Four polymer classes cover almost every chromatography contact surface

A septum that ran 96 injections on acetonitrile at 40 °C came back from the 97th run with a 0.9 mm core punched clean through it and a 3.4% RSD on the internal standard, up from 0.6% the week before. The part was a silicone-backed laminate rated for exactly that solvent. The problem was not the chemistry. It was the grade: the silicone face was a general-purpose 50 Shore A compound with a high volatile extractables load, and at 40 °C over three days the plasticizer migrated into the carrier gas path. Choosing which lab chromatography material belongs in which method is a materials problem before it is a catalog problem, and the solvent, pH, and temperature window decide it before the part number does.

This guide is about polymer selection against method conditions. It does not cover dimensions, fit, or pricing, which are settled elsewhere. What it covers is what each polymer class actually does when you put it in contact with a solvent at a temperature for a run length, and what failure you get when you get it wrong.

Almost every wetted part on a chromatography bench is one of four materials: PTFE (and its modified grades), silicone, EPDM, or a laminate built from two of them. Each has a solvent resistance envelope, a pH window, and a temperature ceiling. The envelope is not a single number. It is a combination of chemical compatibility, extractables, and mechanical creep, and the combination is what fails first.

The practical way to think about it: PTFE is the chemical inertness benchmark. Silicone is the flexibility and reseal benchmark. EPDM is the aqueous and steam benchmark. Laminate is the compromise that tries to get two of those at once, and compromises always have a seam.

What each class is actually good at

• PTFE: essentially universal chemical resistance, service range roughly -200 °C to +260 °C, pH 0 to 14. Virgin PTFE creeps under compression and has poor reseal memory. Modified PTFE (TFM) and filled grades (glass, carbon, graphite) improve creep at the cost of some inertness.
• Silicone: excellent reseal after puncture, service range roughly -60 °C to +200 °C, but swells badly in nonpolar solvents (toluene, hexane, dichloromethane) and has the highest volatile extractables of the four. Platinum-cured medical grade is far cleaner than general purpose.
• EPDM: strong against water, steam, dilute acids, and dilute bases, service range roughly -50 °C to +150 °C. It is a poor choice against aromatic and chlorinated hydrocarbons, which swell it and eventually dissolve the crosslink network.
• Laminate: typically a PTFE face bonded to a silicone core, or a silicone face on a PTFE barrier. The bond line is the weak point, and delamination is the failure mode nobody predicts because the data sheet quotes the face material, not the adhesive.

PTFE is the default until temperature or creep says otherwise

PTFE is where I start for any wetted path that sees a broad solvent menu, because it tolerates almost all of them. A PTFE-lined cap, a PTFE-faced septum, a PTFE filter membrane, a PTFE transfer line: all of these are the safe answer to “will this survive my gradient?” The catch is mechanical, not chemical.

Virgin PTFE has a high coefficient of thermal expansion (roughly 12 × 10⁻⁵ /°C, several times that of stainless steel) and it cold-flows under sustained compression. A PTFE ferrule torqued to spec at 25 °C and then run at 250 °C in a GC inlet will relax, and the seal will leak. This is why filled PTFE grades exist. Glass-filled PTFE reduces creep and holds torque, at the cost of slightly higher extractables and a lower upper temperature limit (typically +250 °C rather than +260 °C).

Where PTFE fails

PTFE does not fail by dissolving. It fails by creep, by permeation of small molecules (helium through thin PTFE is real, and it matters in a long transfer line), and by particulate shedding at a cut edge. A PTFE-faced septum that reseals poorly after 50 punctures will leak carrier gas into the inlet, and the symptom is a drifting retention time on early-eluting peaks that you will chase through the column before you think to look at the septum.

For filtration, the same rules apply. A PTFE membrane filter is the right choice for aggressive solvents and for gas and venting applications, and the wrong choice for aqueous mobile phases without a pre-wet in alcohol, because PTFE is hydrophobic and will not wet through on its own. That single pre-wet step is the most common PTFE filtration mistake I see on a bench.

Silicone reseals beautifully and extracts quietly into your baseline

Silicone is the material you choose when the seal has to survive repeated needle punctures at a temperature below 200 °C. Its reseal memory is the best of the four, which is why it dominates septum applications. The trade is extractables and solvent swelling.

General-purpose silicone carries plasticizers and low-molecular-weight siloxanes that migrate into the sample path. At elevated temperature, those compounds show up as a rising baseline in a GC run and as a cluster of siloxane peaks in a mass spec scan at m/z 73, 147, 207, 281. If you see those ions in a blank injection, your septum or your inlet liner O-ring is the source, not the column.

Platinum-cured versus general purpose

Platinum-cured silicone has a much lower volatile extractables load than peroxide-cured general purpose compound, and it is the grade to specify for trace analysis and for any method with a low detection limit. The cost difference is small relative to the cost of a failed blank. For a method targeting sub-ppb detection, I specify platinum-cured or I move to a laminate with a PTFE sample face.

Silicone’s other weakness is solvent swelling. Toluene and dichloromethane will swell a silicone septum by double digits over a run, which changes the puncture geometry and eventually tears the core. If your method uses those solvents, a silicone-only septum is the wrong part.

EPDM is the aqueous workhorse and the wrong answer for aromatics

EPDM (ethylene propylene diene monomer) is the material for water, dilute acid, dilute base, and steam. Its pH window runs roughly 2 to 12 and it holds up to +150 °C in continuous service. In an aqueous mobile phase method (ion chromatography, or an aqueous HPLC method at neutral pH), EPDM tubing and O-rings are a sensible choice and cheaper than PTFE for the same job.

The failure mode is aromatic and chlorinated solvents. EPDM in toluene swells, softens, and loses tensile strength, and the seal that was snug on day one is loose by day five. The symptom is a slow leak at a fitting that was torqued correctly, and the fix is not more torque, it is a different polymer.

Where EPDM earns its place

Peristaltic pump tubing, aqueous transfer lines, and any O-ring in a water-based system. It is also a good choice for a steam-sterilized or autoclaved component, where PTFE’s creep and silicone’s swelling both cause trouble. But put EPDM in a hexane/ethyl acetate gradient and you will be replacing tubing on a schedule instead of on condition.

Laminate is a compromise with a bond line you cannot see

A laminate gives you a PTFE face for chemical resistance and a silicone core for reseal, or the reverse. The data sheet quotes the face material properties, which is exactly what you want for the sample contact side. The failure mode lives at the bond line and in the core, and neither appears on the data sheet.

The common laminate for septa is a PTFE film bonded to a silicone disc. The PTFE face protects the sample from the silicone extractables, and the silicone provides the reseal. This works well until the adhesive fails at elevated temperature, at which point the PTFE film lifts, the needle punches through two layers instead of one, and the reseal is gone. I have seen this at 180 °C inlet temperature after about 40 injections on a laminate rated to 200 °C.

What to check on a laminate

• Bond-line temperature rating, not face-material rating. Ask for the adhesive spec.
• Core material and cure system (platinum-cured core extracts less).
• Face thickness. A very thin PTFE face is more prone to pinholing and lifting.
• Puncture count to reseal failure, measured at your method temperature, not at ambient.

If the vendor cannot give you the bond-line rating, treat the laminate as rated to the lower of the two component temperatures and derate further for thermal cycling.

Solvent, pH, and temperature decide the polymer before anything else does

The selection order is fixed: solvent first, then pH, then temperature, then mechanical duty (punctures, torque, cycling). Each step narrows the field. The table below is the decision matrix I use, with the numbers I actually specify.

Read the table as a filter, not a scorecard. A method with a toluene mobile phase at 60 °C eliminates silicone and EPDM immediately, leaving PTFE and laminate. A method with an aqueous buffer at pH 7 and 30 °C eliminates nothing, so you choose on mechanical duty and extractables instead.

The failure each wrong choice produces, and how to read it

Wrong polymer choices fail in recognizable ways. Learning the signatures saves you from replacing the wrong part twice.

• Silicone in an aromatic solvent: swelling, then a torn core, then a leak. Symptom is a widening injection-to-injection RSD and visible septum distortion on removal.
• EPDM in a chlorinated solvent: softening and a slow leak at a correctly torqued fitting. Symptom is a pressure drop you cannot trace to the pump.
• PTFE under sustained compression at high temperature: creep and a loss of torque. Symptom is a leak that appears only after thermal cycling, not at ambient.
• Laminate at the top of its temperature range: bond-line failure and PTFE face lift. Symptom is a sudden reseal loss after a run count well below the septum’s rated puncture life.
• General-purpose silicone in trace analysis: siloxane peaks at m/z 73/147/207/281 in the blank. Symptom is a baseline that rises with inlet temperature.

For method-specific context on how these failures interact with the inlet and autosampler, the vendor technical libraries are worth a read: Agilent’s GC consumables documentation and Restek’s technical resources both publish extractables and temperature data that go beyond the catalog blurb. Thermo Fisher’s chromatography pages carry the same kind of grade-level data for sample handling components.

Two field cases where the material, not the method, was the problem

Case 1: the siloxane blank that looked like a column bleed

A lab running a GC-MS method for semivolatile organics saw a rising baseline and a repeating siloxane cluster in every blank. The first instinct was column bleed, so they baked the column at 300 °C overnight. The cluster came back on the next blank. The measurement that identified it was a blank run with the inlet at 250 °C versus 200 °C: the siloxane signal scaled with inlet temperature, which pointed at the inlet, not the column. The septum was a general-purpose silicone disc. The fix was a platinum-cured septum and a fresh inlet liner O-ring, and the blank went clean. The lesson: siloxane ions that track inlet temperature are a septum problem, and the cure is a cleaner silicone grade, not a hotter bake-out.

Case 2: the laminate that lifted at 180 °C

A lab running high-temperature GC on a PTFE/silicone laminate septum rated to 200 °C started losing reseal after about 40 injections. The RSD on the internal standard jumped from 0.8% to 2.9%, and on removal the PTFE face was visibly lifted at the puncture site. The measurement that identified it was a puncture-count-to-failure test at the actual inlet temperature: the septum failed at 40 punctures at 180 °C but held past 100 at ambient. The bond line, not the face material, was the limit. The fix was a septum with a higher-temperature adhesive and a shorter replacement interval set from the measured puncture count, not the catalog number. That is the same logic behind setting replacement intervals from measured life rather than rated life, which applies across the consumables you buy.

A selection checklist you can run before the next method transfer

Before you transfer a method to a new instrument or a new lab, run this checklist against every wetted polymer in the flow path. It takes ten minutes and catches the failures above.

• List every wetted polymer: septum, liner O-ring, cap liner, filter membrane, transfer tubing, vial septum, pump tubing. Most methods have five to eight.
• Write the solvent menu, including the strongest solvent in the gradient, not the average.
• Write the pH range across the run, including any wash steps.
• Write the maximum temperature the part sees, including bake-out and cleaning cycles.
• Check each polymer against the table above and flag any part within 20 °C of its ceiling or in a “poor” solvent cell.
• Confirm the grade: platinum-cured versus general purpose silicone, virgin versus filled PTFE, and the laminate bond-line rating.
• Set the replacement interval from measured puncture count or measured extractables, not the catalog rating.
• Run a blank at the method temperature and check for siloxane ions or a rising baseline before you trust the method.

Two supporting details worth having on hand: a PTFE-lined cap on a COD vial is the right call for aggressive digestion chemistry, and a HPLC filter with the correct membrane chemistry is the difference between a clean injection and a clogged frit. Both are material decisions before they are part-number decisions. For the broader consumables context, the lab chromatography resource pages cover the rest of the bench.

One note on standards: where a method is regulated, the polymer choice still has to survive the validation. EPA SW-846 methods, for example, specify performance but leave material selection to the analyst, so the extractables and temperature data you collect become part of your method documentation. The SW-846 test method library is the reference for that, and the material justification belongs in your validation file.

The decision rule: match the polymer to the worst condition in the run, then derate

Pick the polymer that survives the worst solvent, the widest pH, and the highest temperature the part sees in a single run, then derate the temperature rating by 20 °C and the puncture life by 30% for real-world cycling. If no single polymer survives all three, split the flow path: PTFE for the sample contact side, silicone or EPDM for the mechanical seal, and a laminate only when you can get the bond-line rating in writing. The wrong grade does not announce itself. It shows up as a drifting RSD, a rising blank, or a leak that appears only after thermal cycling, and by then you have usually replaced the column first.

References & further reading

Related resources from our lab consumables network:

• Lab Chromatography at labchromatography.com
• Hplc Filter at hplcfilters.com
• Ptfe Membrane Filter at ptfemembranefilter.com
• Cod Vial at codvial.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