The line that costs the most money is the one with no test method attached

Last month a contract lab ran 2,400 headspace samples through a new lot of 20 mL crimp-top vials. The supplier datasheet said “septum: PTFE/silicone, resealing after 3 punctures.” The lab’s autosampler used a 23-gauge, 51 mm needle with a bevel that had been resharpened once. At sample 1,900, carryover on the blank check vials jumped from 0.03% to 1.7%. The root cause was not the needle, not the wash station, not the liner. It was a datasheet that listed a reseal spec without stating the test needle gauge, the puncture speed, the injection depth, or whether the value was tested on every lot or just on the first qualification batch. If you buy lab vials and you cannot tell which numbers on a lab vials datasheet specifications and what each value means are measured and which are just nominal, you will eventually own that same failure. This article goes line by line through a typical spec sheet and separates the numbers that were earned in a test from the numbers that were copied from a drawing.

Every vial datasheet has a column for “specification” and a column for “value.” What most sheets do not have is a column for “how we know.” That missing column is where the expensive mistakes hide. A value like 12 x 32 mm for a 2 mL autosampler vial means nothing until you know whether someone measured the finished vial with a calibrated gauge or whether the number is the nominal drawing dimension from the mold cavity. Those two numbers can differ by 0.2 mm or more, and 0.2 mm is enough to change how a vial seats in a tray, how the cap crimps, and whether the needle tip reaches the sample at the correct depth.

Start by asking one question for every line on the sheet: is this value tested, guaranteed, or nominal? A tested value has a defined method, an instrument, a sample size, and an acceptance limit. A guaranteed value is one the supplier will replace product over if it fails, whether or not they test it on every lot. A nominal value is a design target. It is often printed to three decimal places on a sheet where the actual process tolerance is ten times looser. None of these three categories are inherently bad. A nominal value for outer diameter is fine if your application does not care about 0.15 mm of variation. A nominal value for septum reseal is not fine if you are running 100 injections per vial in a regulated method.

When I evaluate a supplier, I look for the test method references first. A sheet that says “ISO 8362-1” for glass dimensions or “ISO 11040” for finish geometry is telling me the value was produced by a defined measurement procedure. A sheet that says “12 x 32 mm” with no standard, no gauge description, and no tolerance is telling me the number came from the CAD file. Both sheets might sell the same physical vial. The difference is what happens when the lot drifts. The first supplier will catch the drift and reject the lot. The second supplier will ship it and let your autosampler find the problem.

Glass dimensions: the difference between a go/no-go gauge and a caliper reading

The first block on most vial datasheets covers the glass body: outer diameter, height, wall thickness, and bottom configuration. These numbers look simple. They are not. The measurement method changes the number itself, and buyers who do not know this will compare two suppliers’ sheets as if the numbers were interchangeable.

Outer diameter (OD) for a 2 mL screw-top vial is typically specified as 11.6 mm or 12.0 mm, depending on the tray system. A reputable sheet will list a tolerance, usually ±0.15 mm or tighter. Here is the part most people miss: OD is not a single number. Glass tubing has ovality. The OD measured at one rotational position can differ from the OD measured 90 degrees away by 0.05 mm to 0.2 mm, depending on how well the tubing station controlled the draw. A sheet that lists one OD value with a tolerance is implicitly claiming that the maximum ovality is within that tolerance band. A sheet that does not mention ovality at all is either testing it and not saying so, or not testing it. Both are problems for you.

Ask the supplier how they measure OD. A go/no-go ring gauge checks whether the vial passes through a hole of a fixed diameter. It tells you the maximum OD at any point but not the minimum, not the ovality, and not where along the body the maximum occurs. A laser micrometer or a coordinate measuring machine (CMM) gives you actual dimensional profiles. If your autosampler tray has tight pockets, you need the laser or CMM data. If your tray has generous clearance, the ring gauge is probably enough. The datasheet will rarely tell you which method was used. You have to ask, and the answer tells you a lot about the supplier’s quality system.

Height is the other dimension that seems trivial and is not. A 32 mm vial is never exactly 32 mm. The tolerance is usually ±0.25 mm for molded vials and tighter for tubing vials, around ±0.15 mm. Height matters because the cap must compress the septum to a specific range. If the glass is short and the cap is designed for a taller vial, the septum gets less compression and the seal fails early. If the glass is tall, the cap may not fully engage the crimp or thread. The datasheet lists a height. What it does not list is whether the height is measured with the vial standing on its base or hanging from its finish. For a flat-bottom vial, those are the same. For a vial with a slight convex or concave bottom, they differ by 0.1 mm or more. Again, the method matters.

One more line on the glass block that hides a lot: wall thickness. A typical 2 mL vial has a wall thickness of 0.9 to 1.1 mm. The datasheet may list “1.0 mm” as the value. What it does not say is where that thickness is measured. Wall thickness varies from the shoulder to the base due to the forming process. A vial that is 1.0 mm at the midpoint can be 0.7 mm at the shoulder, and that thin shoulder is where thermal shock cracks start. If your method involves freezing samples at -80°C and then thawing, ask for the minimum wall thickness, not the nominal. A supplier who cannot give you a minimum is not measuring thickness distribution.

Finish dimensions: the threads and crimp neck that never get checked on incoming inspection

The finish is the part of the vial that mates with the cap. For screw-top vials, the relevant standard is GPI (Glass Packaging Institute) finish 13-425 or 15-425, where the first number is the approximate outer diameter in millimeters and the second is the thread pitch in turns per 25.4 mm. For crimp-top vials, the finish is defined by the 11 mm or 20 mm neck geometry from ISO 8362-1 or the equivalent DIN standard.

Here is what I have seen in the field. A lab buys 15-425 screw-top vials from a new supplier because the datasheet says “15-425 finish.” The caps from their existing supplier do not seal. The lab measures the finish with a thread gauge and finds the pitch is correct but the thread depth is 0.15 mm shallower than the caps expect. The datasheet said the right thing. The process did not hold the thread depth. The supplier’s incoming inspection checked the pitch with a go gauge and never checked the depth. The fix was not a new cap. It was a supplier who could show thread profile data, not just a go/no-go result.

For crimp-top vials, the critical finish dimensions are the neck outer diameter, the neck height, and the bead radius. The bead is the raised ring that the aluminum cap crimps over. If the bead radius is too large, the cap does not seat fully and the septum gets uneven compression. If the bead radius is too small, the cap bottoms out on the glass and the septum is over-compressed, which can cause the septum to extrude into the vial or the cap to split. The datasheet rarely lists the bead radius at all. It is one of those dimensions that gets checked during mold qualification and then assumed stable. For molded vials, that assumption is usually fine for a few years. For tubing vials that are flame-formed, the bead is created by a different process and can vary more. If you are buying tubing vials for a crimp application, ask for bead geometry data for the specific lot, not the catalog drawing.

The practical test for finish quality is simple and I recommend doing it on every incoming lot. Take 20 vials, cap them with your production caps at your production torque or crimp force, and then measure the septum indentation depth with a depth micrometer. The indentation should be consistent within ±0.05 mm across all 20. If it varies more than that, the finish dimensions are varying, and the datasheet’s nominal numbers are not being held.

Septum specs: the reseal number is the most misread line in the catalog

The septum section of a datasheet usually lists the material layers, the thickness, the hardness, and a reseal or puncture count. Every one of those lines has a hidden test method, and the reseal line is the worst offender.

Material layers are typically listed as “PTFE/silicone” or “PTFE/silicone/PTFE.” The PTFE face is the barrier that contacts the sample. The silicone core provides the resealing elasticity. The thickness of the PTFE layer is the critical number for chemical resistance. A sheet that says “PTFE/silicone” without a PTFE thickness is not giving you enough to compare. A 0.25 mm PTFE face behaves very differently from a 1.0 mm PTFE face when you are injecting chlorinated solvents. The thin face is fine for aqueous samples. The thick face is what you need for aggressive organics. The thinner face also costs less, which is why some suppliers default to it and do not print the number prominently.

Hardness is listed in Shore A durometer for the silicone layer, typically 40 to 60 Shore A. This number is tested with a durometer per ASTM D2240 or ISO 7619. The test is cheap and fast, so most suppliers actually do measure it. The problem is that Shore A is a bulk property. It does not tell you how the septum behaves under a 23-gauge needle at 40°C after 50 injections. Two septa with the same Shore A can have different compression set, different tear strength, and different reseal performance. The Shore A number is a useful lot-to-lot consistency check, not a predictor of application performance.

The reseal or puncture count line is where I have seen the most expensive misunderstandings. A sheet that says “reseals after 3 punctures” without stating the needle gauge, needle type, puncture speed, and test temperature is not giving you a specification. It is giving you a marketing claim. The reseal performance of a septum depends on all four of those variables. A septum that reseals after 3 punctures with a 26-gauge conical needle may not reseal after 2 punctures with a 23-gauge side-port needle. The side-port needle cuts a larger, differently shaped hole. The conical needle pushes material aside and the silicone can recover. The side-port needle removes material and the recovery is worse.

The test method for reseal is usually based on ISO 8362-5 or a supplier-specific method. The ISO method specifies a needle gauge and a test procedure, but many suppliers run a modified version and still cite the standard. When you see “reseals after X punctures,” ask for the full method: needle gauge, needle type, puncture speed, injection depth, test temperature, and the acceptance criterion. What does “reseals” mean in that test? Does it mean no visible leak under a specified pressure? Does it mean no sample loss above a specified weight? Does it mean no carryover above a specified concentration? Those are different standards, and the number of punctures a septum survives under each will differ.

For a practical incoming test, I recommend running your own puncture test with your production needle and your production injection volume. Take 10 septa, puncture each one 10 times at your normal injection speed, and then run a blank injection. If carryover on any blank exceeds your method limit, the septum lot is not acceptable regardless of what the datasheet says. This test takes 30 minutes and it will catch problems that a supplier’s qualification data will not.

Temperature and pressure ratings: the numbers that are tested on a different product than the one you are buying

Many datasheets list a temperature range for the vial, such as -80°C to 250°C, and a pressure rating for headspace vials, such as 150 psi or 10 bar. These numbers are real, but they are often tested on a specific configuration that may not match your use.

The temperature range for a glass vial is limited by the glass composition and the thermal shock resistance, not by a single continuous-use temperature. A borosilicate glass vial can survive brief exposure to 250°C in a GC inlet, but it will crack if you take it from a -80°C freezer and place it directly into a 100°C water bath. The datasheet lists the endpoints but not the rate. The thermal shock resistance is tested per ISO 718 or a similar method, which specifies a temperature differential and a transfer time. If your process uses a faster transfer than the test, the rating does not apply. Ask for the thermal shock test conditions, not just the temperature range.

The pressure rating on a headspace vial is the number I trust least without seeing the test. A 20 mL headspace vial is typically rated for 150 psi (10.3 bar) at room temperature. That number is derived from the burst pressure of the glass with a safety factor, usually 3x or 4x. The test is done on a limited sample of vials, often during initial qualification. The production lots are not burst-tested. They are checked for dimensional conformance and visual defects, and the pressure rating is assumed to hold. That assumption is usually fine if the glass source and forming process do not change. It is not fine if the supplier switches glass tubing sources or changes the forming parameters. When a headspace vial fails under pressure, it fails catastrophically, and the glass shards end up in your autosampler or your operator’s hand. I treat the pressure rating as a design value, not a lot-level guarantee, unless the supplier can show burst test data for the specific lot I am buying.

For headspace applications, the relevant standard is EPA SW-846 Method 5021 or similar, which specifies the equilibration temperature and time for volatile organic analysis. The vial must withstand the equilibration pressure at the method temperature. A vial rated for 150 psi at 25°C may only be rated for 60 psi at 90°C because the glass strength decreases with temperature. The datasheet rarely states the pressure rating at the method temperature. You have to calculate it or ask the supplier for a derating curve. Most suppliers do not have one. That is a red flag if you are running high-temperature headspace methods.

The certificate of analysis: where the tested numbers actually live

The datasheet is the catalog page. The certificate of analysis (CoA) is the document that tells you what was actually measured on your lot. The two are not the same thing, and buyers who treat them as equivalent are buying on faith.

A proper CoA for a lab vial lot should list the lot number, the production date, the glass source, and the measured values for the critical dimensions and performance parameters. The measured values should be actual numbers, not “conforms.” A CoA that says “OD: conforms” without giving the measured value is not a CoA. It is a statement of faith. A CoA that gives measured values for OD, height, and finish dimensions, plus the test method for each, is evidence that the supplier actually measured the lot.

Here is what I look for on a CoA. The OD measurement should include the number of samples measured and the range, not just the average. A lot of 10,000 vials where 5 samples measured 11.55 mm, 11.58 mm, 11.52 mm, 11.60 mm, and 11.54 mm is a different lot from one where the average is 11.56 mm but the range is 11.40 to 11.72 mm. The average alone hides the outliers. The septum hardness should include the Shore A value and the test standard. The reseal test, if run, should include the needle gauge and the number of punctures tested. If the CoA does not list the needle gauge, the reseal number is not traceable to your application.

Many suppliers will provide a CoA only on request. Some will provide it automatically with every shipment. The difference is not trivial. A supplier who provides a CoA automatically has a quality system that generates the data as part of production. A supplier who provides it only on request may be generating it after the fact from the same nominal values that are on the datasheet. Ask for a CoA from the last three lots, not just the current one. If the measured values are identical across all three lots to the third decimal place, the supplier is not measuring. Real measurements vary. Identical numbers across lots are copied numbers.

How to spot a sheet that hides its test method

After 15 years of reading these sheets, I have a short list of tells. None of them are definitive on their own, but two or more together should make you ask harder questions.

First, look for the absence of a standard reference. A sheet that lists dimensions without citing ISO 8362-1, ISO 11040, GPI, or an equivalent is not necessarily bad, but it is not giving you a way to verify the method. A sheet that cites the standard and then lists values that are not in the standard should make you suspicious. I have seen sheets that say “ISO 8362-1” for a vial finish that does not match any finish defined in that standard. The citation was decoration.

Second, look for precision that exceeds the process capability. A sheet that lists “OD: 11.600 mm ± 0.005 mm” for a molded vial is claiming a tolerance that no molding process holds. The real tolerance for a molded vial is more like ±0.15 mm. The extra zeros are a sign that the number came from the CAD file, not from a measurement. A sheet that lists realistic tolerances is more credible because it reflects an actual process.

Third, look for missing test conditions on performance claims. “Reseals after 3 punctures” without needle gauge, speed, and temperature is not a test result. “Pressure rated to 150 psi” without temperature and safety factor is not a test result. “Temperature range -80°C to 250°C” without thermal shock conditions is not a test result. A sheet that gives you the test conditions for performance claims is a sheet written by someone who has actually run the test.

Fourth, look for the word “typical” on any value that affects your method. “Typical OD: 11.6 mm” means the supplier measured some vials at some point and got that number, but they are not promising that your lot will match. “Typical” is a legitimate word for values that vary and do not affect performance, like the exact color of the glass. It is a dangerous word for values that do affect performance, like OD, height, or reseal count. If a critical dimension is listed as “typical,” ask for the guaranteed range. If the supplier will not give one, the value is not controlled.

A decision checklist for reading any vial datasheet

Here is the checklist I use when evaluating a new vial supplier. It takes about 20 minutes per sheet and it has saved me more money than any other single practice.

Two field cases illustrate how this checklist catches problems before they reach the bench. The first was a pharmaceutical QC lab switching to a lower-cost 2 mL HPLC vial. The datasheet listed “12 x 32 mm, 9 mm screw finish.” The checklist question about finish depth revealed that the supplier’s thread depth was 0.10 mm shallower than the incumbent’s. The caps still engaged, but the septum compression was lower, and after 48 hours of storage at 4°C, the samples showed evaporation losses of 0.8% compared to 0.1% for the incumbent. The measurement that identified the problem was a thread profile scan on a CMM, not a go/no-go gauge. The fix was renegotiating the finish tolerance with the supplier or staying with the incumbent. The lab stayed with the incumbent.

The second case was an environmental lab running EPA SW-846 volatile organic methods in 40 mL VOA vials. The new supplier’s datasheet said “septum: 0.125 in PTFE/silicone, reseals after 2 punctures.” The lab’s method used a 22-gauge needle with a side port, and the reseal test had been run with a 25-gauge conical needle. The first sign of trouble was a 0.5% carryover on the method blank after the second puncture. The measurement that confirmed it was a leak test on punctured septa at 5 psi, which showed visible bubbling on 6 of 10 septa after 2 punctures with the 22-gauge needle. The fix was a septum with a thicker silicone core and a PTFE face of 0.5 mm instead of 0.25 mm, sourced from a supplier who could provide the reseal test method with the production needle gauge.

When you are comparing vials across suppliers, the datasheet is a starting point, not a decision document. The lab vials you buy will perform according to the actual dimensions and material properties of the lot that ships, not the numbers on the PDF. The suppliers who understand this are the ones who publish test methods and CoAs with measured values. The suppliers who do not are selling you a drawing. The difference shows up in your carryover, your sample loss, and your audit findings.

For the specific components, I have found that Aijiren brand vials and septa tend to come with more complete CoA documentation than generic imports, but the same checklist applies regardless of brand. A brand name is not a test method. Run the checklist on every new lot, every new supplier, and every new datasheet revision.

The decision rule is this: if a datasheet value affects your method and the supplier cannot tell you the test method, the sample size, and the acceptance limit, treat that value as nominal. If the value is nominal and it affects your method, you are accepting risk that belongs to the supplier. Push it back. The suppliers who can answer the test method question are the ones who will stand behind the product when the lot drifts. The ones who cannot are the ones who will point to the datasheet and say the product met the specification. The difference is whether the specification was ever real.

References & further reading

Related resources from our lab consumables network:

• Lab Vials at labvials.com
• Aijiren Brand at aijirenvials.com
• Voa Vial at voavial.com
• Septa at septacaps.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