A 47 mm PTFE filter that had run 200 acetonitrile gradients without a problem dissolved at the seam on the third injection of a 60% tetrahydrofuran method. The membrane was fine. The polyester-backed laminate holding it to the support ring was not. The bubble point test on the failed lot read 8 psi against a specification of 22 psi minimum, and the seam had opened within 90 minutes of contact with the THF. That is what happens when you pick the polymer by habit instead of by solvent and temperature class. Choosing which PTFE membrane filter material to use for a given solvent and temperature is a decision you can make on paper in about four minutes, with a solvent compatibility chart, a thermometer reading, and one question about whether the method is aqueous or organic. This guide is about that decision: PTFE, silicone, EPDM, and laminate construction, what each one survives, and the specific failure each wrong choice produces.
Most laboratories treat “PTFE membrane filter” as a single product. It is not. The membrane itself is usually unsintered or lightly sintered polytetrafluoroethylene, but the device you load into a filter holder or a syringe is a composite: membrane, support backing, and in the case of capsule and inline units, a housing and a seal. The polymer you select is the one in contact with your sample, and the failure you get when you pick wrong is specific to the polymer class.
PTFE: chemically inert, mechanically soft
PTFE is the reference material for chemical resistance. It is unaffected by concentrated nitric and hydrochloric acid, by ketones, esters, aromatic hydrocarbons, and by nearly every HPLC mobile phase in routine use. Continuous service temperature runs to about 260 °C, with short excursions higher. It has essentially no pH limit; you can run pH 0 to 14 without swelling, hydrolysis, or leaching of polymer fragments into the eluate.
The failure mode for PTFE is not chemical, it is mechanical. PTFE creeps under compression. A capsule housing torqued to 25 in-lb and left in a heated autoclave cycle will relax its seal and leak at the ferrule. PTFE also cold-flows at the edges of a punch-cut disc, which is why coring at the housing seat is a real failure in inline units. And unsupported PTFE membranes are fragile: a 0.2 µm disc will burst at a differential pressure around 15 to 20 psi if you push aqueous solvent through it too fast, because water does not wet PTFE. That is the single most common PTFE failure I see, and it looks like a chemical failure until you notice the burst disc is intact and the membrane is torn radially.
Silicone: flexible, permeable, and a poor choice for organics
Silicone (polydimethylsiloxane) is the go-to when you need compliance and elasticity: septum material, gasket material, and low-pressure inline seals. It is not a filtration membrane in the analytical sense. Silicone swells in toluene, hexane, dichloromethane, and most nonpolar organics, sometimes by 40 to 80% by volume. It is highly permeable to gases, which makes it useless as a barrier in any degassed or headspace application. Its temperature window is wide (-60 to 200 °C) and its pH range is roughly 3 to 11; outside that, silicone hydrolyzes slowly and sheds low-molecular-weight siloxanes into the sample.
The failure you get: a silicone septum on a 2ml vial used for a hexane extraction will swell, extrude into the crimp, and produce a ghost peak at retention time 4.2 minutes on every subsequent injection. The peak is siloxane, and it will not go away until you replace the septum and recondition the inlet.
EPDM: good for aqueous, poor for aromatic
Ethylene propylene diene monomer is the workhorse seal material for aqueous and mild-acid service. It handles pH 1 to 13, temperatures to about 150 °C in continuous service, and it is far more resistant to compression set than silicone. It is the right choice for water, dilute buffers, and dilute caustic. It is the wrong choice for aromatic hydrocarbons, chlorinated solvents, and ketones. EPDM in toluene swells by 100 to 150% and loses tensile strength within hours.
In a filter housing, EPDM seals fail by extrusion: the swollen O-ring is pushed through the gap at the housing shoulder under pressure, and the leak appears at the outlet fitting, not at the seal. You will chase the fitting torque for an hour before you realize the seal has grown.
Laminate: the backing decides the fate
A laminated PTFE membrane is PTFE bonded to a support (typically polypropylene, polyester, or PTFE nonwoven). The laminate gives you mechanical strength and a defined pore structure that an unsupported membrane cannot hold. But the laminate is only as chemically resistant as its weakest layer. Polyester backing is fine against aqueous and mild organic solvents; it hydrolyzes in strong base and dissolves in tetrahydrofuran, dimethylformamide, and dimethyl sulfoxide. Polypropylene backing handles most organics and pH 1 to 13 but is attacked by strong oxidizers and chlorinated solvents at elevated temperature.
The failure mode for a laminate is delamination: the backing separates from the membrane, and you get a filter that passes particulate and a bubble point that drops to near zero. That is exactly what happened in the THF case above.
Solvent and temperature windows, side by side
The table below is the one I keep taped inside the filter cabinet. It is not exhaustive, but it covers the solvent classes and temperature ranges that account for most of the failures I have seen in 15 years of running a filter line.
| Polymer | pH window | Continuous temp | Survives | Fails against | Failure signature |
|---|---|---|---|---|---|
| PTFE (unsupported) | 0 to 14 | 260 °C | Acids, bases, ketones, aromatics, THF, DMF, DMSO | Mechanical burst below 15 psi wetting pressure; creep under compression | Radial tear, seal relaxation, leak at ferrule |
| PTFE on polypropylene backing | 1 to 13 | 120 °C | Most HPLC mobile phases, dilute buffers, alcohols | Strong oxidizers, chlorinated solvents above 60 °C | Backing embrittlement, pore collapse |
| PTFE on polyester backing | 2 to 12 | 100 °C | Aqueous, mild organic, acetonitrile, methanol | THF, DMF, DMSO, strong base above pH 10 | Delamination, bubble point collapse |
| Silicone | 3 to 11 | 200 °C | Water, alcohols, dilute acids | Toluene, hexane, DCM, gas permeation | Swelling 40 to 80%, siloxane ghost peaks |
| EPDM | 1 to 13 | 150 °C | Water, buffers, dilute caustic, mild acid | Aromatics, chlorinated solvents, ketones | Swelling 100 to 150%, extrusion at shoulder |
Read the table by column, not by row. Start with the solvent. If it is THF, DMF, DMSO, or a chlorinated solvent above 60 °C, you need unsupported PTFE or a PTFE-on-PTFE laminate, and nothing else on this list will do. If it is aqueous buffer, EPDM seals with a PTFE-on-polypropylene membrane are the sensible default. If it is a hexane extraction for a COD vial, silicone is disqualified by swelling before temperature ever enters the conversation.
How temperature changes the answer, not just the rating
Temperature ratings are usually quoted as a single continuous number, and that number is misleading. What matters is the combination of temperature and solvent, because heat accelerates every chemical attack and every mechanical relaxation.
Take an EPDM seal rated to 150 °C. That rating assumes aqueous service. Run the same seal in toluene at 80 °C and it will swell and extrude in under four hours, because the solvent attack and the thermal softening compound. The temperature rating is not a property of the polymer in isolation; it is a property of the polymer in a specific chemical environment.
PTFE is the exception, and this is why it dominates regulated methods. Its temperature window is wide enough that solvent choice rarely changes the answer. But PTFE’s mechanical weakness gets worse with temperature: creep accelerates above 150 °C, and a capsule that held 25 in-lb of torque at room temperature will lose about 30% of its sealing force after 24 hours at 200 °C. If you are running high-temperature filtration, plan on retorquing or on a graphite-filled PTFE gasket, which cuts creep by roughly half.
The practical rule: if your method runs above 60 °C in an organic solvent, treat every non-PTFE polymer as suspect and verify with a 24-hour soak test at method temperature before you commit a production run.
What the wrong choice actually looks like on the bench
Two cases from the last few years, both of which cost real money before anyone looked at the filter.
Case 1: the THF delamination
A contract lab running a size-exclusion method switched from THF to a THF/water mobile phase and started seeing a baseline rise and a loss of resolution after about 90 minutes of run time. The symptom looked like column degradation. The column was replaced twice. The actual measurement that identified the problem was a bubble point test on the inline filter after the run: 8 psi against a 22 psi specification. The filter was a PTFE-on-polyester laminate, and the polyester backing had hydrolyzed in the THF. The fix was a switch to unsupported PTFE in a stainless housing, and the baseline rise disappeared on the next injection. The lesson is that a laminate is only as good as its backing, and the backing is not always on the specification sheet.
Case 2: the EPDM extrusion
A water-testing lab running EPA method 300.1 through a capsule filter started getting intermittent leaks at the outlet fitting. The technician retorqued the fitting three times and replaced the ferrule twice. The leak persisted. The measurement that found it was a dimensional check on the O-ring after a week of service: it had grown from 1.78 mm cross-section to 2.4 mm, a 35% increase, and had extruded into the housing gap. The cause was a switch to a buffer containing 5% methanol at pH 9, which is within EPDM’s pH window but outside its methanol tolerance. The fix was a PTFE-encapsulated EPDM seal, which keeps the EPDM compliance but puts a PTFE barrier between the elastomer and the solvent. The leak stopped and stayed stopped.
Both cases share a pattern: the failure looked like hardware, and the root cause was polymer selection. If you are troubleshooting a leak or a baseline problem and you have already checked torque and fittings, the next question is what the seal and the backing are made of.
A decision checklist you can run before the method is written
This is the sequence I use when a method is being developed or transferred. It takes about four minutes and it catches most of the wrong-polymer choices before they reach the bench.
- Step 1: identify the solvent class. Aqueous, alcohol, acetonitrile, ketone, aromatic, chlorinated, or polar aprotic (THF, DMF, DMSO). Polar aprotic and chlorinated solvents are the ones that eliminate most polymers immediately.
- Step 2: check the pH. Below 2 or above 12, PTFE is the only safe membrane and EPDM is out for seals. Between 3 and 11, you have options.
- Step 3: check the method temperature. Above 60 °C in an organic solvent, assume non-PTFE polymers will fail and verify with a soak test.
- Step 4: identify the backing. Polyester is fine for aqueous and acetonitrile; it is not fine for THF, DMF, DMSO, or strong base. Polypropylene is a better general-purpose backing but still loses to strong oxidizers.
- Step 5: check the wetting requirement. Aqueous filtration through unsupported PTFE needs a pre-wet with alcohol (typically methanol or isopropanol) or a low-pressure wetting step, or the membrane will not pass water at any reasonable pressure. If your method cannot tolerate a pre-wet, you need a hydrophilic PTFE or a different membrane entirely.
- Step 6: verify with a bubble point and a 24-hour soak. Run the actual method solvent through a coupon at method temperature for 24 hours, then measure bubble point against specification. A drop of more than 20% is a fail.
If you are running a regulated method, the acceptance criteria for the filter are usually written into the method or into the validation package. The EPA SW-846 series, for example, specifies filtration steps but leaves the filter material to the analyst in many methods, which means the selection is yours to defend. The U.S. EPA SW-846 test methods are the reference to check before you commit a material to a method that will be audited. For chromatography-specific guidance on solvent compatibility and filter selection, the technical resources at Restek and the sample-handling documentation from Thermo Fisher are worth keeping on the shelf, and the consumables and column compatibility tables from Agilent are useful when you are matching a filter to a specific column chemistry.
Where the specification sheet stops and the bench starts
Most filter data sheets list the membrane polymer and a temperature rating. Very few list the backing polymer, the seal material, or the compression set of the housing. Those are the three things that fail first. When you are evaluating a supplier, ask for the backing material by name, the seal elastomer by name, and a compression set figure (ASTM D395, typically 25% max at 70 hours and 100 °C for a good EPDM). If the supplier cannot tell you the backing, assume polyester and treat the filter as incompatible with polar aprotic solvents until proven otherwise.
One more thing worth checking: the membrane’s own pore structure. A PTFE membrane that has been stretched to a high ratio will have a higher bubble point but a lower burst pressure, and the two trade off against each other. For high-pressure inline service, a lower-stretch membrane with a bubble point around 18 to 22 psi in isopropanol is more robust than a high-stretch membrane at 30 psi. The high-stretch membrane will filter finer, but it will also tear at the housing seat under a pressure spike. Match the membrane to the pressure, not just to the pore size.
If you are choosing a filter for a new method and you want a starting point, the PTFE membrane filter range at ptfemembranefilter.com is organized by backing and seal material, which makes the compatibility check faster. For methods that go through an autosampler, the vial and insert you pair with the filter matter too: a vial insert with a PTFE-lined septum keeps the seal chemistry consistent with the filter chemistry, and that consistency is what keeps a method from developing a ghost peak six months after validation.
The decision rule is simple: start with the solvent, not the pore size. If the solvent is polar aprotic or chlorinated, use unsupported PTFE and a PTFE or graphite-filled seal, and do not accept a laminate without a backing specification. If the solvent is aqueous or a simple alcohol, a PTFE-on-polypropylene membrane with EPDM seals is the efficient choice, provided the pH stays between 1 and 13 and the temperature stays below 120 °C. If the solvent is a nonpolar hydrocarbon, silicone and EPDM are both disqualified by swelling, and you are back to PTFE. Verify the choice with a 24-hour soak and a bubble point test before the method is validated, because the failure you prevent on paper is the one that does not cost you a column and a week of rework.
References & further reading
Related resources from our lab consumables network:
- Ptfe Membrane Filter at ptfemembranefilter.com
- Cod Vial at codvial.com
- Vial Insert at hplcinsert.com
- 2Ml Vial at 2mlsamplevial.com
External references and standards:
