A purchasing agent forwarded me a datasheet last month for a 9 mm screw-top vial. It listed “Residual volume: < 5 µL" and "Septum resealability: > 50 injections.” Both numbers looked fine on paper. The first value was never tested on that lot. The second was tested on a different septum, from a different production year, against a 26-gauge conical needle at 23 °C. The agent had already placed a PO. Reading an HPLC vial datasheet is not about finding the best numbers. It is about knowing which numbers were measured, which were copied from a master specification, and which were borrowed from a test that was never run on the product in front of you.
Tested values carry a method, nominal values carry a tolerance that may not be inspected
Every line on an HPLC vial datasheet falls into one of two buckets. A tested value has a defined measurement procedure, a sample size, and a pass/fail limit that the manufacturer actually checks on production lots. A nominal value is a design target. It is what the mold or the tubing was supposed to produce. It may be checked once during tooling qualification and never again, or it may be checked on every tenth lot. The datasheet rarely tells you which bucket you are looking at.
I have seen a spec sheet for a 12 x 32 mm vial that listed “Outer diameter: 11.6 ± 0.2 mm.” That looks like a tested value because it has a tolerance. The manufacturer measured OD on 5 vials per mold cavity during first article inspection in 2019. The mold has since run 4 million cycles. Nobody re-measures OD on production lots because the autosampler tray tolerances are loose enough that a worn mold would need to drift far outside spec before a tray jam occurred. The number is real, but it is not lot-release data.
Ask the supplier one question for every line that matters to your method: “What is the AQL and sample plan for this parameter?” If they cannot answer within a day, the value is nominal. That does not automatically disqualify the vial. It means you need to verify the parameter yourself on incoming lots, or accept the risk. Many labs run 2 mL vials with nominal OD and never see a problem. The risk is not the nominal value itself. The risk is treating it as guaranteed.
Bore diameter and residual volume: the two most misread lines on the sheet
The inner bore of an HPLC vial is specified as a diameter, but it is formed by a mandrel, not machined. The mandrel wears. A 6.0 mm bore on a freshly tooled line may drift to 6.3 mm after 2 million cycles before the manufacturer replaces the mandrel. The datasheet still says 6.0 mm. The actual bore changes the residual volume, which changes the minimum working volume, which changes how much sample you can inject from a 300 µL limited-volume insert.
Residual volume is the volume of liquid that remains in the vial or insert after the autosampler needle can no longer aspirate it. It is not a fixed geometric property. It depends on needle gauge, needle tip geometry, needle depth setting, vial bottom shape, and surface wetting. A datasheet that lists “Residual volume: 2 µL” without stating the needle gauge and depth is giving you a number that may be off by a factor of three in your lab. I have measured residual volume on the same conical insert with a 23-gauge bevel needle and a 26-gauge conical needle. The 23-gauge left 4.1 µL. The 26-gauge left 1.8 µL. Same insert, same lot, same solvent. The datasheet said 2 µL. It was right for one needle and wrong for the other.
When you see a residual volume spec, check for a footnote or a test condition line. If it says “measured with 26-gauge conical needle, 2 mm from bottom,” the number is tested. If it just says “residual volume: < 2 µL," treat it as a nominal design estimate. The only way to know your real residual volume is to weigh a dry vial, fill it with your mobile phase, aspirate with your autosampler settings, and weigh it again. The difference in mass divided by density is your residual volume. That takes 10 minutes and removes the guesswork.
Septum puncture count is a test result, not a product property
No other spec-sheet line causes more field failures than puncture count. A septum rated for 100 punctures was tested under specific conditions: a specific needle gauge, a specific needle tip style, a specific injection speed, a specific temperature, and often a specific solvent. Change any one of those and the real puncture count moves.
The standard test most manufacturers use is a 23-gauge bevel needle at 20 to 25 °C, piercing the septum at the same point repeatedly until the needle drops a weight or the septum leaks under a defined pressure. That test tells you almost nothing about your method if you use a 26-gauge conical needle at 4 °C with acetonitrile. A conical needle tears a different hole profile than a bevel. Cold PTFE/silicone septa stiffen. Acetonitrile swells some silicone formulations.
I ran a comparative test on three septum lots from three suppliers. All three datasheets claimed “> 100 punctures.” Using a 26-gauge conical needle at 6 °C with 70:30 acetonitrile:water, the first lot leaked at 38 punctures. The second at 52. The third at 61. All three were within their own test specification. None met my method conditions. The fix was not to buy a “better” septum. It was to requalify the septum under my actual autosampler parameters and then lock the supplier to the lot that passed. If your method requires more than 40 injections per vial, do not trust the datasheet puncture count. Run your own test with your needle, your solvent, and your sample temperature.
Glass type and hydrolytic class: when the number is a material property and when it is not
Datasheets list glass as “Type I, Class A” or “33 expansion borosilicate” or sometimes just “borosilicate.” These are material designations, not lot-level test results. The hydrolytic class per USP <660> is a test of the glass surface’s resistance to water attack. It is run on crushed glass or on the formed container, depending on which part of the chapter applies. A manufacturer that lists “USP Type I” has certified the glass formulation. That does not mean every vial from every production run has been hydrolytically tested.
The practical question for an HPLC user is not whether the glass is Type I. It is whether the vial interior has been treated or coated, and whether that treatment survives your sample matrix. Untreated Type I glass can adsorb basic analytes. Silanized or deactivated vials reduce that adsorption, but the deactivation layer is a surface treatment applied after forming. It is not a bulk material property. A datasheet that says “deactivated” without specifying the silanization chemistry or the residual silanol activity is giving you a marketing term.
Ask for the test method. A proper deactivation spec includes a test analyte, a concentration, a contact time, and a recovery percentage. For example: “Recovery of 1 ng/mL chlorhexidine after 24 h storage at 4 °C: > 90%.” That is a tested value. “Deactivated glass” is not.
Insert dimensions: the tolerances that actually matter are not the ones printed
A 300 µL fixed insert datasheet lists length, OD at the top flange, and capacity. The dimension that causes autosampler crashes is the flange OD and the flange thickness. If the flange is 0.1 mm too thick, the insert sits 0.1 mm higher in the vial. The autosampler needle depth is calibrated to the vial shoulder, not the insert flange. A 0.1 mm offset can mean the needle bottoms out on the insert cone instead of aspirating from the liquid pool. I have diagnosed three “low recovery” complaints that turned out to be needle bottoming on inserts whose flange thickness was at the upper end of the tolerance band.
The datasheet may not list flange thickness at all. If it does, the tolerance is often ± 0.05 mm or looser. That is enough to cause the problem. When you qualify a new insert lot, measure flange thickness on 10 inserts with a micrometer. If the spread exceeds 0.08 mm, reject the lot or adjust the needle depth offset in your method. The autosampler does not know the insert is there. It only knows the vial geometry.
Certificates of analysis: the difference between a datasheet and a lot release
A datasheet is a general document. A certificate of analysis (CoA) is lot-specific. The CoA lists the parameters that were actually measured on the lot you are buying, with the results and the pass/fail limits. If your supplier cannot provide a CoA for the specific lot, every number on the datasheet is a promise, not a measurement.
For HPLC vials used in regulated methods, the CoA is not optional. USP <621> and EPA SW-846 methods require that you demonstrate the suitability of your sample containers. A CoA is not proof of suitability, but it is the starting point. Without it, you are qualifying a product family, not a product lot.
Look at the CoA for the parameters that matter to your method. A typical CoA for a 9 mm screw-top vial with a PTFE/silicone septum lists: glass type (tested by composition or expansion), hydrolytic class (tested per USP <660>), vial dimensions (tested on a sample), septum thickness and hardness (tested), and puncture count (tested under stated conditions). If the CoA lists “residual volume” with no test condition, it was copied from the datasheet. If it lists “particle count” with a method reference like ISO 8871, it was measured.
Spotting a sheet that hides its test method
There are four patterns I see on datasheets that are designed to look rigorous while avoiding accountability.
First, the sheet lists a parameter with a tolerance but no test method. “Inner diameter: 6.0 ± 0.1 mm” is a claim. “Inner diameter: 6.0 ± 0.1 mm, measured per ISO 8362-1” is a test. If the method reference is missing, the tolerance is aspirational.
Second, the sheet lists a parameter with a test method that is not appropriate for HPLC vials. I have seen datasheets that cite ISO 8362, which is the standard for injection vials for parenteral use, not for analytical autosampler vials. The dimensions overlap, but the performance tests do not. A puncture test designed for a hypodermic needle through a rubber stopper at 45 degrees tells you nothing about an autosampler needle piercing a laminated septum at 90 degrees.
Third, the sheet borrows a value from a related product. A manufacturer that makes both crimp-top and screw-top vials may list the same septum puncture count for both. The crimp-top septum is compressed differently than the screw-top septum. The puncture mechanics are not the same. If the datasheet does not state the closure type for the puncture test, assume the number was generated on the easiest configuration.
Fourth, the sheet lists a range that is so wide it is meaningless. “Residual volume: 1 to 10 µL” covers every possible outcome. A range that wide is not a specification. It is a disclaimer.
Field case 1: the 96-well plate transfer that lost 30% of sample
A core lab was running a peptide assay with 25 µL of sample in a 300 µL polypropylene insert inside a 12 x 32 mm vial. The datasheet listed “Residual volume: 3 µL” for the insert. The lab was recovering 70% of their expected analyte. They assumed adsorption, switched to a silanized insert, and saw no improvement. I asked them to weigh the insert dry, add 25 µL of their mobile phase, aspirate with their autosampler settings, and weigh again. The residual volume was 7.8 µL, not 3 µL. The autosampler needle was set to 4 mm above the vial bottom, which is fine for a conical insert but wrong for a flat-bottom polypropylene insert. The needle could not reach the liquid pool at the bottom. The fix was to raise the needle depth setting to 1.5 mm above the bottom. Recovery went to 94%. The datasheet was not wrong. It was tested with a different needle depth than the lab was using.
Field case 2: the septum lot that passed the datasheet but failed the method
A contract lab running a stability study for a lyophilized drug product used a 9 mm screw-top vial with a PTFE/silicone septum rated for 100 punctures. After 60 injections, they saw increasing carryover and a rising baseline. They blamed the autosampler wash. The real problem was septum coring. The septum was rated for 100 punctures with a 23-gauge bevel needle. The lab used a 22-gauge blunt-tip needle for reconstitution and sampling. The blunt tip punched a core out of the septum on the first puncture. That core fell into the vial and floated in the sample, causing the baseline noise. The fix was to switch to a pre-slit septum designed for blunt-tip needles. The pre-slit septum had a lower puncture count rating on the datasheet, but it was the correct product for the needle geometry. The lab had optimized for the wrong number.
Decision checklist: what to verify before you buy
Use this checklist when a new vial or septum datasheet lands on your desk. Each line forces a specific verification step.
| Datasheet line | What to ask | Pass condition |
|---|---|---|
| Residual volume | Needle gauge, tip style, depth setting used for the test | Test condition matches your autosampler, or you run your own measurement |
| Septum puncture count | Needle gauge, tip style, temperature, solvent used for the test | Test condition matches your method, or you requalify on your own system |
| Glass type | USP <660> hydrolytic class, test method, CoA for the lot | CoA lists the glass type and class for your specific lot |
| Insert dimensions | Flange thickness tolerance, measured on how many pieces per lot | Flange thickness spread < 0.08 mm on your incoming inspection |
| Particle count | Method reference (ISO 8871, USP <788>), limit, sample size | Method reference present and limit stated in particles per mL |
| Deactivation / silanization | Test analyte, concentration, contact time, recovery % | Recovery % stated with a specific analyte and time point |
What the supplier’s own test lab will not tell you
When you ask a supplier for test data, you get the data they chose to generate. They will not volunteer that the puncture test was run on a different septum lot than the one you are buying. They will not tell you that the residual volume was measured on a prototype insert, not a production insert. They will not mention that the CoA they sent is a template with the lot number typed in, not a measurement record.
The only way to close that gap is to run your own qualification on the first lot you receive. That means measuring the dimensions that matter to your autosampler, running your own puncture test under your method conditions, and verifying residual volume with your needle and depth settings. It takes an afternoon. It is cheaper than a failed sequence or a rejected stability timepoint.
For labs that run high-throughput methods or regulated assays, I recommend qualifying at least two lots from the same supplier before locking the spec. Lot-to-lot variation in septum hardness and insert flange thickness is real. If you qualify one lot and then buy a different lot, you have not qualified your supply. You have qualified a snapshot. The chromatography consumables you buy in month six are not the same product you qualified in month one unless the CoA proves otherwise.
External resources can help you understand the test methods behind the numbers. The Agilent chromatography consumables and columns documentation includes autosampler needle specifications that affect residual volume and puncture behavior. Thermo Fisher chromatography and sample handling resources cover vial and closure compatibility in detail. For labs running environmental methods, U.S. EPA SW-846 test methods for waste and materials specify container cleanliness and compatibility requirements that a datasheet must support. Restek technical resources for chromatography publish chromatograms and recovery data that show how vial surface chemistry affects real analytes.
The number that is never on the sheet: your autosampler’s actual needle path
Every vial datasheet assumes a generic autosampler. Your autosampler has a specific needle path, a specific needle depth offset, a specific injection speed, and a specific needle wash sequence. None of those are on the vial datasheet. They are in your instrument method and your autosampler service manual.
Before you evaluate any vial spec, write down your autosampler’s needle gauge, tip style, depth setting from the vial shoulder, injection speed, and sample temperature. That is your reference condition. Every datasheet number you care about must be evaluated against that reference. If the datasheet does not state its test condition, you cannot evaluate it. You can only guess.
For lab vials in regulated environments, this reference condition belongs in your method validation documentation. When an auditor asks how you qualified your vials, the answer is not “the supplier’s datasheet says Type I.” The answer is “we measured residual volume, puncture count, and dimensional fit on our autosampler under method conditions, and the CoA for each lot confirms the glass type and cleanliness.” That answer holds up. The datasheet alone does not.
The Aijiren brand publishes test conditions on many of its vial and septum datasheets, which makes the qualification step faster. But even with a transparent supplier, you still run your own verification on the first lot. The supplier’s test lab does not have your autosampler.
Read the datasheet as a map, not a guarantee. The tested values tell you what the manufacturer measured on some sample at some point. The nominal values tell you what the design intended. Neither tells you what your autosampler will do with the vial in your hand. That measurement is yours to make.
References & further reading
Related resources from our lab consumables network:
- Hplc Vial at vialforhplc.com
- Chromatography Consumables at chromatographyconsumables.com
- Lab Vials at labvials.com
- Aijiren Brand at aijirenvials.com
External references and standards: