A 2 mL crimp-top vial lot arrived with a datasheet listing “septum hardness: 55 Shore A.” The lab ran 500 injections and saw split peaks on every twentieth vial. The hardness number was real, but it was a nominal value from the silicone compound supplier, not a measurement taken on the finished septum after molding and PTFE lamination. That distinction, between a number someone wrote down and a number someone actually tested, is the whole game when you read a gc vial datasheet specifications and what each value means.

Every line on a spec sheet falls into one of three buckets: a guaranteed value with a defined test method and an accept/reject limit, a typical value measured on a reference lot but not checked on yours, or a nominal value that describes the design target or the raw material input. Most buyers cannot tell which bucket they are reading because the sheet does not say. That is not an accident. A vendor who omits the method is usually selling a nominal number as if it were a guarantee.

I have sat on both sides of this table. When I ran a chromatography consumables operation, we shipped vials with a certificate that listed twelve parameters. Four were tested on every lot. The other eight were nominal, inherited from the glass tubing certificate or the septum compound datasheet. We were not hiding anything, but we also did not volunteer the distinction. Buyers who asked got straight answers. Buyers who did not ask assumed the whole sheet was verified. That assumption costs money.

The glass ID is a range, not a point

A 12 x 32 mm gc vial datasheet will list the outer diameter as 12.0 mm and the height as 32.0 mm. Both are nominal. The actual OD on a converted tubing line runs from 11.85 to 12.15 mm because the vial is drawn from borosilicate tubing, not machined. The autosampler tray pocket is cut to accept that spread. The number you need to care about is the inner diameter, because the ID sets the insert fit and the residual volume. A sheet that says “ID: 9.0 mm” without a tolerance is telling you the design target. A sheet that says “ID: 9.0 ± 0.15 mm” is telling you something about process control.

The ID is measured with a plug gauge or an optical comparator. Plug gauges are fast and cheap. An optical comparator gives you a true profile and catches ovality that a two-point plug gauge misses. If a vial is oval by 0.2 mm, the insert sits cocked, the autosampler needle clips the insert wall, and you get carryover. I have seen a lot rejected for an ID that passed the go/no-go gauge but failed the comparator because the ovality was concentrated in one axis. Ask the vendor which instrument they use. If they cannot answer, they are reselling someone else’s vials.

Septum hardness is the most abused number on the sheet

Hardness appears as Shore A, typically 45, 50, or 55 for GC septa. The number is measured with a durometer on a flat, molded plaque of the raw silicone compound before lamination. The finished septum is not flat, it is not the same thickness, and it has a PTFE layer bonded to one side. The durometer reading on the finished part is always higher, sometimes by 8 to 12 points, because the PTFE dominates the indentation response at shallow depths.

What actually matters for sealing is compression set and resealability after puncture. A septum with 55 Shore A hardness and 5% compression set will outperform a 45 Shore A septum with 25% compression set every time. The hardness number tells you almost nothing about how the septum behaves after the needle goes through it. The test that matters is a puncture-reseal test: a 23-gauge needle, 10 punctures at 0.5 mm offset, then a helium leak check at 40 psi. A good silicone/PTFE septum holds under 0.5 mL/min leak rate after 10 punctures. A datasheet that lists hardness but not puncture count or leak rate is telling you what the compound supplier told them, not what the finished septum does.

PTFE thickness is tested, but the bond is not

The PTFE layer on a GC septum is specified as 0.010 in (0.25 mm) or 0.020 in (0.50 mm) thick. That number is measured on the incoming PTFE film with a micrometer. It is a real measurement, but it is taken before lamination. The bond between PTFE and silicone is the failure point that no standard datasheet line captures. I have seen septa where the PTFE delaminated after three injections because the plasma treatment was too light or the silicone was contaminated with mold release.

The test for bond integrity is a peel test: a 90-degree pull on the PTFE edge at a defined rate, with the force recorded. A good bond fails cohesively in the silicone, not adhesively at the interface. A datasheet that lists “PTFE thickness: 0.010 in” and nothing else is giving you the film spec. Ask for the peel force or the delamination test result. Most vendors will not have it because they buy finished septa from a converter. That is fine, but it means the septum is a commodity input to their vial, not something they control.

Residual volume is calculated, not measured

The residual volume of a 2 mL vial is listed as 50 µL or 80 µL or 120 µL depending on the bottom geometry. That number is almost always calculated from the drawing, not measured on the finished part. The calculation assumes a perfect cone or a perfect flat bottom with a defined radius. The real bottom of a converted vial has a tool mark, a slight crown, or a radius that varies with the mandrel wear. The measured residual volume can be 20 to 40 µL higher than the calculated value on a worn tool.

The measurement method is straightforward: weigh the empty vial, fill it with water, withdraw with a syringe until the meniscus just touches the bottom center, and weigh again. The difference divided by density gives the residual volume. It takes five minutes per vial. Almost no one does it on a production basis because it is slow and the number is not a critical acceptance criterion for most methods. But if you are running 50 µL sample volumes in a 2 mL vial, a residual volume that is actually 90 µL means your autosampler is pulling from a pool that is nearly twice what you planned. The sheet that says “residual volume: 50 µL” is almost certainly a CAD number.

The neck dimensions are where the seal is won or lost

A 9 mm crimp neck has three dimensions that decide whether the cap actually seals: the neck OD, the neck height from the shoulder to the rim, and the rim radius. The neck OD is specified as 9.0 mm with a tolerance of ± 0.05 mm on a good sheet. The neck height is 8.0 mm nominal. The rim radius is the one that never appears on a datasheet, and it is the one that causes the most field failures.

If the rim radius is too sharp, the aluminum cap tears the septum at the crimp point. If it is too flat, the cap does not develop enough radial compression and the septum leaks under inlet pressure. The radius is controlled by the forming tool, and it wears over time. A vendor who measures neck OD with a go/no-go ring gauge but never checks the rim profile on a comparator is shipping vials that pass the spec sheet and fail in the autosampler. I have traced a 5% leak rate on a 500-vial lot to a rim radius that had drifted from 0.3 mm to 0.15 mm over six months of tool use. The OD was still in tolerance. The sheet looked perfect.

Glass type is a claim, not a measurement

“Type I borosilicate 33 expansion” appears on nearly every GC vial datasheet. The expansion coefficient, 3.3 x 10⁻⁶ K⁻¹, is a material property of the glass composition, not a measurement taken on your lot. The glass tubing supplier certifies the composition. The vial converter buys the tubing certificate and passes the claim through. If the converter is honest, they will list the tubing manufacturer and the lot number so you can trace the claim. If they are not, they will just write “Type I” and hope you do not ask.

The test that matters for a GC vial is hydrolytic resistance, measured per USP <660> or ISO 4802. The glass is ground, autoclaved in water, and the extract is titrated. Type I glass gives a titration volume under 1.0 mL of 0.01 N HCl per gram. That is a real test, and it is destructive, so it is done on a sample from each melt, not on every vial. A datasheet that lists the expansion coefficient but not the hydrolytic class is giving you the tubing spec, not the vial spec. The two are linked but not identical, because the vial forming process can introduce surface defects that change the extractable profile.

Silanization is the line that hides the most

A deactivated vial datasheet will say “silanized” or “deactivated” with no further detail. That word covers at least four different processes: vapor-phase DMDCS, liquid-phase DMDCS, HMDS, or a proprietary silane blend. The processes give different surface coverage, different hydrolytic stability, and different extractable profiles. A vapor-phase DMDCS treatment on a clean, dry surface gives a water contact angle of 95 to 105 degrees. A sloppy liquid-phase dip gives 70 to 80 degrees and leaves unreacted silane that bleeds into your solvent.

The test for silanization quality is a contact angle measurement or a dye adsorption test with a polar probe. Few vial vendors run either on a production basis. They treat the vials and ship them. The datasheet says “deactivated” because the process was performed, not because the surface was verified. If you are running sub-nanogram pesticide or PAH methods, ask for the contact angle range and the batch record for the silanization step. If the vendor cannot produce it, the deactivation is a nominal claim.

Certificate of analysis: which lines are actually checked on your lot

A COA that lists fifteen parameters looks impressive. The question is which of those fifteen were measured on the lot in your hand, and which were copied from a type test or a raw material certificate. A type test is a one-time qualification on a reference lot. A raw material certificate is a document from the glass tubing supplier or the septum compound supplier. Neither tells you anything about the lot you received.

The lines that are almost always tested on every lot are: outer diameter, height, neck OD, and visual defects (cracks, chips, stones, bubbles). The lines that are usually type-tested are: hydrolytic resistance, silanization quality, PTFE bond strength, and septum puncture reseal. The lines that are usually nominal are: residual volume, glass expansion coefficient, septum hardness, and PTFE thickness. A COA that does not distinguish these categories is not a certificate, it is a compilation.

I once audited a supplier who shipped a COA with “silanization: pass” on every lot. The test method was a visual inspection of the vial surface for water beading. The operator held the vial under a tap and looked at it. That is not a measurement. That is a vibe. The lot passed the vibe test and failed a real contact angle measurement at 72 degrees. The lab saw analyte tailing on polar compounds for three weeks before we traced it back.

How to spot a sheet that hides its test method

A vendor who tests their vials will tell you how. The test method is not proprietary. A durometer reading is a durometer reading. A plug gauge is a plug gauge. The only reason to omit the method is that the number was not generated by a method at all. Here is the checklist I use when I read a gc vial datasheet specifications and what each value means:

• Every number has a tolerance or a range. A point value with no tolerance is a nominal target, not a measurement. “ID: 9.0 mm” is a drawing. “ID: 9.0 ± 0.15 mm” is a process.
• Every tested value has a method reference. Look for USP <660>, ISO 4802, ASTM D2240 for hardness, or a defined internal method. “Measured by our QC department” is not a method.
• Every method has an instrument and a sample size. “Optical comparator, n=32 per lot” tells you something. “Conforms to specification” tells you nothing.
• The COA separates tested, type-tested, and nominal values. If all fifteen lines look identical in status, the vendor is not distinguishing them. That is a red flag, not a convenience.
• The sheet lists the raw material lot numbers. Glass tubing lot, septum compound lot, PTFE film lot. If these are missing, you cannot trace a failure back to the input.

Field case: the septum that passed every line and still leaked

A contract lab running EPA 8270 semivolatiles switched to a lower-cost 2 mL crimp vial. The datasheet listed 55 Shore A hardness, 0.010 in PTFE, Type I glass, and a COA with twelve passing lines. Within two weeks, the lab saw erratic injection volumes on the autosampler. The symptom was a 15% RSD on the internal standard area, up from 3% on the previous vial lot.

The measurement that identified the failure was a septum compression test on the finished assembly. We crimped caps onto vials from both lots and measured the septum indentation depth with a dial indicator. The old lot compressed 0.45 mm under the crimp. The new lot compressed 0.25 mm. The difference was the silicone compound’s crosslink density, which the hardness number did not capture. The new lot was harder at the surface but softer in the bulk, so the crimp did not develop enough clamping force. The fix was to switch back to the previous septum compound and add a compression set requirement of under 10% after 22 hours at 70°C to the spec. The hardness line stayed at 55 Shore A. The failure was invisible on the original sheet.

Field case: the insert that did not fit the ID that was in tolerance

A pharmaceutical QC lab bought 100,000 2 mL vials with a spring insert. The datasheet listed the ID as 9.0 ± 0.15 mm. The insert OD was 8.6 mm. On paper, the clearance was 0.4 mm nominal, 0.25 mm minimum. In practice, 3% of the vials had the insert jam at the bottom, and the autosampler needle bent when it hit the cocked insert.

The ID measurement on the failed vials was 8.92 mm, inside the tolerance. The problem was ovality. The ID measured 9.05 mm in one axis and 8.92 mm in the perpendicular axis. The plug gauge passed because it measured the minimum axis. The insert jammed because the maximum axis was not enough to clear the insert at the bottom taper. The fix was to add an ovality requirement of under 0.05 mm measured on an optical comparator at three depths. The vendor had to change their forming mandrel to hold that spec. The original sheet was technically accurate and functionally useless.

What a datasheet should look like if the vendor actually tests

A serious gc vial datasheet from a factory that runs its own QC will have a table that looks like this. I am using the parameters I would ship if I were still running the line today:

That table tells you exactly what the vendor knows and what they are passing through. The nominal lines are not worthless, they just are not verified on your lot. The tested lines are the ones you can hold them to. If a vendor sends you a sheet where every line is a point value with no method and no sample size, they are not testing. They are typing.

What to ask before you place the order

You do not need to audit the factory. You need to ask five questions that separate a tested datasheet from a compiled one. I have used these questions with every supplier I have qualified, and the answers predict field performance better than the sheet itself.

Question 1: Which lines on this COA were measured on the lot I will receive? If the answer is “all of them,” ask for the test records. If the answer is “the dimensional ones,” you have an honest vendor. If the answer is a pause, you have a reseller.

Question 2: What is the sample size for the dimensional checks? A lot of 100,000 vials checked with n=5 is not a statistical sample. A vendor who checks n=50 or n=100 per lot is running a real process. The sample size tells you their confidence in their own line.

Question 3: Do you measure ovality or just ID? This question catches the vendors who use a plug gauge and call it a day. If they do not know what ovality means, they are not measuring the ID in a way that predicts insert fit.

Question 4: What is the compression set of the septum compound? If they cannot answer, they are buying finished septa and passing through the hardness number. That is acceptable if you are running routine methods with wide tolerances. It is not acceptable if you are running 8270 or 525.2 at low levels.

Question 5: Can you send the raw material lot numbers for the glass tubing and the septum compound? If they can, you have traceability. If they cannot, any failure investigation stops at the vendor’s door. For regulated work, that is a non-starter.

The decision rule

A gc vial datasheet is only as good as the distinction between what was measured and what was assumed. The dimensional lines you can verify on arrival with a caliper and a plug gauge. The material lines you have to take on trust, unless the vendor gives you the raw material lots and the type test records. The septum lines are where most failures hide, because hardness and PTFE thickness are nominal values that do not predict sealing performance. The decision rule is simple: buy from a vendor who separates tested, type-tested, and nominal values on the sheet, and who can produce the test records for the tested lines when you ask. If the sheet is a wall of point values with no methods and no sample sizes, the price difference is not a discount. It is the cost of the testing they are not doing, and you will pay it in failed runs. The same logic applies when you buy an hplc autosampler vial or a sample vial for a different workflow. A reagent bottle datasheet has the same three buckets, and the same questions apply. For method-specific requirements, check the Agilent chromatography consumables and columns documentation or the Thermo Fisher chromatography and sample handling guides for the autosampler constraints. If you are running environmental methods, the U.S. EPA SW-846 test methods for waste and materials will tell you which vial parameters are actually specified in the method. And if you want a second opinion on septum performance, the Restek technical resources for chromatography have published puncture and reseal data that is worth comparing against your vendor’s claims.

References & further reading

Related resources from our lab consumables network:

• Gc Vial at gcvialfactory.com
• Reagent Bottle at reagentbottleforsale.com
• Hplc Autosampler Vial at hplcautosamplervials.com
• Sample Vial at samplevialprice.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