The run stopped at injection 47 of a 96-well plate with a G2613A autosampler pickup error on an Agilent 7693A. The vial was intact. The cap was on. The septum was new. What had actually happened was that the tray had been loaded with a partially seated cap that sat 1.8 mm proud of its nominal height, and the gripper closed on air above the flange. That is the shape of most “chromatography supplier autosampler pickup error and misalignment” tickets I see: not a failed part, a failed fit. The supplier gets blamed because the supplier sold the vial, but the fault is a stack-up between tray datum, gripper tolerance, and cap seating that nobody measured.
This article is about that stack-up. It assumes you already know what a vial is and what an autosampler does. It goes into the three dimensions that actually govern pickup, the error codes each misalignment throws on the common platforms, and the measurements that separate a supplier defect from a fit defect.
When an autosampler fails to grab a vial, the instinct is to blame the vial. Almost always the vial is within tolerance. The failure is the sum of four tolerances that were never designed to be summed:
- Tray Z-height: the vertical position of the tray floor relative to the gripper’s closed position. Nominal is set by the instrument; real trays drift with wear, bent locating pins, and swapped tray inserts.
- Vial height: body height plus cap height. For a 2 mL 12 x 32 mm vial to GPI 8-425 or 13-425, body height runs 31.5 to 32.0 mm and the seated cap adds roughly 5.0 to 6.5 mm depending on closure style.
- Cap seating: whether the cap is fully down on the thread, or cross-threaded, or a crimp that never fully collapsed.
- Gripper tolerance: the open and closed jaw positions, and the sensor that decides “I have the vial.”
Each of those is fine on its own. Add them and you get a pickup error. The job is to find which term moved. The chromatography supplier can only control vial height and cap seating; the other two belong to the instrument and the operator.
The three dimensions that decide whether a vial gets picked up
Tray height and the datum you are actually referencing
The gripper does not know where the tray is. It moves to a stored coordinate and closes its jaws. If the tray floor has risen or fallen by more than half the gripper’s vertical capture window, the jaws close above or below the vial neck. On most modern autosamplers that capture window is 1.5 to 2.5 mm. That is the whole budget.
Two things move the tray datum. First, tray inserts and adapters: a 10-position tray insert that is 0.5 mm thicker than the one the instrument was calibrated with shifts every well by 0.5 mm. Second, worn locating pins. I have seen a tray that had been swapped between two instruments for three years and had 0.9 mm of lateral play at the front corner because the pin hole had ovalised. Every vial in the front row picked up intermittently; the back rows were fine.
Measure it directly. Put a dial indicator on the tray floor at the four corners and the centre, zero on the corner the instrument uses as its reference, and read the spread. If the spread exceeds 0.5 mm, the tray is out of flat or the pins are worn. Re-seat or replace before touching the vials.
Gripper tolerance: open, closed, and the sensor in between
The gripper has a jaw opening and a jaw closing position, and a sensor (optical or force) that confirms capture. The jaw opening must clear the cap diameter with margin. For a 13-425 cap the outside diameter is about 15.5 to 16.0 mm; for an 8-425 cap it is about 10.5 to 11.0 mm. If the jaws open to 15.0 mm and the cap is 15.8 mm, the gripper will push the vial rather than straddle it, and the sensor may still report “captured” because it felt resistance.
The sensor is the part that lies. A force-based sensor that trips at, say, 2 N will register a cross-threaded cap the same way it registers a properly seated one. That is why the error often appears as a pickup error rather than a jam: the gripper thinks it has the vial, lifts, and the vial drops back into the well.
Check gripper tolerance with a go/no-go gauge, not by eye. A machined pin at nominal cap OD plus 0.3 mm should pass between the open jaws; a pin at nominal plus 0.1 mm should not. If the open jaws cannot clear a 16.3 mm pin for a 13-425 cap, the gripper needs service.
Cap seating: the term the supplier actually owns
This is where most real defects live, and it is the one the vial closures supplier can measure. A cap that is not fully seated adds height at the top of the stack, exactly where the gripper closes. Two failure modes:
- Cross-threading: the cap goes on at an angle, sits 1 to 2 mm high on one side, and the gripper grabs the high side. The vial tilts and drops.
- Under-torque on screw caps: a cap that is finger-tight but not fully down sits 0.5 to 1.0 mm proud. Enough to miss the capture window on a worn tray.
For screw caps, the seating target is a defined torque, not “tight.” On a 13-425 screw cap with a PTFE/silicone septum, I set the capping head to 0.9 to 1.2 N·m and verify with a torque tester on the first five vials of every lot. Below 0.8 N·m you get proud caps; above 1.5 N·m you start to extrude the septum and lose the seal. For crimp caps, seating is a crimp dimension, not a torque: the crimp diameter should close to 0.1 to 0.2 mm under the cap skirt OD, and a crimp that leaves the cap rocking on the neck is a pickup error waiting to happen.
Run a seating gauge on 30 vials from each lot. Cap height above the body should be within ±0.15 mm of nominal. If the spread is wider, the capping head is the problem, not the caps.
Which error codes each misalignment throws on common instruments
The error code tells you which term in the stack-up moved, if you know how to read it. The same physical misalignment produces different codes on different platforms because the sensors are different. Here is what I have mapped from field logs:
| Misalignment | Physical signature | Agilent 7693A / G2613A | Thermo TriPlus / AS3000 | Waters 2777C |
|---|---|---|---|---|
| Tray Z too high | Jaws close on cap shoulder, vial pushed down | “Vial not found” / pickup error, often row-specific | “Gripper error” at the same well each time | “Vial present” fault, plate rejected |
| Tray Z too low | Jaws close above cap, air grab | Pickup error after lift, vial left in well | “No vial detected” on lift | Pickup failure, no plate advance |
| Cap proud by >1.5 mm | Gripper grabs high side, vial tilts | Intermittent pickup error, same well | Gripper error, vial dropped | Vial present then absent |
| Cross-threaded cap | Cap angled, sensor sees resistance | Pickup error, then needle bend on next attempt | Gripper error plus needle fault | Pickup fault, cap left in gripper |
| Gripper open too narrow | Vial pushed sideways, not lifted | Pickup error, vial displaced in tray | “Vial position error” | Plate misalignment fault |
The pattern to watch is row-specific or well-specific repetition. A single well that fails every time is a tray or gripper problem at that position. A whole row that fails is a tray datum problem. Random wells across the plate are a cap seating problem, because seating varies vial to vial. That single distinction saves hours.
On Agilent platforms, the 7693A reports pickup errors through the controller and the Agilent chromatography consumables and columns documentation gives the gripper service tolerances. On Thermo platforms the TriPlus and AS3000 use different sensor logic, and the Thermo Fisher chromatography and sample handling service notes list the gripper calibration procedure. Read the code against the table before you touch a vial.
Field case: the front row that failed every third injection
Symptom. A 60-position tray on an Agilent 7693A failed pickup every third injection, always in the front row, always the same three wells. The operator had replaced the vials twice and the caps once.
Measurement. I put a dial indicator on the tray floor at the front-left corner and the rear-left corner with the tray in its locating pins. Front-left read 0.8 mm higher than rear-left. The front locating pin hole was ovalised; the tray was rocking on the pin. The gripper was closing 0.8 mm high on the front row, right at the edge of its capture window, and the third injection happened to be the one where seating variance pushed it over.
Fix. New tray with fresh pins, re-datum the instrument, and a 30-vial seating check on the current cap lot. The seating check showed a 0.25 mm spread, within tolerance, which confirmed the caps were never the problem. Pickup errors stopped. The lesson: the front-row pattern pointed at the tray, and the dial indicator proved it in ten minutes.
Field case: the lot that drifted 0.4 mm mid-run
Symptom. A contract lab running EPA SW-846 methods on a Thermo AS3000 started getting gripper errors on one well per plate, moving around the plate. No row pattern. The vials were from a single lot of OEM vial stock.
Measurement. I pulled 50 vials from the lot and measured cap height above the body with a height gauge. The first 30 were within ±0.10 mm. The last 20 ran 0.3 to 0.4 mm proud. The capping head on the supplier’s line had drifted over the lot. The gripper was fine; the caps were sitting high enough on the last third of the lot to miss the capture window on random wells.
Fix. The supplier re-capped the lot and provided a seating-height certificate. We added a 30-vial seating gauge check to incoming inspection for that closure style, with an accept limit of ±0.15 mm. This is exactly the kind of defect an analytical testing lab should catch at goods-in, not at injection 47.
A fit-and-alignment checklist to run before you call the supplier
Work through this in order. Each step isolates one term in the stack-up. Stop when the measurement fails.
- Tray flatness: dial indicator at four corners and centre, zero on the reference corner. Accept if spread is ≤0.5 mm. Fail means re-seat or replace the tray.
- Tray lateral play: try to move the tray by hand at each locating pin. Accept if play is ≤0.2 mm. Fail means worn pins.
- Gripper open clearance: go/no-go pin at nominal cap OD plus 0.3 mm. Accept if it passes. Fail means gripper service.
- Gripper closed position: confirm the closed jaws grip the cap at mid-height, not the shoulder. Use a pin at nominal cap OD minus 0.2 mm.
- Cap seating height: 30 vials from the lot, height gauge, accept ±0.15 mm of nominal.
- Screw cap torque: torque tester on the first five vials, accept 0.9 to 1.2 N·m for 13-425.
- Crimp dimension: crimp diameter 0.1 to 0.2 mm under skirt OD, no rocking.
- Error code mapping: match the code to the table above and confirm the pattern (well-specific, row-specific, or random).
If steps 1 through 4 pass and step 5 or 6 fails, the supplier owns it. If 1 through 4 fail, the instrument or operator owns it. That is the whole decision, and it takes under an hour with the right gauges. Restek’s technical resources have vendor-neutral dimensional references if you need nominal values for a closure style you do not stock.
What to specify so the fit problem does not come back
The fix is not a better vial. It is a tighter specification on the two dimensions that feed the stack-up, plus a seating check at goods-in. Three things I put on every purchase spec now:
- Cap seating height tolerance of ±0.15 mm, measured on 30 vials per lot, with the certificate reporting the actual spread, not just pass/fail.
- Cap OD and body OD stated to two decimal places so the gripper clearance can be checked against the instrument’s capture window before the lot is accepted.
- A torque or crimp-dimension target for the closure, with the capping head verified on the first five units of each shift.
None of that is exotic. It is the difference between a supplier who ships vials and a supplier who ships a fit. The chromatography supplier conversation should be about seating height and cap OD, not about “pickup errors,” because the supplier cannot see your tray and you cannot see their capping head unless you ask for the number.
The decision rule is simple. If the pickup error is well-specific or row-specific, measure the tray and gripper first and leave the vials alone. If it is random across the plate, measure cap seating height on 30 vials from the lot and check the spread against ±0.15 mm. Only when the seating spread is out of tolerance does the supplier own the failure. Everything else is a datum problem, and a dial indicator finds it faster than a replacement lot ever will.
References & further reading
Related resources from our lab consumables network:
- Chromatography Supplier at chromatographysupplier.com
- Oem Vial at oemsamplevials.com
- Vial Closures at vialclosures.com
- Analytical Testing at labanalyticaltesting.com
External references and standards: