A lot of 9 mm screw-thread vials arrived with a datasheet claiming 0.8 mm septa and a 60 Shore A durometer. The lab ran 96 injections on the autosampler and got three split peaks and one coring event at injection 61. The septa measured 0.61 mm on a micrometer, and the durometer read 47 Shore A on a hand-held gauge. Nothing on the sheet was a lie. It was all nominal. That is the single most expensive distinction in sample vial procurement: a spec sheet is a mix of tested values, calculated values, purchased-component values, and numbers copied from a mold drawing that nobody has verified since the tool was cut. Reading a sample vial datasheet specifications correctly is less about chemistry than about knowing which lines carry a test report behind them and which carry a wish.
I have bought vials from four continents and rejected two lots in the last three years. Both rejections came down to reading the sheet more carefully than the supplier expected me to. This guide is the checklist I use now, line by line, with the measurement method I ask for and the tolerance I accept.
Every number on a vial datasheet falls into one of three buckets, and suppliers rarely label them.
- Tested (guaranteed): a value measured on a sample from the lot, with a stated method and a stated acceptance range. If the lot fails, the supplier owes you a replacement. Dimensional values from a CMM or optical comparator usually live here.
- Typical: a value measured on development lots or a historical average, with no per-lot verification. Septum durometer and residual extractables often sit here. The number is real but not a commitment.
- Nominal: a design value from the drawing. Wall thickness, neck ID, and crimp skirt diameter are frequently nominal. The mold wears; the drawing does not.
The word “typical” on a datasheet is a legal signal. It means the supplier will not replace the lot if your measurement differs. When I see typical next to a number that matters to my method, I ask for the tested value and the method. If they cannot produce it, I treat the number as decoration.
The dimensional block: what a CMM actually measures and what it does not
Dimensional lines are the most trustworthy part of a good sheet and the most misleading part of a bad one. The reason is that “9 mm” is a family name, not a specification. The real numbers that decide fit are the ones below.
| Spec line | Typical printed value | How it should be measured | My acceptance window |
|---|---|---|---|
| Neck OD (crimp) | 11.0 mm nominal | Optical comparator, 10 pieces per lot | 10.95 to 11.05 mm |
| Neck ID | 6.0 mm nominal | Pin gauge, go/no-go | 5.95 to 6.05 mm |
| Body OD | 9.0 mm nominal | CMM, 10 pieces | 8.90 to 9.10 mm |
| Height (crimp, 12×32) | 32.0 mm nominal | Height gauge, 10 pieces | 31.7 to 32.3 mm |
| Wall thickness | 1.0 mm nominal | Sectioned piece, optical | 0.85 to 1.15 mm |
| Bottom thickness | 1.0 mm nominal | Sectioned piece, optical | 0.80 to 1.20 mm |
Two things to notice. First, the wall and bottom are almost always nominal and almost always the first to drift as the mold wears. A vial that reads 0.75 mm at the bottom will crack under a crimper set for 1.0 mm. Second, the neck ID is the number that governs septum seating. If the sheet gives you 6.0 mm with no tolerance, ask for the go/no-go pin gauge result. A neck that runs 5.85 mm will let a 6.0 mm septum walk during injection.
If your supplier cannot tell you which of these are tested per lot, assume they are all nominal and plan your incoming inspection accordingly. The sample vial dimensional drawing on file is a starting point, not a certificate.
Septum specs: durometer, thickness, and the puncture count that is never on the sheet
Septum lines are where nominal values do the most damage because the failure mode is silent. A soft septum does not fail at injection one. It fails at injection 40 to 80, after the needle has carved a channel you cannot see until the peak splits.
Durometer
A sheet will say 55 Shore A or 60 Shore A. Ask whether that is a molded-sheet value or a finished-part value. The two differ by 3 to 5 points because the finished septum is thinner and the gauge reads differently on a curved surface. I ask for the durometer on the finished part, measured per ASTM D2240, with a tolerance of ±5. Anything printed without a tolerance is typical, not tested.
Thickness
For 11 mm crimp vials, the standard septum is 1.0 mm nominal, and the sheet will often print 1.0 with no range. Measured values on incoming lots run 0.85 to 1.15 mm. A 0.85 mm septum in a crimper set for 1.0 mm will over-compress and split at the skirt. I reject below 0.90 mm and above 1.10 mm on a 10-piece sample.
Puncture count
This is the number that is almost never on the sheet, and it is the one I care about most. Ask for the puncture count to failure, with the needle gauge stated. A PTFE-faced silicone septum on a 22-gauge needle typically holds 60 to 100 punctures before the coring rate exceeds 1 in 100. If the sheet does not state needle gauge, the puncture number is meaningless. A 26-gauge needle will triple the count on the same septum.
For methods that run more than 50 injections per vial, I specify a septum rated for at least 80 punctures at 22 gauge and I verify it on a 20-vial sample by running a dye test. If the sheet omits the needle gauge, I ask. If they cannot answer, I buy elsewhere.
Chemical and thermal lines: which are certificates and which are copied from a resin data sheet
The chemical compatibility block on a vial datasheet is usually copied from the resin supplier’s data sheet, not tested on the finished vial. That is not dishonest, but it means the number describes the polymer, not the part. The finished septum has been molded, cured, and washed, and its extractables profile is not the resin’s profile.
What to look for instead:
- Extractable test method. A real sheet cites a method, for example USP <661> for plastic packaging or an in-house GC-MS screen at a stated temperature and time. If the line says “low extractables” with no method, it is marketing.
- Thermal rating. A septa sheet may say 200 °C maximum. Ask whether that is continuous or short-term. Continuous ratings on silicone septa are usually 20 to 30 °C lower than the headline number.
- Autoclavability. If the vial is sold as sterile or autoclavable, the sheet should state the cycle (for example 121 °C, 20 min) and whether the certification is per lot or per shipment.
For methods that follow U.S. EPA SW-846 protocols, the extractables question is not academic. A volatile screen on a septum with unstated extractables will show siloxane peaks at m/z 73, 147, and 207 that you will spend a day chasing. I ask for a GC-MS blank run at the method temperature before I approve a new lot.
Case one: the 0.61 mm septum that passed incoming inspection
We had a supplier whose sheet printed 1.0 mm ± 0.1 mm on the septum. Incoming inspection used a go/no-go gauge set at 0.85 mm minimum, and the lot passed. But the gauge measured the center of the septum, and the lot had been die-cut with a worn punch that thinned the edge to 0.61 mm. The autosampler needle entered near the edge on a 96-well tray and cored the septum at injection 61.
The measurement that identified it was a cross-section under a stereo microscope at 20x, taken from five septa per lot. The center read 0.98 mm; the edge read 0.61 mm. The fix was twofold: we changed the incoming gauge to a five-point thickness check (center plus four quadrants), and we added a needle-entry position check to the autosampler method so the needle lands within 1.5 mm of center. The supplier retooled the punch. We now reject any septum with more than a 0.15 mm center-to-edge variation.
Case two: the “200 °C” septum that outgassed at 180
A different supplier sold a PTFE-faced silicone septum rated 200 °C maximum. Our method ran at 180 °C inlet. Within two weeks the baseline on a GC-FID drifted upward by 40 pA over a 30-minute run and the column showed siloxane bleed at m/z 73. The sheet was not wrong about the material’s thermal limit. It was wrong about the finished part, because the cure was incomplete.
The measurement that found it was a blank injection at 180 °C with the inlet purge off for 60 seconds, followed by a temperature ramp. The blank showed a siloxane envelope starting at 175 °C. We asked the supplier for the post-cure schedule. They had none. We switched to a septum with a stated 2-hour post-cure at 200 °C and a lot-level GC-MS blank certificate. The baseline drift dropped to under 5 pA.
The lesson: a thermal rating without a cure schedule is a resin number, not a part number. Ask for the post-cure conditions and the blank-run certificate. A supplier who cannot produce either is selling you the resin’s reputation, not the septum’s performance.
The lines that hide their test method: a five-question filter
When a sheet looks complete but feels vague, I run it through five questions. Any sheet that fails two or more goes back to the supplier before I place an order.
- Does every number that matters have a tolerance? A number without a tolerance is nominal. Ask for the tested range.
- Is the test method cited? ASTM D2240 for durometer, USP <661> for extractables, a stated needle gauge for puncture count. No method, no guarantee.
- Is the value per lot or per shipment? A certificate of analysis that covers a shipment of 50,000 vials is not a per-lot test. Ask for the sampling plan.
- Is the value on the finished part or the raw material? Resin data sheets are not vial data sheets.
- Does the sheet state the failure mode? A good sheet tells you what the part is not rated for. A sheet with no exclusions is a sheet with no testing.
The comparison table below is the one I keep on the wall. It separates the lines I will accept on trust from the lines I verify on every incoming lot.
| Spec line | Trust level | Verification I run | Reject threshold |
|---|---|---|---|
| Body OD | Tested, per lot | CMM, 10 pieces | Outside 8.90 to 9.10 mm |
| Neck ID | Tested, per lot | Go/no-go pin gauge | Fail on either pin |
| Septum thickness | Nominal on most sheets | 5-point micrometer, 10 pieces | Below 0.90 or above 1.10 mm |
| Septum durometer | Typical on most sheets | ASTM D2240 on finished part | Outside ±5 of stated |
| Puncture count | Rarely stated | Dye test, 20 vials, 22-gauge needle | Below 80 punctures |
| Extractables | Resin data sheet | GC-MS blank at method temperature | Any siloxane envelope below method temp |
| Thermal max | Resin data sheet | Blank run at rated temperature | Baseline drift above 10 pA |
What a crimp or screw-thread certificate should contain, and what it usually omits
A certificate of analysis for a crimp vial lot should list the lot number, the molding date, the resin lot, the septum lot, and the results of the dimensional and septa tests. In practice, most certificates list only the lot number and a pass/fail stamp. That is a shipping document, not a certificate.
What I ask for, and what a competent supplier will provide:
- Lot traceability from resin to finished vial, with the molding date and the mold cavity number.
- Dimensional results for body OD, neck ID, and height, with the measurement method and sample size.
- Septum results for thickness, durometer, and puncture count, with the needle gauge.
- Extractables blank at the method temperature, or a statement that the lot was not tested and the value is typical.
If the certificate omits the method, ask for the method. If the supplier cannot produce it, the certificate is a form letter. The vials with caps assembly is only as good as the weakest certificate in the chain, and the septum certificate is usually the weak one.
Troubleshooting: reading a sheet against a failing run
When a run fails, the sheet tells you where to look first. This is the checklist I use, in order.
- Split peaks or ghost peaks after injection 40. Check the septum puncture count and needle gauge on the sheet. If the sheet does not state the gauge, assume it is the cause and run a dye test.
- Baseline drift at method temperature. Check whether the thermal rating is continuous or short-term, and whether a post-cure schedule is stated. Run a blank at method temperature.
- Cracked vials after crimping. Check wall and bottom thickness against the nominal. A bottom below 0.80 mm will crack at standard crimp torque.
- Needle bending or septum walk. Check neck ID against the go/no-go pin result. A neck below 5.95 mm will not seat a 6.0 mm septum.
- Siloxane peaks in a volatile screen. Check the extractables method. If the sheet cites only the resin, run a GC-MS blank before blaming the column.
For method-specific troubleshooting, the technical notes at Restek and the application libraries from Agilent and Thermo Fisher are worth keeping on file, but they will not tell you what your supplier tested. That is a conversation with the supplier, not a literature search.
The buying rule: treat every untested line as a risk you are accepting
The decision rule is simple. For every line on the datasheet, ask whether the supplier will replace the lot if your measurement differs. If the answer is yes, the line is tested and you can plan around it. If the answer is no, the line is nominal or typical, and you are accepting the risk. Price the risk accordingly: a septum that is nominal on thickness and durometer is worth 20 to 30 percent less than one that is tested per lot, because you will spend that difference on rejected runs and incoming inspection. When you compare quotes, compare the tested lines, not the printed ones. The supplier who can show you the method, the sample size, and the needle gauge is the one whose sheet you can actually read. The rest are selling you a drawing. If you are sourcing across a full consumables line, the chromatography products catalog and the manufacturing notes at Aijiren are useful references for what a tested specification looks like, but the certificate in your hand is the only document that binds the lot.
References & further reading
Related resources from our lab consumables network:
- Sample Vial at samplevialprice.com
- Chromatography Products at chromatographyselling.com
- Aijiren at aijirentech.com
- Vials With Caps at vialswithcaps.com
External references and standards: