, and keeps the vial headspace isolated from the room. That sounds simple. In practice the part has to satisfy four competing requirements at once, and every material choice trades one against another.

Resealability: the puncture must close well enough to hold headspace pressure and prevent solvent loss during a 20 to 60 minute autosampler queue.
Chemical inertness: no extractables that show up as ghost peaks, and no swelling or degradation in the solvent you are actually using.
Mechanical durability: enough puncture life to survive a full plate or tray without coring, and enough hardness to resist needle bending.
Thermal stability: hold shape and seal at autosampler tray temperatures and at injection port temperatures up to 350 °C for GC inlet septa.

Failure modes fall into a small number of buckets, and knowing which one you are fighting tells you which specification to change:

Coring. The needle punches a plug of elastomer into the vial. You see it as an intermittent high peak, a blocked needle, or a visible fragment floating in the sample. Usually caused by a needle that is too large for the septum, a blunt or burred needle, or a septum that is too hard for the needle gauge.
Bleed. Elastomer breakdown products elute into the column. Silicone septa are the classic offender in GC, showing up as siloxane peaks at characteristic retention times. Bleed gets worse with inlet temperature and with septum age.
Extractables and ghost peaks. Plasticizers, antioxidants, and curing agents leach into the sample or the headspace. In HPLC this shows as baseline drift or an unidentified peak that follows the septum lot, not the sample.
Back-pressure and blow-by. A septum that is too soft extrudes into the cap opening under headspace pressure, or the needle pushes the septum down and the seal fails. Common with low-durometer silicone at high headspace pressure.
Solvent loss and evaporation. A reseal that is marginal lets volatile analytes escape between injections. You see it as declining response across a long sequence, worse for low-boiling analytes.

The first thing I check when a customer reports an intermittent problem is whether the septum is the right type for the job, not whether it is defective. Most “bad lot” complaints I have investigated turned out to be a material mismatch. If you are specifying from scratch, start from the septa catalog at septacaps.com and work backward from your instrument and method, not forward from price.

The material map: what each elastomer is actually good for

There are five families you will encounter in a working lab. I have ranked them by where they earn their place.

Silicone (polydimethylsiloxane)

The default for general-purpose septa because it is soft, reseals well, and tolerates a wide temperature range, roughly -60 °C to +200 °C in continuous service and higher for short excursions. The problem is bleed. Silicone is the source of the siloxane background that plagues sensitive GC methods. Plain silicone is fine for HPLC vials and for room-temperature headspace work; it is a poor choice for trace GC at high inlet temperature unless it is a low-bleed grade. A 55 Shore A silicone disc is a reasonable general-purpose hardness.

Silicone with a PTFE face

This is the workhorse for HPLC and for anything where the sample touches the septum. A thin PTFE film, typically 0.002 to 0.005 in (50 to 125 µm), is bonded to the sample side. The PTFE provides chemical resistance to almost everything and stops extractables from reaching the sample; the silicone underneath provides the reseal. The trade-off is that the PTFE face is what the needle must pierce, and a thick PTFE layer increases coring risk and needle drag. For HPLC autosamplers running acetonitrile, methanol, and buffered mobile phases, a silicone/PTFE septum is what I specify nine times out of ten. If you are matching septa to a specific HPLC vial, confirm the PTFE is on the sample side, because a reversed laminate is a silent extractables source.

PTFE-faced silicone, pre-slit

A slit cut partway or fully through the septum. The slit lets the needle enter without coring and reduces the force needed to pierce, which matters on autosamplers with light needle drives and on high-throughput systems doing thousands of injections. The cost is a slightly weaker reseal, so pre-slit septa are a poor choice for long headspace equilibration or for methods where solvent loss between injections matters. I use pre-slit for short HPLC sequences and avoid it for headspace.

Butyl and bromobutyl rubber

Low gas permeability and good reseal make butyl the standard for pharmaceutical vial closures and for headspace where you need to hold a pressure. It is not a general HPLC septum; its extractables profile is worse than silicone/PTFE and it is harder, so coring risk rises with large needles. Temperature ceiling is around 120 to 130 °C.

EPDM and fluorocarbon (FKM)

EPDM handles aqueous and polar solvents and steam better than silicone, with a service range roughly -50 °C to +150 °C. Fluorocarbon (Viton-type) is the choice for aggressive solvents and higher temperatures, up to about 200 to 230 °C, but it is hard, expensive, and reseals less well, so it is usually reserved for GC inlet septa on methods that destroy silicone. Neither is a first choice for routine HPLC.

GC inlet septa versus autosampler vial septa: do not conflate them

These are different parts solving different problems, and buying the wrong one is the most common specification error I see.

A GC inlet septum sits in the injection port, sees the full inlet temperature (often 250 to 300 °C, sometimes 350 °C), and is pierced repeatedly by the syringe as it injects. It must resist thermal degradation and bleed, and it is usually a molded disc, 11 mm or 12.5 mm diameter, held by a retaining nut torqued to a specified value. Inlet septa are consumables with a defined injection life, commonly 50 to 100 injections for a standard silicone septum and more for a low-bleed or pre-conditioned grade. Replace on a schedule, not on failure.

An autosampler vial septum sits in a crimp or screw cap on a 2 mL vial (or a 4 mL, 10 mL, or 20 mL headspace vial), sees tray temperature (usually 4 to 40 °C), and is pierced once per injection. It must resist extractables and hold a seal for the queue duration. These are the septa you buy by the thousand. The vial septa range at vialsepta.com covers the common 9 mm, 11 mm, and 20 mm formats.

Matching septum to cap and vial format

The septum has to fit the cap and the vial neck, and the dimensions are not interchangeable. The table below is the working reference I use.

Format	Nominal septum diameter	Thickness (typical)	Typical use
9 mm short screw	9 mm	1.0 mm	2 mL HPLC autosampler vials
11 mm crimp	11 mm	1.0 to 1.5 mm	2 mL GC and HPLC vials, headspace
13 mm	13 mm	1.5 mm	Larger vials, some headspace
20 mm	20 mm	3.0 mm	20 mL headspace vials
GC inlet	11 mm or 12.5 mm	3 to 5 mm molded	Injection port

The cap thread standard matters too. For screw caps, the GPI 8-425, GPI 13-425, and GPI 15-425 glass finishes are the common ones on 2 mL vials, and a 9 mm cap is not the same as a 13-425 cap even though both look small in the hand. Buying a 9 mm septum for a 13-425 vial gives you a loose seal and a leak you will chase for a week. Check the glass finish printed on the vial carton before you reorder. When in doubt, match the autosampler vial format to the cap and septum as a set.

Selection criteria with numbers

Here is the decision logic I run through with a customer. It is ordered so that the first question eliminates the most options.

1. What temperature does the septum see?

Below 40 °C (autosampler tray, refrigerated tray): almost any material works. Choose on extractables, not temperature.
40 to 120 °C (headspace, some GC inlets): silicone, butyl, or EPDM. Check bleed if GC.
120 to 200 °C (GC inlet, hot methods): low-bleed silicone or fluorocarbon.
Above 200 °C (high-temperature GC inlet up to 350 °C): fluorocarbon or a dedicated high-temperature inlet septum. Plain silicone will degrade and bleed.

2. What are you injecting, and how sensitive is the detector?

If you are running trace analysis on a mass spectrometer or an electron capture detector, bleed and extractables dominate. Specify low-bleed silicone or fluorocarbon for the inlet, and silicone/PTFE for the vial. If you are running a UV detector at 254 nm on a 1 mg/mL sample, a standard silicone/PTFE septum is fine and you should not pay a premium for low-bleed. Match the septum to the detection limit, not to the catalog’s best grade.

3. What needle gauge and how many injections?

Needle outside diameter drives coring risk. A 22 gauge (0.72 mm OD) needle is gentle; a 26 gauge (0.46 mm OD) is finer and cores less but bends more easily; a large-bore or blunt needle cores aggressively. For high-throughput work, a pre-slit septum or a softer durometer reduces coring. Puncture life is a real number you can plan around: a standard silicone/PTFE 9 mm septum holds a reseal for roughly 1 to 3 punctures in the same spot before leakage becomes measurable, which is why autosamplers move the needle or use a fresh vial per injection. For GC inlet septa, plan on 50 to 100 injections per septum for standard grades and replace on schedule.

4. What is the headspace pressure and queue length?

Headspace vials pressurized to 10 to 20 psi and held for a 30 to 60 minute equilibration put real load on the septum. Butyl or a thicker silicone septum holds pressure better than a thin pre-slit disc. For long HPLC queues with volatile analytes, a non-slit silicone/PTFE septum with a good crimp or a properly torqued screw cap is the safer choice.

5. What is the cap torque specification?

This is where operators cause most of their own problems. Over-torquing a screw cap deforms the septum, thins it under the cap, and either splits it or drives it into the vial neck. Under-torquing leaves a leak. For 9 mm short screw caps, a hand-tight plus a quarter turn is the practical rule, and if you use a torque tool, the working range is roughly 8 to 12 in-lb for a 9 mm cap and 15 to 20 in-lb for an 11 mm crimp cap. Crimp caps are set by the crimper, not by hand: a properly adjusted crimper produces a uniform crimp with the septum compressed but not cut. I have seen a mis-set crimper split a hundred septa in a row and the operator blame the supplier.

A decision checklist you can hand to a buyer

This is the version I give to procurement so they can qualify a supplier without a chemist in the room. Each line has a number or a yes/no.

Format: 9 mm, 11 mm, 13 mm, 20 mm, or GC inlet (11/12.5 mm). Confirm against the vial glass finish (GPI 8-425, 13-425, 15-425) and the cap.
Material: silicone, silicone/PTFE, pre-slit silicone/PTFE, butyl, EPDM, or fluorocarbon. PTFE face on the sample side if extractables matter.
Temperature rating: at least 20 °C above your maximum service temperature. Do not run a 200 °C septum at 200 °C continuously.
Hardness: 45 to 55 Shore A for general autosampler use; harder for high-pressure or large-needle applications.
Thickness: 1.0 mm for 9 mm caps, 1.5 mm for 11 mm crimp, 3.0 mm for 20 mm headspace.
Puncture life: stated by the supplier in injections or punctures. If it is not stated, ask. A supplier who cannot give you a number is guessing.
Extractables: request a lot certificate and, for trace work, an extractables profile or a blank-injection chromatogram.
Bleed: for GC, a bleed specification at your inlet temperature, not at 250 °C if you run 300 °C.
Lot traceability: lot number on the bag. This is non-negotiable for regulated work.
Packaging: individually bagged or cleanroom-packed for trace work; bulk bags are fine for general HPLC.

If you want an independent sanity check on bleed and inertness before you commit a lot, the application notes from instrument and column vendors are the fastest route. Agilent’s GC columns and accessories documentation covers inlet septum bleed and replacement intervals in detail, and Restek’s technical resources have good material on septa selection for trace analysis. For sample handling and vial compatibility, Thermo Fisher’s chromatography and sample handling pages are a useful cross-reference. If your lab works under a regulated method, the U.S. EPA SW-846 test methods give you the acceptance criteria you need to document.

Operational precautions that prevent most failures

Specification gets you a good part. Handling keeps it good. These are the practices that have saved me the most rework.

Storage and shelf life

Silicone and rubber septa age. Store them in the original sealed bag, away from UV, ozone sources (motors, copiers), and solvent vapors. Most manufacturers give a shelf life of 2 to 3 years from manufacture; I treat anything older than 18 months as suspect for trace work and run a blank before using it. A bag of septa left open on a bench next to an open solvent bottle will absorb vapors and show ghost peaks. I have proved this to myself more than once.

Handling

Do not handle septa with bare fingers if you are running trace analysis. Skin oils and plasticizers transfer. Use powder-free nitrile gloves and forceps. Do not pre-pierce a septum “to make it easier” unless it is a pre-slit part; a hand-pierced hole does not reseal the way a molded slit does.

Installation

For crimp caps, seat the septum flat in the cap, set the crimper to the vial’s glass finish, and check the crimp on the first vial of every batch. For screw caps, torque to the range above and check that the septum is not distorted. For GC inlets, install the septum with the correct orientation (PTFE face toward the sample, if the part has one), and torque the retaining nut to the instrument’s specification, typically hand-tight plus a small fraction of a turn; over-tightening deforms the septum and causes leaks and coring.

Replacement schedule

Replace GC inlet septa on an injection count, not on failure. A septum that has started to bleed has already contaminated your column and your baseline. For autosampler vials, use a fresh septum per vial and do not re-pierce a used one more than the supplier’s stated reseal limit allows. For long queues, cap and septum the vials immediately before loading, not the day before.

Troubleshooting checklist

When a run misbehaves and you suspect the septum, work this list in order. It is ordered by how often each cause turns out to be the answer in my experience.

Ghost peak that follows the septum lot, not the sample: extractables. Switch to silicone/PTFE with the PTFE on the sample side, or request a lower-extractables grade. Run a blank injection of mobile phase through a fresh septum to confirm.
Siloxane peaks in GC at high inlet temperature: bleed. Lower the inlet temperature if the method allows, switch to low-bleed or fluorocarbon, and shorten the replacement interval.
Intermittent high peak or blocked needle: coring. Check needle gauge and condition, switch to pre-slit or softer septum, and inspect the vial for a fragment.
Declining response across a sequence: solvent loss through a marginal reseal. Check cap torque, switch to a non-slit septum, and shorten the queue.
Leak at the cap: wrong septum diameter for the glass finish, or over-torqued cap deforming the septum. Verify GPI finish and re-torque to spec.
Septum pushed into the vial: under-torqued cap or oversized needle. Re-torque and check needle alignment.
Baseline drift on HPLC: plasticizer leaching from a non-PTFE septum, or a reversed laminate. Confirm PTFE orientation and switch lots.
Split or cut septum after crimping: crimper out of adjustment. Re-set the crimper and check the first five vials of each batch.

What I recommend, by workflow

Here is the short version, with concrete parameters, for the four workflows I get asked about most.

Routine HPLC, UV detection, 2 mL vials: 9 mm silicone/PTFE septum, 1.0 mm thick, PTFE on the sample side, 45 to 55 Shore A, non-slit. Torque 9 mm screw caps to 8 to 12 in-lb. Replace the septum with each new vial.
High-throughput HPLC autosampler: 9 mm pre-slit silicone/PTFE, 1.0 mm, for reduced coring and lower needle force. Keep queue length under the reseal limit and cap vials immediately before loading.
GC with a sensitive detector (MS, ECD): low-bleed silicone or fluorocarbon inlet septum, 11 or 12.5 mm, rated at least 20 °C above inlet temperature. Replace every 50 to 100 injections and log the count. Use silicone/PTFE vial septa for the samples.
Headspace: butyl or thicker silicone septum, 20 mm format for 20 mL vials, non-slit, rated for the pressurization and equilibration time. Check cap torque and crimp before every batch.

The single change that has saved me the most trouble is buying septa against a written specification (format, material, thickness, hardness, temperature rating, puncture life, lot traceability) and recording the lot number against each run. That one habit turns a mystery ghost peak into a two-minute check, and it turns a supplier conversation from a complaint into a specification review. The septa catalog, the vial septa range, the HPLC vial selection, and the autosampler vial options in our network are organized around those specifications, so you can match part to method without guesswork.

Treat the septum as a specified component with a temperature rating, a hardness, a thickness, and a puncture count, and most of the failures in that shoebox under my desk stop happening. The part is still cheap. The specification is what makes it reliable.

References & further reading

Related resources from our lab consumables network:

External references and standards: