Syringe Filters for HPLC and GC: PTFE vs Nylon vs PVDF, and 0.22 vs 0.45 micron

A syringe filter looks like the simplest consumable in the lab, which is exactly why it causes so many hidden problems. Choose the wrong membrane, and it can absorb your analyte, leach contaminants into the sample, or clog halfway through the first injection. Choose well, and you protect your column, your injector, and your results. Here is how to pick the right one.

Why filter samples at all?

Particles are the main cause of blocked frits, rising backpressure, and shortened column life. Filtering samples before injection is cheap insurance, and it works best alongside a guard column and properly filtered mobile phases. It also protects autosampler needles and valves, which are costly to repair.

Step 1: Match the membrane to your solvent

• PTFE: chemically resistant and compatible with aggressive organic solvents, acids, and bases. It is naturally hydrophobic, so for aqueous samples it needs to be wetted with an organic solvent first. A common choice for organic extracts and GC samples.
• Nylon: hydrophilic and works well for aqueous and many organic samples. It can bind some analytes, so test recovery for sensitive compounds.
• PVDF: low protein binding and good general-purpose compatibility, widely used for aqueous and mixed solvents in bioanalysis.
• PES (polyethersulfone): low binding and fast flow, a good choice for aqueous, biological, and dilute samples.
• Regenerated cellulose: hydrophilic with low binding, suited to aqueous and some polar organic solvents, but check solvent compatibility carefully.

Always check the manufacturer's chemical compatibility chart for your exact solvent mix, since membrane and housing material both matter.

Step 2: Pick the pore size

• 0.45 micron: the standard choice for conventional HPLC columns with 3-5 micron particles.
• 0.22 micron: recommended for UHPLC and sub-2 micron columns, and for samples where you want to remove fine particulates or bacteria.

Finer filters clog faster and need more pressure. If your samples are dirty, a glass-fibre prefilter layer helps the main membrane last longer.

Step 3: Choose the diameter

• 4 mm and 13 mm: for small volumes, typically up to a few millilitres, ideal for autosampler vial preparation.
• 25 mm and 33 mm: for larger volumes or particulate-heavy samples.

A filter that is too small for your sample will clog and tempt you to push harder, which can burst the housing. Match size to volume.

Step 4: Watch for extractables and sample loss

Some membranes and housings release trace compounds into the sample, and some adsorb analytes. For trace work, discard the first few drops of filtrate, filter a solvent blank to check for ghost peaks, and test recovery against an unfiltered standard. Unexpected extra peaks may come from the filter, so also see our guide to HPLC ghost peaks.

Three quick scenarios

Routine HPLC assay of a tablet extract in methanol and water.

Use a 13 mm, 0.45 micron nylon or PVDF filter and discard the first few drops. If recovery of the active ingredient drops, switch to PVDF or PES and repeat the check.

UHPLC analysis of a diluted plasma sample.

Use a 0.22-micron low-binding filter such as PES or PVDF. Protein precipitation followed by centrifugation first will stop the filter clogging halfway through the batch.

GC analysis of a hexane or dichloromethane extract.

Use a 0.45-micron PTFE filter in a size that matches the volume, and confirm that both the membrane and the housing tolerate the solvent.

In every case, the final check is the same:

Run a filtered blank and compare it with an unfiltered one. If the two chromatograms match, the filter is doing its job without adding anything.

Quick selection guide

• Aqueous samples, general use: PVDF, PES or nylon, 0.45 micron
• UHPLC, aqueous: PVDF or PES, 0.22 micron
• Strong organic solvents, acids, bases: PTFE
• GC samples in organic solvents: PTFE
• Digested samples for ICP: see microwave digestion for ICP-OES and ICP-MS

Common syringe filter mistakes and how to avoid them

Filtration looks trivial, but these errors turn up in method troubleshooting again and again.

• Using nylon for an analyte that binds to it. Compare recovery from a filtered and an unfiltered standard when you validate a method, particularly for low-concentration or protein-related compounds.
• Pushing aqueous samples through dry PTFE. Water will not wet a hydrophobic membrane, so the syringe builds pressure, and nothing comes out. Pre-wet the filter with a little methanol or choose a hydrophilic membrane.
• Forcing a clogged filter. Heavy pressure can rupture the membrane or the housing, and it lets particles through. If the filter is hard to push, use a larger diameter, a prefilter, or centrifuge the sample first.
• Reusing filters. A used filter can release material it has already trapped and carry analytes between samples. Use one filter per sample.
• Ignoring the first drops. Some filters shed trace extractables at the start. For trace analysis, discard the first few drops and run a filtered solvent blank.
• Filtering volatile samples carelessly. Evaporation during filtering changes concentration, so work quickly and cap vials straight away.
• Skipping a centrifuge step for very dirty samples. Spinning the sample first removes bulk solids so the filter can do its real job. See our guide on how to choose the right centrifuge speed.

Frequently asked questions

What is the difference between 0.22 and 0.45 micron syringe filters?

The number is the pore size. A 0.45 micron filter is the usual choice for standard HPLC with 3-5 micron particle columns. A 0.22 micron filter removes finer particles and is recommended for UHPLC and sub-2 micron columns, but it clogs faster.

Is PTFE or nylon better for HPLC samples?

It depends on the sample. PTFE is chemically resistant and suits strong organic solvents, acids, and bases. Nylon works well for aqueous and many organic samples but can bind some analytes. Check compatibility and test recovery for your compounds.

Do I need to filter every sample before HPLC injection?

If the sample may contain particles, yes. Clean standards in filtered solvent may not need it, but real samples such as extracts, formulations, and digests almost always do. Filtering protects the column frit, injector, and autosampler.

Can the syringe filter contaminate my sample?

It can. Some membranes and housings release trace extractables, which may appear as extra peaks. Discard the first drops and run a filtered solvent blank to check, particularly in trace analysis. See HPLC ghost peaks for related causes.