Run 47 on the autosampler came back with a coring fragment lodged in the injection port liner, a 0.8 µL injection volume that had drifted to an apparent 0.3 µL, and a peak area RSD of 11.4% across six injections that should have sat under 2%. The vial septum had a clean 1.03 mm disc of silicone and PTFE punched out of it, sitting at the bottom of the glass insert. That is aijiren coring, fragments and needle damage diagnosis in its most common form: not a seal failure, not a worn part, but a mechanical punch that removed material from the closure and carried it downstream. Once you know what to look for under a loupe, the diagnosis takes about ten minutes. Getting to the root cause takes longer, because the punch is almost always the result of a needle geometry and guide dimension that were never matched to the septum in the first place.

Coring is not a random event. It is a repeatable sequence that happens every time the needle passes through the closure, and understanding the stages tells you which parameter to change.

Stage 1: the needle bevel contacts the septum surface

A standard HPLC or GC autosampler needle has a point style. Most are either a 22-gauge bevel (0.72 mm OD), a 23-gauge (0.64 mm OD), or a 26-gauge (0.46 mm OD) for small-volume work. The bevel angle matters more than most people realize. A 12° bevel slices through the septum; a 30° or 45° bevel tends to push material ahead of the tip before it cuts. That push is the beginning of a plug. If the septum is cold (below about 15°C) or the silicone is a high-durometer formulation, the material does not yield cleanly and the bevel starts to compress rather than cut.

Stage 2: the needle shaft drags material into the bore

As the needle advances, the outer surface of the shaft carries a thin film of septum material into the hole it just made. On a 0.64 mm OD needle passing through a 0.5 mm pre-slit, the clearance is nominally 0.07 mm per side. That is enough to shear a clean hole. On a 0.72 mm OD needle through the same slit, the clearance drops to 0.04 mm per side, and the material has nowhere to go except into the needle bore or back up the shaft. I have pulled needles where the bore was packed with silicone and PTFE flakes after fewer than 200 injections. That is a geometry problem, not a septum quality problem.

Stage 3: the guide and the needle axis misalign

This is the stage most people miss. The needle guide (the small bore in the autosampler head that keeps the needle vertical) has an internal diameter. If that ID is more than about 0.05 mm larger than the needle OD, the needle can deflect laterally by a few hundredths of a millimeter. That deflection turns a perpendicular punch into an angled punch, and an angled punch removes a plug rather than making a hole. A 0.72 mm needle in a 0.80 mm guide has 0.08 mm of play. Over 500 injections, that play translates into an oval hole, then a torn hole, then a plug. The guide is a wear item, and it is almost never on anyone’s PM schedule.

Stage 4: the plug is ejected into the flow path

Once the plug is free, it either stays on the needle tip (and gets carried into the next vial), drops into the vial insert (and blocks the needle on the next injection), or is pushed into the injection port liner (and shows up as a ghost peak or a split peak three runs later). The stage at which you catch it determines how much of your data you have to throw away.

Reading the debris under a loupe: what the fragment shape tells you

A 10× to 20× loupe and a clean white background will tell you more about the failure than any instrument log. The shape of the fragment is a fingerprint of the mechanism.

  • Perfectly circular disc, 0.8 to 1.1 mm diameter, clean edges: a true punch. The needle was perpendicular, the bevel was sharp, and the septum material was soft enough to be removed as a plug. This is the classic coring signature and it points to a needle that is too large for the septum or a septum that is too thin for the needle.
  • Oval or teardrop fragment: the needle was deflecting. Check guide ID versus needle OD. A teardrop is the single most reliable indicator of guide wear.
  • Crescent or half-moon fragment: the needle entered at an angle or the vial was not seated squarely in the tray. Look at the vial height tolerance (a 32 × 11.6 mm vial should sit within ±0.2 mm of nominal) and the tray spring tension.
  • Stringy, elongated fragment with ragged edges: the septum was cold or the silicone was a high-durometer formulation. Check the septum hardness (Shore A) and the sample tray temperature.
  • Fine flakes or powder rather than a discrete fragment: the needle bevel is dull or burred. A dull bevel abrades rather than cuts. Replace the needle.

I keep a small logbook of fragment photos next to the autosampler. When a new analyst reports a problem, the first question is “what does the debris look like under the loupe?” Nine times out of ten, the shape narrows the cause to one of two things before anyone touches the instrument.

The needle and guide dimensions that cause coring

This is the part that gets skipped in most troubleshooting guides. The numbers matter, and they interact.

Needle outside diameter versus septum pre-slit

The rule I use is simple: the needle OD should not exceed 80% of the pre-slit width. A 0.5 mm pre-slit handles a needle up to 0.40 mm OD cleanly. A 0.72 mm needle in a 0.5 mm slit is a 144% oversize condition, and it will core. If your method needs a 22-gauge needle, you need a septum with a 0.9 to 1.0 mm pre-slit, or a septum formulated for high puncture counts (some silicone/PTFE laminates are rated to 500+ punctures at 22 gauge before fragment shedding becomes measurable).

Needle bevel angle and point style

A 12° bevel is the standard for analytical autosamplers because it cuts cleanly. A 30° bevel is common on prep-scale and some GC needles, and it pushes material. If you are seeing circular discs with a 30° bevel, the fix is either a different needle or a softer septum. Do not try to compensate with injection speed; slower insertion makes the push worse, not better.

Guide internal diameter and wear tolerance

The guide ID should be within 0.02 to 0.03 mm of the needle OD for a new guide. At 0.05 mm of clearance, you are at the edge of acceptable. Beyond 0.08 mm, lateral deflection becomes measurable and teardrop fragments appear. I replace guides at 0.06 mm clearance as a preventive measure, which on a typical autosampler running 200 injections a day works out to roughly every 8 to 12 months. The part costs less than a single repeat run.

Septum thickness and durometer

A 1.0 mm thick silicone/PTFE septum at 45 to 55 Shore A is the sweet spot for most HPLC autosamplers. Thinner than 0.8 mm and the needle punches through without enough material to seal; thicker than 1.5 mm and the needle has to push harder, which increases the coring risk. High-durometer formulations (60+ Shore A) resist coring better but seal worse after multiple punctures, which is a different failure mode and outside the scope here.

Parameter Safe range Coring risk threshold What to check
Needle OD vs pre-slit ≤ 80% of slit width > 100% of slit width Caliper the needle, measure the slit under loupe
Guide ID minus needle OD 0.02–0.03 mm > 0.08 mm Pin gauge the guide, replace at 0.06 mm
Needle bevel angle 12° ≥ 30° Visual inspection or needle spec sheet
Septum thickness 1.0–1.2 mm < 0.8 mm or > 1.5 mm Micrometer on a removed septum
Septum durometer 45–55 Shore A > 60 Shore A Durometer or supplier COA
Vial height tolerance ±0.2 mm of nominal ±0.5 mm Height gauge on 10 vials per lot
Crimp torque (11 mm) 1.2–1.5 N·m < 1.0 or > 1.8 N·m Torque tester on 5 closures per lot

That table is the checklist I hand to a new analyst on day one. It covers the dimensions that actually cause coring, in the order they fail. If you are sourcing closures and need the dimensional baseline, the Vials With Caps reference set is a useful starting point, and the broader Vial Supplier catalog covers the 11 mm and 9 mm crimp and screw formats that most autosamplers use.

Two field cases where the debris told the story

Case 1: the teardrop that was a worn guide

A contract lab running a 1260 Infinity II with a 22-gauge needle started seeing split peaks on early-eluting compounds and a slow drift in retention time of about 0.04 min per run. The analyst assumed column degradation and replaced the column twice. The peak shape did not improve. I asked for the septum fragments from the last 20 vials. Under a 20× loupe, 14 of the 20 fragments were teardrop-shaped, with the point of the teardrop consistently oriented in the same direction. That orientation is the giveaway: a worn guide allows the needle to deflect in one direction, so the fragments all point the same way.

The measurement: the guide ID was 0.78 mm against a 0.72 mm needle, a clearance of 0.06 mm. That is at the replacement threshold. We replaced the guide, re-ran the same method, and the retention drift disappeared within two injections. The split peaks resolved because the fragments had been partially occluding the needle bore, which changed the effective injection volume. The column was never the problem.

Case 2: the circular disc that was a septum mismatch

A pharma QC lab running a GC method with a 0.5 mm pre-slit septum and a 23-gauge needle (0.64 mm OD) reported ghost peaks at 8.2 and 11.6 minutes that appeared in blank injections. The instrument log showed nothing unusual. The injection port liner, when pulled, had a 1.0 mm disc of silicone sitting on the glass wool. Under the loupe, the disc was perfectly circular with clean edges, which is the classic true-punch signature.

The measurement: the needle OD (0.64 mm) was 128% of the pre-slit width (0.5 mm). That is well past the 80% rule. The fix was not a new needle; it was a septum with a 0.9 mm pre-slit rated for 23-gauge needles. We also replaced the liner and the glass wool, because the disc had been sitting there long enough to adsorb and then release compounds. Ghost peaks gone in the next blank. The lab had been about to send the GC out for service, which would have cost three weeks of downtime and found nothing.

Diagnostic sequence: from symptom to root cause in under an hour

When a run fails with suspected coring, I work through this sequence in order. It is designed to eliminate the cheap causes before anyone touches the instrument.

  1. Pull the septum from the failed vial. Look at the hole under a 10× loupe. Is it a clean hole, an oval, a tear, or a missing plug? That single observation eliminates half the possible causes.
  2. Check the vial insert and the bottom of the vial. Is there a fragment? If yes, the plug was ejected on this injection, not a previous one. If no, the plug went downstream or stayed on the needle.
  3. Inspect the needle tip and bore. A loupe and a bright light. Is the bevel sharp? Is the bore clear? A packed bore explains a low injection volume even when the septum looks fine.
  4. Measure the needle OD and the guide ID. Caliper for the needle, pin gauge for the guide. Clearance above 0.06 mm is a replacement candidate; above 0.08 mm is a confirmed cause.
  5. Check the septum specification against the needle. Pre-slit width, thickness, durometer. The 80% rule applies.
  6. Check the crimp torque. An over-crimped closure (above 1.8 N·m on an 11 mm crimp) deforms the septum and makes it more likely to core. An under-crimped closure (below 1.0 N·m) lets the septum move, which also cores. Target 1.2 to 1.5 N·m.
  7. Check the vial height and tray seating. A vial that sits 0.5 mm high changes the insertion depth and the angle. Ten vials per lot, height gauge, ±0.2 mm.
  8. Only then look at the instrument. If all of the above are within spec, the problem is in the autosampler head or the tray, and that is a service call.

Steps 1 through 3 take about fifteen minutes and resolve most cases. Steps 4 through 7 are where the real root cause usually lives. Step 8 is rare.

Consumable selection: the numbers that prevent coring before it starts

The cheapest fix is to buy the right closure and needle combination in the first place. Here is what I specify, and why.

  • Septum pre-slit: at least 1.25× the needle OD. For a 0.72 mm needle, use a 0.9 mm pre-slit minimum. For a 0.64 mm needle, 0.8 mm is acceptable. For a 0.46 mm needle, 0.6 mm.
  • Septum thickness: 1.0 to 1.2 mm. Below 0.8 mm, the needle has too little material to seal against after multiple punctures. Above 1.5 mm, insertion force rises and coring risk goes up.
  • Durometer: 45 to 55 Shore A. Softer seals better but cores more easily; harder resists coring but seals worse. The middle is where most methods land.
  • Puncture rating: ask the supplier for a tested puncture count at your needle gauge. A good silicone/PTFE laminate is rated to 300 to 500 punctures at 22 gauge before fragment shedding becomes measurable. If the supplier cannot give you a number, that is a data point.
  • Guide replacement interval: every 8 to 12 months at 200 injections per day, or whenever clearance exceeds 0.06 mm. Put it on the PM schedule.
  • Crimp torque: 1.2 to 1.5 N·m on 11 mm crimp closures. Verify with a torque tester on five closures per lot.

For the instrument-side consumables (liners, septa, ferrules), the manufacturer technical resources from Agilent, Thermo Fisher, and Restek are worth bookmarking. They publish the dimensional specs that let you check needle-to-septum compatibility before you buy. The U.S. EPA SW-846 methods are also useful as a reference for the injection volumes and needle gauges that regulated methods actually use, which tells you what your closure has to survive.

If you are standardizing across a lab, the Aijiren consumables line covers the 11 mm and 9 mm formats with published septum dimensions, which makes the compatibility check straightforward. For method development work where injection volumes and needle gauges change frequently, the Lab Chromatography reference material on injection technique is a useful companion to the dimensional specs.

What to do when the fragment is already downstream

If the plug has reached the injection port liner, the run is compromised and the liner needs to come out. A fragment sitting on glass wool adsorbs and then releases compounds over dozens of injections, which is exactly how ghost peaks appear three runs after the coring event. Replace the liner and the glass wool, re-run a blank, and confirm the ghost peaks are gone before you trust any data from that sequence. If the fragment reached the column, you are looking at a column flush at minimum and possibly a column replacement; a 0.5 mm silicone disc in a 2.1 mm ID column is a partial occlusion that will show up as increased backpressure and peak broadening, not as a clean failure.

The decision rule is this: any visible fragment in the flow path means the liner comes out and the column gets flushed before the next injection. Do not try to run through it. The cost of a liner and a flush is minutes; the cost of re-running a sequence because of ghost peaks is a day.

And the preventive rule, which is the one that actually matters: match the needle OD to the pre-slit at 80% or less, keep guide clearance under 0.06 mm, and replace the guide on a calendar schedule rather than waiting for the teardrops to appear. Coring is a dimensional problem before it is a quality problem. The debris under the loupe tells you which dimension failed.

References & further reading

Related resources from our lab consumables network:

External references and standards: