Which materials bleed which compounds, and at what temperature

A 40 °C GC headspace run on a 20 mL crimp vial came back with a siloxane peak at retention time 11.3 min that had not been in the method’s blank for the previous six months. The peak area was 340 pA·s, RSD across six replicates was 22%, and it appeared only when the sample tray sat on the instrument for more than four hours before injection. The analytes were fine. The oem vial ghost peaks, extractables and contamination were coming from the closure, not the sample, and finding that out took three blank injections and one placebo run, not a new column.

Ghost peaks from vialware are the most misdiagnosed failure in trace analysis. I have watched labs replace a $2,400 column, re-passivate an inlet, and swap a detector before anyone thought to inject air into an empty vial. The vial and its closure are the last thing the sample touches and the first thing everyone forgets. This article is about the specific compounds that bleed from specific materials at specific temperatures, and the blank and placebo sequence that isolates the source in under an hour.

Every vial component is a formulation, not a pure material. The polymer, the plasticizer, the antioxidant package, the mold release, and the lubricant on the septum all have something to give up under heat. The question is never whether a component extracts. It is which compound, at what concentration, and above what temperature.

Septum silicone and its siloxane ladder

Silicone septa are polydimethylsiloxane, and PDMS depolymerizes. The cyclic oligomers, principally hexamethylcyclotrisiloxane (D3), octamethylcyclotetrasiloxane (D4), decamethylcyclopentasiloxane (D5), and dodecamethylcyclohexasiloxane (D6), elute as a regular ladder in GC-MS at m/z 73, 207, 281, and 355. On a nonpolar column they land roughly every 60 to 90 retention indices apart. If you see that ladder, stop looking at the sample. D4 through D6 dominate, and they show up in blank headspace runs at 60 °C within 30 minutes. At 40 °C they are still there, just smaller. Above 80 °C the bleed climbs steeply.

The grade of silicone decides how much. A standard gray bromobutyl-faced septum has a silicone coating on the sample side that is measured in micrometers; a pure silicone septum is solid PDMS. The solid one bleeds more, and it bleeds longer. For headspace above 80 °C, a PTFE-faced butyl septum with the PTFE toward the sample is the only defensible choice, and even then the PTFE is a barrier, not a seal against everything.

Polypropylene and the antioxidant package

Vial bodies and caps are polypropylene. The polymer itself is fairly clean, but the stabilization package is not. Hindered phenolic antioxidants such as Irganox 1010 and phosphite co-stabilizers show up in extracts as 2,4-di-tert-butylphenol (m/z 191, 206), BHT, and various phosphate esters. These appear in GC-MS as a cluster of peaks between retention indices 1500 and 1900, often with a strong m/z 191 fragment that looks like a terpene and fools people into chasing a natural product. They extract more readily at elevated autosampler temperatures and after long residence times. A vial body sitting at 40 °C for eight hours in a tray will show more of this than one injected immediately.

Mold release, slip agents, and the erucamide problem

Erucamide and oleamide are slip agents and mold releases. They migrate to the surface of polypropylene and polyethylene during molding and sit there waiting to be dissolved. In LC-MS with electrospray, erucamide gives a strong [M+H]+ at m/z 338.3 and a characteristic ammonium adduct at m/z 355.3. If you run positive-mode LC-MS and see a persistent peak at m/z 338 with a shoulder 17 mass units up, you are looking at the bag the vial shipped in, the cap, or the liner, not your analyte. This one has burned more method development time than any siloxane.

What the temperature actually does

Extraction is Arrhenius-driven, and the practical rule is that every 10 °C of vial residence temperature roughly doubles the extractable concentration reaching the headspace or the solvent. A septum that gives a clean blank at 25 °C on the bench can fail at 40 °C in a tray after four hours. The temperature rating on a closure datasheet is a service limit for the polymer, not a guarantee of analytical cleanliness. A butyl septum rated to 100 °C will not melt at 100 °C. It will still bleed siloxanes and sulfur compounds at that temperature.

The blank and placebo sequence that isolates the source

Do not troubleshoot ghost peaks by swapping parts and re-running the sample. Run a controlled sequence that changes one variable at a time and compare peak areas, not peak presence. The sequence below takes about 45 minutes on a headspace GC and identifies the source in most cases.

• Blank 1, system blank: no vial, empty headspace loop or direct injection of carrier. Establishes instrument and column background. If the peak appears here, the vial is innocent and you should stop.
• Blank 2, empty vial, same lot: the suspect vial and closure, sealed, no sample. This is the vial blank. Any peak here is extractable from the vialware.
• Blank 3, empty vial, different lot: same part number, different manufacturing lot. If Blank 3 is clean and Blank 2 is not, you have a lot problem, not a design problem.
• Placebo: the sample matrix without analyte, prepared in the suspect vial exactly as the method specifies, including any derivatization, heating, or hold time. This catches matrix-catalyzed extraction that a dry empty vial misses.
• Placebo in a reference vial: the same matrix in a vial you trust, ideally a different material or a certified low-bleed closure. The delta between the two placebos is your vial contribution.

Compare peak area ratios, not just presence. A siloxane peak that is 340 pA·s in the suspect vial and 12 pA·s in the reference vial is a vial problem. A peak that is 340 in both is your matrix or your inlet. Run each blank in triplicate so you can report an RSD. If the vial blank RSD exceeds 15%, the closure is not uniform, which itself is a finding.

Reading the mass spectrum before you blame the vial

Take the ghost peak to MS before you change anything. The fragmentation pattern tells you the material family in one injection.

If your peak is not in this table, it is probably not the vial. Silicone, antioxidant, and slip agent cover the overwhelming majority of vial-origin ghosts. A peak with a clean molecular ion and a plausible fragmentation for your analyte is your analyte, and you have a different problem.

Field case: the siloxane ladder that only appeared after four hours

A contract lab running EPA method 8260 by headspace on a 20 mL crimp vial started seeing D4 and D5 at 60 °C in method blanks, but only on the afternoon tray. Morning runs were clean. The measurement that identified it: they ran an empty sealed vial from the same lot at 60 °C with a 5-minute hold and again with a 60-minute hold. The 5-minute hold gave D4 at 40 pA·s. The 60-minute hold gave 380 pA·s. The septum was a standard gray butyl with a thin silicone film, and the film was being cooked by the tray heater, which was set to 45 °C and had drifted. The fix was two-part: they switched to a PTFE-faced butyl septum with the PTFE toward the sample, and they brought the tray heater back to 30 °C and re-validated the hold time. D4 dropped to under 20 pA·s in the 60-minute blank. This is the failure mode that vial closures selection actually controls, and it is a material choice, not a torque problem.

Field case: the m/z 338 peak that was never the sample

An environmental lab running positive-mode LC-MS/MS on a pesticide panel had a persistent peak at m/z 338.3 in every injection, including solvent blanks, with an intensity that varied by a factor of three between vials. They had already replaced the column. The measurement that identified it: they injected the same solvent from the same bottle in three vial types (their standard vial, a glass vial with a PTFE-lined silicone cap, and a glass vial with a polyethylene cap), and the m/z 338 peak tracked the polyethylene cap. The cap was molded with an erucamide slip agent, and the solvent was dissolving it off the cap surface during the 12-hour autosampler run. The fix was to move to a cap without an amide slip agent and to reduce the autosampler residence time by splitting the tray into two shorter runs. The peak dropped below the reporting limit. The lesson is that the cap is part of the sample path, and the chromatography vial as a system includes the closure surface the solvent touches.

How to specify a low-extractable vial without paying for a myth

There is no such thing as a zero-extractable vial. There is a vial whose extractables sit below your method’s reporting limit under your method’s temperature and residence time. Specify to that, and ask for the data that proves it.

• Closure face material: for headspace above 60 °C, specify PTFE-faced butyl, PTFE toward the sample, minimum 0.010 in (0.25 mm) PTFE film. For ambient LC-MS, a PTFE/silicone with a PTFE face is adequate if the solvent does not contact the silicone edge.
• Body resin: specify virgin polypropylene, no regrind, and ask for the antioxidant package. A supplier who cannot name the stabilizer cannot control it.
• Slip agent: ask explicitly whether the cap or liner contains erucamide or oleamide. If the answer is yes, ask for a low-slip alternative for LC-MS work.
• Certification data: request a lot-specific extractables report, not a generic statement. The report should name the test solvent, the temperature, the hold time, and the detection limit. A report that says “passes” without a number is decorative.
• Dimensional and material traceability: the OEM vial you buy should arrive with a certificate of analysis that ties the lot to the resin lot. This is what lets you defend a lot-to-lot shift to an auditor.

For method development, run your extractables screen through a contract lab that can do both GC-MS and LC-MS on the same vial lot, so you see the siloxane ladder and the erucamide peak in one report. The analytical testing group you use should be able to give you a semi-quantitative extractables profile at your method temperature, not a room-temperature soak.

Operational precautions that keep vial extractables out of the run

Most ghost-peak incidents are not a bad lot. They are a good lot used outside its envelope. The controls below cost almost nothing and remove the majority of vial-origin contamination.

• Cap the tray temperature at 30 °C unless the method requires more. Every 10 °C above ambient roughly doubles extractable concentration. Verify the tray setpoint with an external thermometer; I have seen 15 °C of drift on a heater that read correctly on its own display.
• Limit autosampler residence time to under four hours. If the run is longer, split the tray. A vial that is clean at injection can be contaminated by hour six.
• Do not store prepared vials over a weekend in a tray at temperature. The extractables do not care that the instrument is idle.
• Run a vial blank at the start and end of every tray. Two injections per tray, and you have a control chart for vial cleanliness. If the end-of-tray blank is three times the start-of-tray blank, you have a residence-time problem, not a lot problem.
• Keep the septum face dry before crimping. Solvent on the septum face during crimping wicks extractables into the sample. Wipe the crimp area, not the septum.
• Crimp to the closure’s specified compression, not to a feel. For a 20 mm crimp, that is typically 1.5 to 2.0 mm of compression on the seal, set by the crimper jaw, verified with a crimp gauge. Over-crimping extrudes the septum and increases the surface area exposed to the headspace.

For headspace methods, the U.S. EPA SW-846 series and the instrument vendor’s application notes are where you find the method-specific temperature and hold-time envelope. Use them to set the tray temperature, then verify with your own vial blank. The Restek technical library and the Agilent GC consumables documentation both publish extractables data by closure type that is worth reading before you write the specification. For LC-MS work, the Thermo Fisher chromatography and sample handling resources include guidance on slip-agent and plasticizer interference that maps directly to the m/z 338 problem.

Troubleshooting checklist: ghost peak, unknown origin

Work top to bottom. Each step changes one variable and gives you a number to compare.

1. Inject a system blank with no vial. Peak present? The vial is not the source. Stop and look at the inlet, column, and carrier.
2. Inject an empty sealed vial from the suspect lot. Peak present? Go to step 3. Peak absent? The vial is clean dry; go to step 5 (placebo).
3. Inject an empty sealed vial from a different lot of the same part. Peak absent in the second lot? You have a lot-specific extractable. Quarantine the lot and request the supplier’s extractables report.
4. Take the ghost peak to MS. Match the ions to the table above. Silicone ladder means closure face material. m/z 338 means slip agent. m/z 191 means antioxidant.
5. Run the placebo in the suspect vial and in a reference vial. Delta in peak area is the vial contribution. If the delta is under 10% of the peak, the vial is not your main problem.
6. Vary hold time at temperature. If the peak grows linearly with residence time, it is an extraction kinetics problem. Reduce tray temperature or split the run.
7. Vary crimp compression. If the peak grows with compression, the septum is being extruded and the exposed area is increasing. Re-set the crimper to 1.5 to 2.0 mm and re-test.

If steps 1 through 7 do not isolate it, the peak is not from the vialware, and you should look at the sample preparation reagents or the autosampler wash solvent. A wash solvent that sits in a plastic reservoir for a week will pick up plasticizers and carry them onto the needle.

Decision rule

When a ghost peak appears, run the system blank and the empty-vial blank before you touch the instrument, and take the peak to MS before you change a part. If the peak matches the siloxane ladder, change the closure face material to PTFE-faced butyl and drop the tray temperature to 30 °C. If it matches m/z 338, the cap has a slip agent and the solvent is dissolving it. If it matches m/z 191, the body resin’s antioxidant package is extracting and you need a lower-extractable resin specification. Only after those three checks come back clean do you spend money on the column.

References & further reading

Related resources from our lab consumables network:

• Oem Vial at oemsamplevials.com
• Vial Closures at vialclosures.com
• Analytical Testing at labanalyticaltesting.com
• Chromatography Vial at chromvials.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