What a septum actually does in an HPLC or GC workflow--Aijiren HPLC Septa

I have watched a lab lose three weeks of stability data because someone grabbed the wrong bag of septa from the cabinet. The vials were fine, the method was validated, the autosampler was calibrated. The septa were a generic silicone laminate that off-gassed silicone volatiles into the headspace, and every early time point came back with ghost peaks at m/z 73, 147, and 221 on the GC-MS confirmation run. That is the kind of failure this guide exists to prevent.

Septa are the cheapest component in the entire HPLC stack and the one most likely to quietly wreck a run. A bag of 100 costs less than a single hour of instrument time, yet it sits between your sample and your detector, gets punctured dozens of times, and is exposed to whatever your mobile phase, your vial, and your autosampler needle can throw at it. Choosing the right one is not complicated, but it does require matching four or five concrete parameters to your actual workflow. Below is how I specify septa for a lab, in the order I work through the decision.

A septum has three jobs, and they conflict with each other. It must seal the vial so the contents do not evaporate or oxidize between preparation and injection. It must allow a needle to pass through without coring, tearing, or dragging material into the sample. And it must stay chemically inert toward the sample and the headspace above it.

The conflict is the interesting part. The softest, most resealable materials (silicone) are the most permeable and the most prone to off-gassing. The most chemically inert materials (PTFE) are stiff, do not reseal well after multiple punctures, and can shed particles. Most real septa are laminates: a thin PTFE face on the sample side bonded to a silicone backing for resealability. That is why a 1 mm silicone / 0.05 to 0.13 mm PTFE laminate is the default for most HPLC work, and why you see the same construction sold under a dozen brand names.

For HPLC specifically, the septum is doing less work than in GC headspace. You are not heating the vial to 80 °C, so permeability and thermal off-gassing matter less. What matters more is coring (particles that land in the sample and plug the needle or the column frit), needle drag, and whether the septum stays seated in the cap under the autosampler’s gripper. For headspace GC and GC-MS, the calculation flips: thermal stability and low bleed become the dominant criteria.

Material selection: the four families you will actually encounter

Silicone / PTFE laminate

This is the workhorse. A typical 11 mm septum is 1.0 mm silicone with a 0.06 mm PTFE film on the sample-facing side. Temperature rating is usually -40 °C to 200 °C for the silicone layer, though the PTFE face is good to well beyond that. Puncture life is typically 3 to 10 punctures before reseal integrity degrades, depending on needle gauge and how the autosampler positions the needle. Use it for reverse-phase HPLC, autosampler trays that run overnight, and any method where the same vial gets injected more than once.

Failure mode I see most: the PTFE film delaminates at the puncture site after 15 to 20 injections on a 22-gauge needle, and you get a small flap of PTFE sitting in the vial. It rarely causes an immediate problem, but it will eventually find its way into a needle or a frit. If your method injects a single vial more than 10 times, either change vials or move to a pre-slit septum.

Pure PTFE (and PTFE-faced silicone with no exposed silicone)

Pure PTFE septa are the choice for aggressive solvents, high-temperature headspace, and trace analysis where any silicone bleed is unacceptable. They are chemically inert to almost everything in a normal lab, including DMSO, THF, dichloromethane, and concentrated acids. The trade-off is resealability. A pure PTFE septum may only reseal reliably for 1 to 3 punctures, and after that it can leak. They are also stiffer, which means the cap has to be crimped or screwed correctly or the septum will not seat.

If you are running headspace GC-MS at 80 °C or above and seeing siloxane background at m/z 73, 147, 207, and 281, switching to a PTFE-faced septum with a low-bleed silicone backing is the first thing to try. The second is baking the septa at 100 °C for an hour before use, which drives off residual volatiles. I have done this on more than one method and it is worth the hour.

Butyl rubber and bromobutyl

Butyl is the standard for pharmaceutical vial closures because it has low moisture and gas permeability and is available in pharmacopeial grades. It is not usually the first choice for analytical HPLC, but it shows up in vial closures for lyophilized products and in some GC applications where a very low-bleed seal is needed. Bromobutyl has better chemical resistance than plain butyl and is common in 13 mm and 20 mm crimp closures for parenteral products.

The failure mode here is extractables. Butyl contains curing agents, fillers, and antioxidants that can leach into the sample. If you are running trace analysis on a butyl-closed vial, run a blank first and look for sulfur-containing compounds on the LC-MS.

Silicone with a slit (pre-slit)

A pre-slit septum has a small slit cut through it so the needle does not have to core through virgin material. This extends puncture life dramatically (I have run 50+ injections on a single pre-slit septum without reseal failure) and reduces coring. The trade-off is that a slit is a leak path. For volatile samples or long autosampler queues, a pre-slit septum can lose more sample to evaporation than a solid one. Use pre-slit for high-throughput aqueous methods; avoid it for volatile analytes or headspace work.

Matching septum dimensions to your vial and cap

Dimension mismatches are the single most common cause of autosampler crashes and needle bends I have dealt with. The septum has to fit the cap, and the cap has to fit the vial, and the whole assembly has to fit the autosampler’s gripper. Get one of those wrong and you find out at 2 a.m.

The common sizes:

11 mm septa for standard 2 mL HPLC vials with 11 mm crimp or snap caps. This is the most common HPLC size.
13 mm septa for 4 mL and some 2 mL vials, and for many GC headspace vials.
20 mm septa for 20 mL and 22 mL headspace vials, standard in GC headspace work.
22 mm and 24 mm septa for larger vials and some preparative applications.
9 mm screw-thread septa (often called 9 mm short-thread) for the narrow-neck screw-top vials common on Agilent and Waters autosamplers.

The GPI 400 thread finish (13-425) and the GPI 18 mm finish (for 9 mm screw vials) are the standard thread references you will see in vendor catalogs. If your cap is a 13-425, you need a 13 mm septum, full stop. Do not try to force an 11 mm septum into a 13 mm cap; it will sit off-center, the needle will miss the center, and you will bend a needle within a week.

Thickness matters too. Standard HPLC septa are 1.0 mm thick. Headspace septa are often 3.0 mm to survive the higher puncture forces and the thermal cycling. A 3 mm septum will not fit a standard 11 mm crimp cap designed for 1 mm. Check the cap’s internal depth before you buy.

Temperature, pressure, and chemical compatibility

For HPLC, temperature is rarely the binding constraint. Room-temperature autosamplers and 40 °C column compartments do not stress a silicone/PTFE septum. The binding constraints are chemical compatibility and puncture life.

For GC headspace and any method that heats the vial, temperature becomes the primary constraint. Silicone is good to about 200 °C; PTFE-faced laminates are good to similar temperatures on the face but the silicone backing limits the assembly. If you are running headspace at 150 °C or above, use a PTFE-faced laminate rated for the temperature and pre-condition the septa by heating them to the method temperature for 30 minutes before use. This drives off residual volatiles that would otherwise show up as baseline noise.

Pressure matters for high-pressure applications and for any vial that gets heated. A sealed 20 mL headspace vial at 150 °C can see internal pressures of several bar. The septum has to hold that without extruding through the cap. That is why headspace septa are thicker and why the crimp has to be correct. A loose crimp on a heated vial will vent through the septum and give you a low response, not a leak you can see.

Chemical compatibility is where I see the most avoidable mistakes. Silicone swells in nonpolar solvents (hexane, toluene, dichloromethane) and in some ketones. A swollen septum can push out of the cap or deform enough to break the seal. PTFE is inert to almost everything but is mechanically weaker. If your method uses DCM or THF, go PTFE-faced at minimum, and consider pure PTFE if the method is short.

A practical selection matrix

Here is the decision table I use when specifying septa for a new method. The numbers are the ones I have found reliable in practice; your autosampler and needle gauge will shift the puncture counts by a factor of two either way.

Standard reverse-phase HPLC, aqueous or low-organic mobile phase, single injection per vial: 11 mm, 1.0 mm silicone / 0.06 mm PTFE laminate, non-pre-slit. Temperature rating -40 to 200 °C. Puncture life 3 to 10. Lowest cost per vial.
Same method, multiple injections per vial (calibration curves, stability runs): 11 mm pre-slit silicone/PTFE laminate. Puncture life 30 to 50. Watch for evaporation on volatile analytes.
Aggressive solvents (DCM, THF, DMSO, concentrated acid): 11 mm PTFE-faced laminate, or pure PTFE for short methods. Puncture life 1 to 3 for pure PTFE. Verify reseal on a blank before committing.
Headspace GC at 80 to 150 °C: 20 mm, 3.0 mm PTFE-faced silicone, pre-conditioned at method temperature. Puncture life 1 to 5. Low bleed is the priority.
Headspace GC-MS, trace analysis: 20 mm PTFE-faced low-bleed laminate, baked at 100 °C for 1 hour before use. Verify siloxane background at m/z 73, 147, 207, 281 is below method detection limit.
High-throughput screening, aqueous, 96-well or 384-well plate format: pre-slit silicone mat, 1.0 mm, with a PTFE face if the method uses any organic above 20%.

If you are buying for a whole lab rather than a single method, the sensible default is a 11 mm 1.0 mm silicone/PTFE laminate in both pre-slit and solid versions, plus a box of 20 mm headspace septa if you run GC. That covers 80% of what a typical analytical lab does. The HPLC septa catalog is a reasonable place to compare the laminate options side by side.

Puncture count, coring, and needle gauge

Coring is the failure mode that gets blamed on the column. A cored septum sheds a small plug of silicone or PTFE into the vial, the autosampler draws it up, and it lands on the column frit or in the needle. Symptoms are rising back pressure, split peaks, and eventually a plugged needle. If your back pressure climbs over a run and the column is new, check the septum before you check anything else.

Needle gauge drives coring. A 22-gauge needle (0.72 mm OD) is standard for most HPLC autosamplers and will core a 1 mm silicone septum after a handful of punctures if the needle is not sharp. A 26-gauge needle (0.46 mm OD) cores less but bends more easily. The needle should be sharp, straight, and centered in the septum. If the autosampler’s needle guide is worn, the needle will hit the septum off-center and core it immediately.

Practical puncture limits I work to:

Solid 1.0 mm silicone/PTFE laminate: 5 punctures before I expect reseal problems, 10 before I expect coring.
Pre-slit 1.0 mm silicone/PTFE laminate: 30 to 50 punctures before reseal degrades.
Pure PTFE 1.0 mm: 1 to 3 punctures before reseal fails.
3.0 mm headspace PTFE-faced laminate: 3 to 5 punctures at temperature.

These are conservative. You can push them, but if you are running a validated method, do not. Write the puncture limit into your SOP and change the vial when you hit it.

Crimp and torque: the numbers that matter

For crimp caps, the crimper setting is the whole game. An 11 mm aluminum crimp cap on a 11 mm vial should be crimped to a crimp height of about 5.5 mm (measured from the vial base to the top of the cap). Too loose and the seal leaks; too tight and you crush the septum and deform the cap. I have seen labs use a crimper set 0.5 mm too tight and wonder why their septa fail after two punctures. The septum is being pre-compressed before the needle ever touches it.

For screw caps, the torque is lower than most people expect. A 9 mm screw cap on a 9 mm short-thread vial should be tightened to about 0.5 to 1.0 N·m. Hand-tight is usually fine, but if you use a torque tool, do not exceed 1.0 N·m or you will deform the septum and the cap liner. Over-torqued screw caps are a common source of septum extrusion and needle drag.

If you are using vial closures from a supplier, ask for the recommended crimp height and torque for their specific cap. The numbers vary by cap design and by whether the cap has a liner.

Contamination, extractables, and the blank run

Every septum is a source of extractables and leachables. Silicone contains low-molecular-weight siloxanes; butyl contains curing agents and antioxidants; even PTFE can shed oligomers at high temperature. For trace analysis, the septum is often the largest single source of background in the method blank.

The fix is a blank run. Before you commit a new lot of septa to a validated method, run a blank vial (mobile phase or water only) through the full sequence and look at the chromatogram. For LC-UV, look for peaks above 0.5 mAU at the retention time of your analyte. For LC-MS, look for background above your method detection limit. For GC-MS, check the siloxane ions at m/z 73, 147, 207, and 281. If the blank is clean, the lot is good. If it is not, bake the septa or reject the lot.

This is where a lot-to-lot qualification pays off. I keep a reference chromatogram for each approved septum lot and compare the first blank of every new lot to it. If the background shifts by more than 20%, the lot goes back to the supplier. The analytical testing side of the lab can run this qualification in an afternoon and it has saved me more than one method investigation.

Autosampler-specific considerations

Different autosamplers handle septa differently, and the differences matter at the buying stage.

Agilent autosamplers (1100, 1200, 1260, 1290 series) use a needle that pushes through the septum into the vial and draws from a fixed depth. The gripper holds the vial by the cap, so a cap that is too tall or a septum that sits proud of the cap can cause a gripper fault. Agilent’s own chromatography consumables line specifies the cap and septum dimensions for each autosampler, and it is worth matching to their spec rather than guessing.

Thermo autosamplers (Dionex AS, Vanquish, TriPlus) have similar constraints but different gripper geometry. Thermo’s chromatography and sample handling documentation lists the compatible vial and cap formats for each system. If you are running a mixed fleet (Agilent and Thermo in the same lab), standardize on a vial and cap format that both systems accept, which usually means a 2 mL 11 mm crimp or a 9 mm screw vial with a compatible cap.

Waters autosamplers (2695, ACQUITY) use a 12 x 32 mm vial with a 9 mm screw cap in most configurations. The septum has to be centered or the needle will hit the cap. Waters is strict about this and their method validation will flag off-center septa as a system suitability failure.

Troubleshooting checklist

When a method goes wrong and the septum is a suspect, work through this in order. It takes about 20 minutes and saves hours of column swapping.

Ghost peaks in the blank: septum off-gassing. Bake the septa at 100 °C for 1 hour, or switch to a PTFE-faced low-bleed laminate. Check siloxane ions at m/z 73, 147, 207, 281 on GC-MS.
Rising back pressure: coring. Inspect the vial for particles, check the needle for a plug, and replace the septum lot. Check the column frit before replacing the column.
Split or fronting peaks: needle drag or partial injection. Check that the septum is seated flat and the needle is centered. A deformed septum can drag on the needle and change the injection volume.
Low response or missing peaks: leak through the septum. Check the crimp height (5.5 mm for 11 mm caps) or the screw torque (0.5 to 1.0 N·m). A loose crimp on a heated vial will vent.
Gripper faults or vial pick-up errors: septum sitting proud of the cap, or wrong cap height for the autosampler. Measure the cap and compare to the autosampler spec.
Sample evaporation over a long queue: pre-slit septum or a loose seal. Switch to a solid septum for volatile analytes, or shorten the queue.
Lot-to-lot variability in the blank: supplier changed the silicone formulation. Requalify the lot with a blank run and compare to the reference chromatogram.

If the checklist does not resolve it, the septum is probably not the cause. Move on to the needle, the injection port, and the column. The Restek technical resources have good troubleshooting flowcharts for the injection side of the system, and the EPA SW-846 methods give you the acceptance criteria for blanks and calibration that you can use as a sanity check on your method.

Buying: what to put in the specification

When you write a purchase specification for septa, be specific. A generic “11 mm silicone septum” gets you whatever the distributor has in stock, which may be a 1.0 mm laminate or a 1.5 mm solid silicone with no PTFE face. The specification should include:

Diameter: 11 mm, 13 mm, 20 mm, or 9 mm as appropriate.
Thickness: 1.0 mm for HPLC, 3.0 mm for headspace.
Construction: silicone/PTFE laminate with PTFE on the sample side, or pure PTFE, or pre-slit.
Temperature rating: state the maximum method temperature and require the septum to be rated 20 °C above it.
Puncture life: state the minimum number of punctures the septum must reseal for. 5 for solid, 30 for pre-slit.
Extractable specification: a maximum blank absorbance or a maximum siloxane background at the method detection limit.
Lot qualification: a reference blank chromatogram supplied with each lot, or a first-article blank run on receipt.
Packaging: individually bagged or in a clean-room bag, with a lot number on the bag. Loose septa in a cardboard box pick up dust and fibers.

The cost difference between a generic septum and a qualified one is usually 2 to 4 cents per septum. On a 100-vial run that is four dollars. On a failed run it is a week of instrument time. Buy the qualified one.

Takeaway

Choose septa by matching four parameters to your method: material (silicone/PTFE laminate for most HPLC, PTFE-faced for aggressive solvents and headspace, pure PTFE only for short methods), dimensions (11 mm x 1.0 mm for standard 2 mL vials, 20 mm x 3.0 mm for headspace), puncture life (5 for solid, 30 to 50 for pre-slit), and blank performance (verified with a blank run before the lot goes into a validated method). Crimp to 5.5 mm or torque to 0.5 to 1.0 N·m, change vials at the puncture limit, and keep a reference chromatogram for each approved lot. The septum is the cheapest component in the system and the one most likely to cost you a week. Specify it like it matters.

References & further reading

Related resources from our lab consumables network:

External references and standards: