The four dimensions that actually decide fit, in the order they fail

Last spring a lab running EPA 8270 on a Thermo TRACE 1310 with a TriPlus RSH autosampler called me because their internal standard response had dropped 22% over three injection sequences, and the run-to-run RSD on the 50 µg/L check standard had crept from 1.8% to 6.4%. The instrument was fine. The column was three weeks old. The problem was that the 8 mm screw-top vials they had switched to (to save money) sat 1.2 mm lower in the tray than the 9 mm wide-mouth vials the method was written around, and the syringe was bottoming out against the glass instead of drawing from 2 mm above the taper. That is the whole game with aijiren size compatibility to a specific autosampler or inlet: the dimensional stack has to match the instrument’s mechanical drawing, not the catalog’s marketing photo. This article walks through each dimension in that stack, how to measure it, and how to verify it against a named instrument before you commit to a lot.

When a vial does not work in an autosampler, it is almost never because the wrong polymer was chosen or because the septa were the wrong durometer (those are separate articles on this site). It is because one of four numbers is off. I check them in this order because this is the order in which they cause problems on the bench.

Outside diameter: the number that determines whether the vial enters the tray at all

For 2 mL autosampler vials, the two OD families you will encounter are 11.6 mm (the standard 12 x 32 mm format) and 12.0 mm (the “wide” variant used by some Shimadzu and older Waters trays). The 0.4 mm difference sounds trivial until you load a tray of 12.0 mm vials into a carousel bored for 11.6 mm and the gripper stalls on the third position. The GPI finish that goes with each is different: 11.6 mm vials almost always carry a 8-425 or 9 mm screw thread, while 12.0 mm vials use a 10-425 or 13-425 neck. Measure OD with a digital caliper at three points around the circumference and take the maximum; molded glass is not perfectly round, and a vial that mics at 11.62 mm on one axis can be 11.78 mm on another. The tray bore is the limiting dimension, not the nominal.

Height: 32 mm is a family, not a spec

The nominal “32 mm” height on a 12 x 32 vial is measured from the base to the top of the crimp or the top of the cap, and manufacturers differ on which. Agilent’s published drawings for their 2 mL vials give 32.0 ± 0.4 mm base-to-crimp-top; some Asian imports run 31.5 mm and some run 32.6 mm. On a fixed-height tray with a spring-loaded needle guide, 0.8 mm of extra height means the needle guide does not seat and the syringe pierces at an angle, which is how you get coring (a plug of septum pushed into the vial) rather than a clean puncture. On a variable-height tray (most modern autosamplers), height tolerance is absorbed, but insertion depth is not. That is the next number.

Neck finish: GPI thread, crimp, or snap, and the cap that goes with it

The neck finish is the thread form and the sealing geometry. For screw vials, the GPI (Glass Packaging Institute) standard designates the thread as 8-425 (8 mm diameter, 425 thousandths of an inch pitch), 9 mm (a short-form thread used on wide-mouth vials), 10-425, and 13-425. A 9 mm cap will not seal on an 8-425 neck, and vice versa, even though both are “screw top.” For crimp vials, the finish is 11 mm (the standard) or 20 mm (headspace). The crimp requires a specific crimper jaw: an 11 mm crimper on an 11 mm neck with a 0.5 mm too-thick seal will not roll the aluminum skirt properly, and you get a leak that shows up as a 15% loss of the lightest analyte over a 24-hour autosampler hold. This is where the cap and seal selection actually lives, and it is why I send people to Vials With Caps when they need the cap matched to the neck rather than guessing from a part number.

Insertion depth: the number nobody publishes and everybody needs

Insertion depth is the distance from the top of the cap (or the crimp) to the point where the needle should stop. It is not the same as vial height. On a vial with a 300 µL insert, the insert sits on a polymer spring or a glass dimple and its own height is 29.5 to 30.0 mm for a typical 12 x 32 configuration. If the autosampler’s needle guide is set for a 32 mm vial and you load a 30 mm insert-in-vial, the needle travels 2 mm too far and either bends (0.4 mm OD needles bend at about 3 N of lateral load) or punches through the insert bottom into the outer vial, which is how you get glass shards in the syringe. The fix is to program the insertion depth in the autosampler method, not to trust the vial’s nominal dimensions.

How to verify each dimension against a named instrument before you buy

The verification workflow is the same regardless of instrument, but the reference numbers come from the instrument’s own consumables drawing. I use the instrument manufacturer’s published accessory list as the ground truth, then measure the candidate vial against it. Here is the procedure I run for any new lot.

1. Pull the instrument’s consumables drawing. For Agilent 7693A and 7697A, the published tray and needle-guide dimensions are in the site preparation guide; for Thermo TriPlus RSH and AS3000, they are in the autosampler user guide. For Shimadzu AOC-20i and AOC-20s, the tray bore and needle stroke are in the AOC manual. These are the numbers you compare against, not the vial catalog.
2. Measure ten vials from the lot with a caliper. Record OD at three axes, height base-to-cap-top, and insert height if applicable. Compute the mean and the range. If the range on any dimension exceeds the instrument’s stated tolerance, reject the lot before it reaches the tray.
3. Run a fit test on the actual tray. Load a full tray of the candidate vials and cycle the autosampler through 50 empty injections with the needle guide in place. Listen for the gripper stalling or the needle guide not seating. A vial that fits 49 of 50 positions is a vial that will fail on the 50th, and it will fail during a batch you care about.
4. Run a puncture test with the real septa. A 9 mm silicone/PTFE septum on a 8-425 neck takes about 50 to 60 punctures before coring becomes visible; a 10 mm septum on a 10-425 neck takes about 80. If the autosampler’s needle is 0.5 mm OD and the septum is 1.0 mm thick, the puncture count drops by roughly 20%. Verify with a 100-injection sequence and inspect the septum under a loupe at 20x.

If you want the instrument-side reference for Agilent, their chromatography consumables and columns pages list the tray and needle-guide part numbers by autosampler model, which is the fastest way to get the drawing. Thermo’s chromatography and sample handling documentation has the same information for the TriPlus line. Restek’s technical resources are useful for the septum and insert side of the stack, and the EPA SW-846 methods are where the insertion-depth and hold-time requirements for the method actually come from.

A dimension-by-dimension comparison table for the four common autosampler families

The table below is what I hand to a lab manager who is about to standardize on one vial across multiple instruments. It is not exhaustive, but it covers the four families that account for most of the mismatches I see. Every value is a nominal with a tolerance; treat the tolerance as the real spec.

Two things to notice. First, the Shimadzu AOC-20i is the odd one out on OD; if you standardize on 11.6 mm vials for everything else, you need a 12.0 mm vial for the Shimadzu tray or a tray insert. Second, the insertion depth is not the same across families even when the vial height is. That 0.5 mm difference between Agilent and Thermo is the difference between a clean draw and a needle that touches the taper.

Field case one: the 9 mm wide-mouth vial that drifted 0.4 mm and cost a week

A contract lab running a 200-vial batch of EPA 8260 on an Agilent 7697A headspace sampler saw the internal standard area count drop 18% between the first and last 50 vials of the tray. The instrument was leak-checked and passed. The transfer line was new. The column had 40 injections on it. I asked them to measure the vial height on the first and last 20 vials of the lot. The first 20 miced at 32.1 mm; the last 20 miced at 32.5 mm. The lot had drifted within the manufacturer’s stated tolerance, but the headspace sampler’s vial heater block is set for a 32.0 mm nominal with a 0.4 mm tolerance, and at 32.5 mm the vial did not seat fully in the heater block. The headspace equilibration temperature was 80 °C, and a 0.4 mm gap between the vial base and the block face dropped the actual sample temperature by about 4 °C, which shifted the partition coefficients enough to move the light analytes out of the headspace. The fix was not a new instrument; it was a lot rejection and a switch to a supplier who would hold height to ±0.2 mm. The lesson: on a headspace sampler, height tolerance is a temperature-control parameter, not a fit parameter.

Field case two: the 300 µL insert that bent a needle on a Thermo AS3000

A pharmaceutical QC lab on a Thermo AS3000 was getting intermittent needle-bend faults, about one every 80 injections, and only on the low-volume methods that used a 300 µL insert. The autosampler method had been copied from a 2 mL vial method without changing the insertion depth. The insert height was 29.8 mm; the method insertion depth was 30.5 mm. The needle was traveling 0.7 mm past the insert bottom, and on a glass insert with a 0.3 mm wall, that is enough to deflect the needle tip and set up a bend on the next injection. We measured ten inserts and found a 0.4 mm height range within the lot, which meant the worst-case insert was 0.4 mm shorter than nominal and the needle was 1.1 mm past the bottom. The fix was to set the insertion depth to 28.5 mm and add a 1 mm pull-back in the method, plus a lot acceptance criterion of 29.8 ± 0.2 mm on insert height. The needle-bend rate went to zero over the next 2,000 injections.

Both of these cases are the same failure mode in different clothing: a dimension that was nominally correct but out of tolerance in the direction the instrument could not absorb. The measurement that identified each one was a caliper reading on ten vials from the lot, not a service call.

What to put in a lot acceptance spec before the first vial hits the tray

If you are writing a purchasing spec for aijiren vials going into a named instrument, the spec has to be written against that instrument’s drawing, not against a generic 12 x 32 catalog line. Here is the checklist I use, with the numbers that have held up across the lots I have qualified.

• OD: 11.6 ± 0.2 mm for Agilent, Thermo, and Waters trays; 12.0 ± 0.2 mm for Shimadzu AOC-20i. Reject any lot with a range greater than 0.3 mm across ten vials.
• Height: 32.0 ± 0.4 mm for most autosamplers; tighten to ±0.2 mm for headspace samplers, because height is a temperature-control variable there.
• Neck finish: match the GPI thread to the cap you are actually using. 8-425 cap on 8-425 neck, 9 mm cap on 9 mm neck, 10-425 on 10-425. Do not mix.
• Insert height: 29.8 ± 0.2 mm for a 300 µL insert in a 12 x 32 vial. Measure ten per lot.
• Septa: 1.0 mm silicone/PTFE for routine; 0.5 mm for high-temperature headspace above 120 °C, accepting the lower puncture count.
• Crimp seal: 11 mm aluminum, 0.5 mm skirt, crimped to a torque that gives a visible roll without cutting the seal. I use a hand crimper set to 0.9 N·m and check the seal with a 10-second pull test at 50 g.
• Puncture count: retire screw vials at 50 punctures, crimp vials at 60, and any vial that has been held at 4 °C for more than 24 hours after the first puncture.

This spec is the one I would put in front of any supplier, and it is the one I would use to audit a lot on receipt. If a supplier cannot hold these numbers, the instrument will tell you in the first batch. The vial supplier side of the conversation is about whether they will certify to the spec and provide a certificate of analysis with the measured range, not just the nominal.

Troubleshooting checklist when a vial does not fit the instrument it was bought for

When a fit problem shows up, work through this list in order. It is ordered by how often each item is the actual cause, based on the calls I have taken.

• Symptom: gripper stalls or vial tips in the tray. Measure OD at three axes on ten vials. If the max exceeds the tray bore by more than 0.1 mm, the lot is out. If OD is in spec, check the tray bore for residue or a bent spring.
• Symptom: needle bends or cores the septum. Measure insertion depth against the instrument drawing. Check insert height if an insert is used. Check septum thickness (a 1.5 mm septum in a 1.0 mm seat will hold the needle 0.5 mm high and cause coring on the next injection).
• Symptom: response drifts over a tray. Measure height on the first and last 20 vials. On a headspace sampler, height drift is a temperature problem. On a liquid autosampler, height drift is a needle-depth problem.
• Symptom: leak at the crimp. Check the crimper jaw against the neck finish. An 11 mm crimper on an 11 mm neck is correct only if the seal is 0.5 mm; a 0.7 mm seal needs a different jaw. Pull-test ten seals at 50 g for 10 seconds.
• Symptom: intermittent no-draw. Check the vial bottom for a glass dimple or a manufacturing artifact. A vial with a 0.5 mm dimple in the center will leave 20 µL undrawable, which matters on a 100 µL method.

If you are running multiple instrument families in one lab, the cleanest solution is to standardize on one OD and one neck finish per family and keep a separate tray insert for the outlier. The lab chromatography side of this is that the vial is part of the sample path, and the sample path has to be dimensionally consistent across the method, not just chemically consistent. The Aijiren catalog is organized by instrument family for exactly this reason; use the family, not the nominal size, as your starting point.

Decision rule: match the instrument’s drawing, then hold the lot to it

The decision rule is simple and it does not depend on brand loyalty. Pull the consumables drawing for the exact autosampler or inlet model, extract the four numbers (OD, height, neck finish, insertion depth), write them into a lot acceptance spec with tolerances no wider than the instrument’s own, and measure ten vials per lot against that spec before the lot goes into the tray. If a supplier will not certify to those numbers with a measured range, the lot is not qualified, regardless of what the catalog says. The instrument will not negotiate, and neither should your purchasing spec.

References & further reading

Related resources from our lab consumables network:

• Aijiren at aijirentech.com
• Vials With Caps at vialswithcaps.com
• Vial Supplier at vialsupplier.com
• Lab Chromatography at labchromatography.com

External references and standards:

• Agilent chromatography consumables and columns
• Thermo Fisher chromatography and sample handling
• U.S. EPA SW-846 test methods for waste and materials
• Restek technical resources for chromatography