I have pulled more torn septa out of an autosampler needle guide than I care to count, and in almost every case the analyst swore the needle was new and the septa were the expensive ones. Coring is one of those failures that gets blamed on the wrong component for months. The vial is usually the accomplice, not the victim. Here is what I have learned about how vial geometry, septa selection, and cap torque combine to destroy needles and shed rubber into your sample.
Coring is the mechanical removal of a small plug of septum material by the autosampler needle. The plug is typically 0.5 to 1.2 mm in diameter, roughly the outer diameter of a 23-gauge needle. Sometimes it drops into the vial. Sometimes it stays lodged in the needle bore. Sometimes it accumulates on the needle exterior and gets dragged into the next injection.
The visible signs are easy to miss on a first pass. You see a small circular crater on the top of the septum under a loupe, or a tiny disc of silicone floating in the sample. On the instrument side you see a rising injection-to-injection RSD, split peaks, or a needle that fails a pressure test during the autosampler self-check. On an Aijiren autosampler with a 900 µL loop, a partially occluded needle shows up as a 5 to 15% drop in injection volume before the pressure error trips.
I have seen a single cored septum cause a 0.1 to 0.3 minute shift in retention time across a 40-injection sequence because the plug partially blocked the needle and the injection volume drifted. The analyst spent two days re-equilibrating the column before anyone looked at the septum.
Where the plug goes and why it matters
- Into the vial: the plug sits at the bottom and slowly leaches plasticizers or silicone oil into the sample. On a 1.5 mL vial with 500 µL of sample, a 1 mm silicone plug can contribute measurable siloxane background on GC-MS.
- Into the needle: partial occlusion causes low injection volume and rising RSD. Full occlusion triggers a pressure fault or a failed injection.
- Onto the needle exterior: a ring of rubber builds up and drags sample carryover between vials. I have measured carryover of 0.4 to 1.2% from a single cored septum on a 10 µL injection.
- Into the seat or guide: the plug wedges in the needle guide bushing and bends the needle on the next stroke. This is where needle damage becomes expensive.
The mechanical sequence that damages the needle
Needle damage from a vial is almost never a single event. It is a sequence, and it usually starts with a septum that was not designed for the needle gauge or the number of punctures the method demands.
Step 1: the septum is over-punctured
Every septum has a finite puncture budget. A standard silicone/PTFE septum in a 11 mm crimp cap tolerates roughly 3 to 5 punctures before the puncture channel starts to tear. A pre-slit septum tolerates more because the slit opens rather than cuts, but the slit also lets volatiles escape and is not appropriate for headspace or long sequences with volatile analytes.
If you are injecting 100 samples with the same vial, you are asking one septum to do 100 punctures. It will core by puncture 6 or 7. The plug that comes off on puncture 8 is what bends the needle on puncture 9.
Step 2: the needle hits the septum off-center
Vial geometry matters here. A 12 x 32 mm vial in a 2 mL crimp format has a defined neck ID. If the vial is seated at an angle in the tray, or if the cap is not fully crimped, the needle enters the septum at an angle and cuts rather than pierces. A 2 to 3 degree tilt is enough to start a tear.
I have seen this most often with snap-cap vials where the cap was pressed on by hand instead of with a crimper. The cap sits 0.5 mm proud on one side, the needle enters off-axis, and the septum cores within 10 injections.
Step 3: the plug lodges in the needle guide
The needle guide is a small bushing, typically 0.5 to 0.7 mm ID for a 23-gauge needle. A cored plug of 0.8 to 1.0 mm will not pass. It wedges. The next injection stroke drives the needle into the wedged plug, and the needle bends or the guide bushing cracks.
A bent needle is a $40 to $120 replacement on most autosamplers. A cracked guide bushing is a $200 to $400 service part and usually needs a factory-trained tech. The septum that caused it cost 8 cents.
Septa selection: the numbers that actually matter
Most coring problems trace back to a septum chosen for price rather than for the puncture count and needle gauge the method demands. Here is what I look at when specifying septa for a method.
Needle gauge versus septum thickness
A 23-gauge needle has an OD of 0.64 mm. A 26-gauge needle is 0.46 mm. The thinner the needle, the more the septum resists penetration, and the more likely the needle is to deflect and cut rather than pierce. For 26-gauge needles, I specify 1.0 mm thick septa, not 1.5 mm. Thicker septa increase the penetration force and the coring risk with thin needles.
For a 23-gauge needle, a 1.3 to 1.5 mm thick silicone/PTFE septum is fine. The PTFE face on the sample side is essential: it is chemically inert and prevents the silicone from contacting the sample. Without the PTFE face, you get siloxane background on GC-MS and extractables on LC-MS.
Puncture count and durometer
Silicone durometer is measured on the Shore A scale. A 45 Shore A septum is soft and seals well but cores easily. A 55 to 60 Shore A septum is firmer, resists coring better, but requires more penetration force and may not seal as well on a partially crimped cap.
For methods with more than 5 punctures per vial, I specify pre-slit septa or a higher-durometer septum. Pre-slit is the safer choice for LC-UV methods where volatiles are not a concern. For headspace GC, pre-slit is a leak path and I do not use it.
Temperature rating
Silicone/PTFE septa are rated to about 200 °C continuous. Above that, the silicone degrades and sheds particles. If your GC inlet is at 250 to 280 °C, the septum in the vial is not the problem (the inlet septum is), but a hot autosampler tray or a heated sample compartment can still degrade the vial septum.
For high-temperature methods, specify a PTFE-faced silicone with a rated maximum of at least 220 °C and confirm the tray temperature. I have seen coring accelerate on a 60 °C tray because the silicone softened and the needle cut rather than pierced.
A comparison table for septum selection
This is the decision table I use when specifying septa for a new method. The numbers are from my own bench testing and from manufacturer datasheets.
- Standard silicone/PTFE, 1.5 mm, 45 Shore A: 3 to 5 punctures, 23-gauge needle, up to 200 °C. Best for single-injection methods and LC-UV.
- Pre-slit silicone/PTFE, 1.0 mm, 50 Shore A: 10 to 15 punctures, 23 to 26-gauge, up to 200 °C. Best for long LC sequences, not for headspace.
- High-durometer silicone/PTFE, 1.3 mm, 60 Shore A: 6 to 8 punctures, 23-gauge, up to 220 °C. Best for GC methods with hot trays.
- PTFE/silicone/PTFE sandwich, 1.0 mm: 4 to 6 punctures, 23-gauge, up to 200 °C. Best for LC-MS where extractables matter.
- Natural rubber/PTFE: avoid for coring-prone methods. Natural rubber tears more readily than silicone and sheds larger plugs.
For a full catalog of septum options, see Hplc Septa at hplcsepta.com.
Cap torque and crimp geometry
Cap torque is the variable most analysts ignore and the one that causes the most coring. A crimp cap that is under-crimped lets the septum move under needle pressure. An over-crimped cap deforms the septum and pre-stresses it, which makes it core faster.
Crimp cap torque targets
For an 11 mm aluminum crimp cap on a 12 x 32 mm vial, the target crimp is 0.8 to 1.2 N·m on a calibrated crimper. Below 0.6 N·m, the septum moves and cores. Above 1.5 N·m, the aluminum skirt deforms and the septum is compressed unevenly.
I check crimper calibration every 500 caps with a torque wrench. A crimper that has drifted 0.3 N·m is enough to double the coring rate on a 100-injection sequence.
Screw cap torque targets
For a 9 mm screw cap on a 2 mL vial, the target is 0.4 to 0.6 N·m. Hand-tightening is unreliable: I have measured 0.2 to 1.1 N·m across five analysts on the same cap. Under-torqued caps let the septum lift, and the needle cuts the edge of the puncture rather than the center.
The GPI 9 mm and GPI 11 mm thread standards define the cap geometry. A cap that meets the GPI spec but is torqued outside the range will still core. The standard tells you the thread fits; it does not tell you the seal is correct.
Needle damage modes and how to tell them apart
Not every bent needle is caused by coring. Here is how I distinguish the failure modes on the bench.
Bent tip, straight shaft
The needle tip is bent but the shaft is straight. This is almost always a septum that was over-punctured and the needle hit the vial bottom or the seat. Check the septum for a crater and the vial for a dimple at the bottom.
Bent shaft, straight tip
The shaft is bent mid-length. This is a needle guide or seat problem, often caused by a wedged septum plug. Pull the guide and inspect for rubber residue.
Burred or hooked tip
The tip is burred or hooked. This is a needle that pierced a hardened septum or a cap that was not fully seated. Check the septum durometer and the cap crimp.
Plugged bore
The bore is plugged but the needle is straight. This is a cored plug lodged in the bore. Flush with solvent and inspect the septum for a missing plug.
Operational precautions that cut coring to near zero
These are the changes I have made in my own lab and in customer labs that dropped coring-related needle damage from a monthly event to a once-a-year event.
Match the septum to the puncture count
If the method injects the same vial more than 5 times, switch to pre-slit or high-durometer septa. Do not ask a standard septum to do 20 punctures. It will core, and the plug will find the needle guide.
Calibrate the crimper and the screw cap torque
Calibrate the crimper every 500 caps and the screw cap torque wrench every 200 caps. Log the values. A drifting crimper is the single most common cause of coring I have found in customer audits.
Inspect the needle guide every 500 injections
Pull the guide, inspect for rubber residue, and clean with a lint-free swab and methanol. A guide with visible residue is a guide that will wedge a plug and bend a needle.
Use the right vial geometry for the autosampler
A 12 x 32 mm vial in a tray designed for 11 mm caps will seat at an angle. The needle enters off-axis and cores the septum. Check the tray spec against the vial and cap dimensions. The Chromatography Vial at chromvials.com catalog lists the tray compatibility for each format.
Do not reuse vials with cored septa
A cored septum has a puncture channel that will not reseal. Reusing the vial means the next injection leaks and the needle cuts a second channel. Discard any vial with a visible crater on the septum.
Troubleshooting checklist
When coring or needle damage appears, work through this list in order. I have used it on Aijiren autosamplers and on Thermo Fisher systems with the same results.
- Inspect the septum: look for a crater, a missing plug, or a tear at the puncture site. A crater means coring; a tear means off-axis entry.
- Check the puncture count: count injections per vial. More than 5 on a standard septum is a coring risk.
- Measure cap torque: crimp caps at 0.8 to 1.2 N·m, screw caps at 0.4 to 0.6 N·m. Log the values.
- Inspect the needle guide: pull the guide and look for rubber residue. Clean or replace.
- Check the vial seating: confirm the vial sits flat in the tray and the cap is fully seated.
- Check the tray temperature: above 60 °C, silicone softens and coring accelerates.
- Run a blank injection: a cored plug in the needle shows up as a ghost peak or a pressure fault on the blank.
- Review the RSD trend: a rising RSD over a sequence is the earliest sign of partial needle occlusion.
What the standards say and where they stop
The GPI thread standards define cap and vial geometry, and ISO 4799 covers the dimensional requirements for laboratory glassware including vials. USP <1058> covers the analytical performance of autosamplers but does not specify septum puncture counts. EPA SW-846 methods specify vial and septum requirements for environmental samples, and the U.S. EPA SW-846 test methods are worth reading for the sample handling sections alone.
What the standards do not tell you is the puncture count for your specific needle gauge and method. That number comes from bench testing. I test every new septum lot against the method needle gauge and log the puncture count at which coring starts. A lot that cores at puncture 3 instead of 5 is a lot I reject.
For method development resources and application notes, the Restek technical resources library has good data on septum selection and injection technique.
Recommendation
For a routine LC-UV method with a 23-gauge needle and up to 5 punctures per vial, use a 1.5 mm silicone/PTFE septum, 45 to 50 Shore A, crimped at 0.8 to 1.2 N·m. Inspect the needle guide every 500 injections and calibrate the crimper every 500 caps.
For a long sequence with more than 5 punctures per vial, switch to a pre-slit septum, 1.0 mm, 50 Shore A, and confirm the tray is below 60 °C. For headspace GC, do not use pre-slit septa; use a high-durometer silicone/PTFE, 1.3 mm, 60 Shore A and keep the tray at or below 40 °C.
For LC-MS methods where extractables matter, use a PTFE/silicone/PTFE sandwich, 1.0 mm, and verify the septum lot against a blank injection before running the sequence.
For a full selection of vials and caps that match these septum specs, see the Hplc Vial at hplcvialfactory.com catalog and the broader Chromatography Products at chromatographyforsale.com range.
Coring is a mechanical problem with a mechanical fix. Match the septum to the needle gauge and the puncture count, control the cap torque, and inspect the needle guide on a schedule. Do those three things and the needle damage stops. The vial is not the enemy. The mismatch between the vial and the method is.
References & further reading
Related resources from our lab consumables network:
- Chromatography Vial at chromvials.com
- Hplc Vial at hplcvialfactory.com
- Chromatography Products at chromatographyforsale.com
- Hplc Septa at hplcsepta.com
External references and standards:
