Coring is not a septum defect, it is a geometry and force mismatch

I watched a technician destroy eleven septa in a single afternoon, and not one of them failed because of the septum. She was running a 96-position autosampler on 2 mL vials from a lot that had arrived with crimp caps seated at inconsistent heights. The needle was bottoming out against the glass, the septum was taking the strike, and the coring that followed filled the inlet liner with rubber shavings. That was the day I stopped treating “coring” as a septum problem and started treating it as a vial-and-needle geometry problem.

This article is about coring and needle damage specifically: what causes it, how to detect it before it ruins a run, and what to change in your 2 mL vial selection and handling to make it stop. If you have ever pulled a needle out of a septum and found a neat little plug of silicone stuck to the bevel, you are the audience.

The word “coring” gets thrown around loosely. I want to be precise about it because the fix depends on which failure you actually have.

Coring is material removal. The needle punches a disc out of the septum and either carries it into the vial (a “plug”) or leaves it hanging in the injection port. A related failure is “tenting,” where the septum stretches downward instead of being cut, then retracts and partially reseals around the needle, which causes bent needles when the autosampler lifts. A third is “needle bending” or “needle burring,” where the shaft deflects against the glass or the cap and the bevel is deformed. They have different causes and different remedies, and a lab that lumps them together will keep replacing septa without fixing anything.

The physical driver in all three cases is the same: the needle has to pass through a fixed stack of materials (cap, septum, sometimes a liner or seal) at a defined speed, and the force required to do that has to stay below the needle’s buckling threshold. Anything that raises the required force, or lowers the threshold, produces damage. On a 2 mL vial, the variables that matter are septum thickness and durometer, cap crimp or screw torque, vial height tolerance, and the needle gauge and bevel type. Get any one of them out of range and the others have to absorb the error.

What the needle actually experiences

A standard 22-gauge autosampler needle has an outer diameter of about 0.72 mm and a wall thickness in the 0.15 mm range. The buckling load for a 40 mm length of that tubing is not large. I have seen 22-gauge needles deflect at forces in the low single-digit newtons when the bevel is dull or when the septum has hardened. That is why a septum that was fine for 50 injections in January starts coring in July: the elastomer stiffens with age and heat, the required penetration force climbs, and the needle starts to skate across the surface before it cuts.

Screw-cap septa are usually silicone-faced with a PTFE film on the sample side. The PTFE is there for chemical inertness, and it also reduces friction during penetration. When the PTFE film is damaged or missing on a cut-rate cap, the silicone grabs the needle and the coring rate goes up immediately. This is one of the reasons I am fussy about cap sourcing: the film is doing mechanical work, not just chemical work.

The five things that actually cause coring in 2 mL vials

After enough of these investigations, the causes sort into a short list. I have ranked them by how often I have found them in real labs, not by how often they appear in vendor literature.

1. Septum hardness and thickness out of spec

Silicone septa for 2 mL vials are typically specified at a Shore A durometer in the 45 to 55 range and a thickness of about 1.0 mm for standard crimp caps, with thicker options (1.3 to 1.5 mm) for high-pressure or long-run applications. A septum at the top of the durometer range, or one that has aged, requires noticeably more penetration force. If your supplier does not publish durometer and thickness, that is a red flag. Ask for the certificate of analysis and check the lot-to-lot variation, not just the nominal value.

Thin septa core less because the needle travels less material, but they also reseal worse and are more prone to tenting. Thick septa reseal better but core more. There is no free lunch here; you pick based on injection count and pressure.

2. Cap crimp height and torque

For crimp caps, the crimping tool sets the cap height. A properly crimped 11 mm aluminum cap on a 2 mL vial should seat with the septum compressed evenly and the cap skirt rolled under the vial bead. If the crimp is too high, the septum sits proud and the needle has to push it down before it cuts, which is a tenting setup. If the crimp is too low, the septum is over-compressed and hardened, which raises penetration force. Both produce coring.

For screw caps, torque matters. The common 9 mm screw thread on a 2 mL vial wants roughly 0.6 to 0.9 N·m of tightening torque, which is more than most people apply by hand and less than a torque wrench set to “farmer tight.” Over-torquing compresses the septum against the cap and stiffens it. Under-torquing lets the septum move during penetration. I keep a small torque screwdriver on the bench for cap validation and I have never regretted it.

The relevant dimensional standard here is the GPI 9 mm thread finish (GPI 400-24 for the 9 mm screw thread), and for crimp the 11 mm aluminum seal. If your caps are not made to the GPI finish, the geometry will drift and no amount of technique will save you.

3. Vial height tolerance and seat position

This is the one that bit the technician in my opening story. A 2 mL vial is nominally 32 mm tall with an outer diameter of 11.6 mm and an inner diameter around 8 mm. The tolerance on that 32 mm height is where trouble lives. A vial that is 0.4 mm tall pushes the septum 0.4 mm closer to the needle’s bottom-of-travel. If the autosampler’s needle depth is set for nominal, the needle now strikes the septum at a shallower angle relative to the cap, or bottoms out against the glass neck. Either way you get coring or a bent needle.

I have measured lots where the height spread was over 0.5 mm. That is enough to change the penetration geometry across a tray. Good suppliers hold height to ±0.2 mm or better. Ask for the dimensional report. If you are buying through a 2 mL vial supplier who cannot produce one, you are buying lottery tickets.

4. Needle gauge, bevel, and condition

A 22-gauge needle with a standard bevel is the workhorse. Going to 26 gauge reduces the penetration force per unit area but the needle is more fragile and deflects more easily, so the coring rate can actually go up if the septum is hard. Going to 20 gauge increases force but the needle is stiffer. The right gauge depends on your septum, not the other way around.

Needle condition is underappreciated. A bevel that has been used 500 times is not the same bevel that came out of the box. It dulls, it burrs, and a burred bevel tears rather than cuts. I recommend logging injection counts per needle and replacing on a schedule, not on failure. For most labs running aqueous or low-salt samples, 500 to 1000 injections per needle is a reasonable replacement interval. For high-salt or high-protein samples, cut that in half.

5. Sample matrix and temperature

Cold samples stiffen septa. A tray that sits in a 4 °C autosampler compartment for six hours will core more than the same tray run at ambient, because the elastomer is harder and the required force is higher. This is a real effect and it is easy to test: run the same batch at ambient and at 4 °C and count the plugs in the inlet liner. I have seen a threefold difference.

High-salt matrices leave crystalline residue on the septum and the needle, which acts like an abrasive. If you are running EPA SW-846 methods with high ionic strength extracts, expect more coring and plan for it.

A diagnostic checklist you can run in an hour

When a lab calls me about coring, I do not start with a root-cause meeting. I start with a bench test. Here is the sequence, in order, with the numbers I use.

  • Inspect the septum after 10 injections. Pull the cap and look at the puncture site under a loupe. A clean cut with a small dimple is normal. A ragged hole, a lifted flap, or a visible plug missing is coring. Photograph it.
  • Check the inlet liner. Pull the liner and look for silicone shavings or a plug. A single plug in a liner is a warning. Multiple plugs mean the problem is systemic, not a one-off bad cap.
  • Measure vial height on 20 random vials. Calipers, flat on a granite block. Record min, max, and spread. If the spread exceeds 0.3 mm, the lot is suspect.
  • Measure crimp height or check screw torque. For crimp, the cap skirt should be rolled under the bead with no gap. For screw, use a torque screwdriver and confirm 0.6 to 0.9 N·m.
  • Inspect the needle bevel under magnification. A dull or burred bevel is visible. Replace if in doubt.
  • Run a controlled comparison. New needle, new septum lot, ambient temperature, 20 injections. Then repeat with the suspect lot. Count plugs. This isolates the variable in about 40 minutes.

If the comparison shows the suspect lot is the problem, stop. Do not try to fix it with technique. The dimensional or material spec is wrong and the supplier needs to know.

Decision checklist: which septum and cap for your application

This is the table I actually use when specifying a 2 mL vial setup. The numbers are the ones I have validated across multiple instruments.

  • Routine aqueous HPLC, ambient, <100 injections per vial: 1.0 mm silicone/PTFE septum, 45 to 50 Shore A, 11 mm crimp or 9 mm screw. Replace needle every 1000 injections.
  • Long-run sequences, >100 injections per vial, or high pressure (>400 bar): 1.3 mm silicone/PTFE septum, 50 to 55 Shore A, screw cap torqued to 0.7 N·m. Replace needle every 500 injections.
  • Cold autosampler (4 °C): 1.0 mm septum, 45 Shore A, and expect to replace needles at 500 injections. Consider a pre-pierced septum if the method allows.
  • High-salt or high-protein matrix: 1.0 mm septum with PTFE film on both faces, 50 Shore A, and inspect the liner every 50 injections. Replace needle at 300 to 500 injections.
  • Headspace or GC with a thick septum: 1.5 mm silicone, 55 Shore A, and a needle guide. Accept that coring is more likely and plan the liner maintenance.

Note the pattern: as pressure, injection count, or matrix difficulty goes up, septum thickness and durometer go up, and needle replacement interval goes down. That is the trade you are making. If a supplier tells you one septum handles all of it, they are selling you a compromise you did not ask for.

Why the vial, not the septum, is often the root cause

I want to push back on the reflex to blame the septum. In my experience, roughly half of coring complaints trace to the vial or cap geometry, not the elastomer. The septum is just the part that shows the damage.

Consider what a 2 mL vial has to do. It has to present a flat, consistent sealing surface at a defined height, hold a crimp or screw cap without distortion, and survive autosampler handling without chipping at the neck. The neck finish is where this lives. A vial with a poorly formed bead or a neck that is out of round will not seat the cap evenly, and an unevenly seated cap compresses the septum asymmetrically. The needle then hits a tilted septum, which is the classic tenting setup.

This is why I care about the dimensional envelope of the vial itself, not just the cap. The 11.6 mm outer diameter and 32 mm height are nominal; the tolerances around them are what separate a good lot from a bad one. If you are sourcing through a OEM vial program, write the height and neck-finish tolerances into the specification and require a dimensional report with each lot. It is the single most effective thing you can do to reduce coring.

The autosampler is not innocent either

Needle depth is a programmable parameter and it is frequently set wrong. If the needle is set to travel to the bottom of the vial, it will strike glass and deflect, and the next injection will core because the bevel is now damaged. Set needle depth to leave 2 to 3 mm of clearance above the bottom, and verify it with a vial containing a colored solution so you can see where the tip lands. I have found instruments set 4 mm too deep on more than one occasion, and the coring stopped the day it was corrected.

Needle speed matters too. Fast penetration on a hard septum produces more coring than slow penetration. If your instrument allows it, drop the injection speed by 20 to 30 percent and re-test. It costs a few seconds per injection and saves septa.

What good looks like: specifications to write into your next PO

Here is the specification I would put on a purchase order for 2 mL vials and caps intended for autosampler use. It is deliberately specific because vague specs are how you get coring.

  • Vial: 32.0 mm ±0.2 mm height, 11.6 mm ±0.1 mm OD, borosilicate glass, 9 mm screw or 11 mm crimp neck to GPI finish.
  • Septum: silicone/PTFE, 1.0 mm ±0.05 mm thickness, 45 to 55 Shore A, PTFE film on the sample side, lot COA with durometer and thickness.
  • Cap: aluminum, GPI 9 mm or 11 mm as applicable, crimp height within ±0.1 mm of nominal, screw torque validated at 0.6 to 0.9 N·m.
  • Dimensional report per lot: height, OD, neck finish, septum thickness, durometer.
  • Needle replacement interval documented in the SOP: 500 to 1000 injections depending on matrix.

If you are buying HPLC consumables alongside the vials, apply the same discipline to the septa and liners. The HPLC consumables you pair with the vial are part of the same fluid path, and a cheap septum will cost you more in liner replacements than it saves. I have written before about how a chromatography supplier relationship is really a specification relationship; this is exactly the case where that pays off.

Field fixes that actually work

Not every lab can change suppliers overnight. Here are the fixes I have used in the field, ranked by effort and impact.

Immediate (same day)

Replace the needle. Inspect and replace the inlet liner. Re-set needle depth to leave 2 to 3 mm clearance. Reduce injection speed. These four steps resolve a surprising number of coring complaints because they remove the damaged components and the geometry error at the same time.

Short term (this week)

Measure vial height on the current lot and quarantine any lot with a spread over 0.3 mm. Check crimp height or screw torque on a sample of caps. Switch to a septum with a documented durometer and thickness. Log injection counts per needle and start a replacement schedule.

Structural (this quarter)

Write the specification above into your PO. Require lot dimensional reports. Qualify a second supplier so you are not hostage to one lot. Train the bench on cap torque and needle depth. Add a liner inspection to the preventive maintenance schedule.

For method-specific guidance, the instrument vendors publish useful application notes. Agilent’s chromatography consumables and columns documentation covers septum selection and inlet maintenance in detail, and Thermo Fisher’s chromatography and sample handling resources have good material on autosampler needle care. Restek’s technical resources are worth bookmarking for septum and liner selection charts. None of these will fix a bad vial lot, but they will help you specify the right one.

What I would do if I were standing in your lab tomorrow

I would pull the current vial lot, measure 20 vials with calipers, and write the height spread on a whiteboard. I would pull the inlet liner and count the plugs. I would check the needle bevel under a loupe. Then I would set needle depth to leave 3 mm of clearance, replace the needle, and run 20 injections with a fresh septum lot. If the plugs stop, the problem was the lot or the setup. If they continue, the problem is the method or the matrix, and I would move to a thicker, harder septum and a shorter needle interval.

The takeaway is simple: coring on 2 mL vials is almost never a septum defect in isolation. It is a mismatch between the force the needle applies and the force the septum-and-cap stack can absorb without deforming. Fix the geometry first (vial height, crimp or torque, needle depth), then the materials (septum durometer and thickness), then the consumable schedule (needle replacement interval). Do those three in that order and the shavings in your liner will stop appearing.

References & further reading

Related resources from our lab consumables network:

External references and standards: