A 0.22 µm filter removes organisms from a peptide solution and removes some of the peptide with them, and at low concentration the loss can be large enough to matter. The mechanism is adsorption to the membrane rather than exclusion by pore size — a peptide is far smaller than the pores it passes through. It sticks to the material the filter is made of, and how much sticks depends on which material that is.
Why a filter removes something it cannot exclude
A 0.22 micron pore is enormous relative to a peptide. A fifteen-residue chain is a few nanometres across, so nothing about the geometry stops it. What stops some of it is surface: a filter is a high-surface-area membrane, and peptides adsorb to surfaces through the same hydrophobic and electrostatic interactions that take material onto tube walls and pipette tips.
The difference is scale. A microcentrifuge tube presents a few square centimetres of wall; a syringe filter presents a membrane with a vastly larger effective area packed into a small housing. That is what makes filtration one of the more expensive single steps a dilute solution passes through.
| Membrane | Protein and peptide binding | Notes |
|---|---|---|
| PES (polyethersulfone) | Low | The usual first choice for peptide work |
| PVDF | Low to moderate | Widely available; low-binding grades exist |
| Cellulose acetate | Low | Good for aqueous solutions; poor solvent tolerance |
| Nylon | High | Binds peptides and proteins readily; avoid |
| PTFE | Low for solvents | Hydrophobic; needs wetting for aqueous use |
How much peptide does a filter actually take?
It depends almost entirely on concentration, and the relationship is not proportional. Adsorption fills a roughly fixed number of binding sites on the membrane, so the mass lost is similar whether the solution is concentrated or dilute — which means the fraction lost rises sharply as concentration falls.
At milligram-per-millilitre concentrations the loss is usually a rounding error. At microgram-per-millilitre it can be a visible fraction of what was there. At the low end, where a working dilution sits, it is possible to filter a solution and recover substantially less than was put in, with nothing visibly wrong.
Does pre-rinsing the filter help?
Considerably, and it is the single most effective thing to do. Passing a small volume of buffer, or a sacrificial portion of the solution itself, saturates the binding sites before the material that matters goes through. The principle is identical to pre-wetting a pipette tip: the first pass pays the adsorption cost so the second does not.
Where the solution is precious and the volume small, the sacrificial-portion approach is better than a buffer rinse, because it conditions the membrane with the same molecule rather than a different one. The cost is the portion discarded, which is known rather than invisible.
When filtration is necessary and when it is habit
The honest answer is that a good deal of routine filtration is habit carried over from protein work, and the question worth asking is what the filter is for.
Removing organisms. This is what a 0.22 µm filter does well, and it is the right reason to use one — typically where a solution is going into a cell culture and will sit there for days.
Removing particulate. A solution with visible particulate has a problem the filter will hide rather than solve. Undissolved material or an aggregate filtered out is peptide removed from the solution, and the concentration afterwards is not the concentration calculated.
Removing endotoxin. It does not. Lipopolysaccharide passes a 0.22 µm filter freely, because it is a molecule rather than an organism. A filtered solution is not a low-endotoxin solution and treating it as one is the commonest misunderstanding in this area.
Is filtering a reconstituted vial worth doing at all?
Usually not, if the vial was sterile to begin with and was entered cleanly. A sealed vial reconstituted with a bacteriostatic solution through an alcohol-wiped stopper is already about as clean as the filtration would make it, and the filter adds a loss step for no gain.
Filtration earns its place when the solution has been open, handled repeatedly, or made up in a non-sterile diluent, and is then going somewhere that cannot tolerate growth. Good technique at the vial removes most of the reasons to filter later.
What about the volume held inside the filter?
It is a separate loss from adsorption and it is often larger. A syringe filter retains a hold-up volume in the housing and membrane — typically a few tens of microlitres — which never comes out. On a 200 µL sample that is a substantial proportion of the volume, quite apart from anything that stuck to the membrane.
For small volumes the answer is usually a smaller filter. Low-dead-volume formats exist precisely for this, and choosing one is simpler than trying to chase the residue out with a buffer push, which dilutes whatever is recovered.
A practical order of operations
Where filtration is genuinely needed, the sequence that loses least is short.
- Choose a low-binding membrane. PES is the usual answer for aqueous peptide solutions. Nylon is the one to avoid.
- Match the filter to the volume. A large filter on a small sample loses more to hold-up than to adsorption.
- Pre-rinse or pre-condition. Buffer if the solution is plentiful, a sacrificial portion if it is not.
- Filter at the highest workable concentration. Filter the stock, then dilute, rather than filtering the working dilution. The fraction lost falls as concentration rises.
- Record that you filtered. It is a processing step and it belongs in the record alongside the concentration.
Why filter the stock rather than the dilution?
Because the loss is roughly a fixed mass rather than a fixed percentage. Taking the same milligram-scale hit out of a concentrated stock costs a fraction of a percent; taking it out of a working dilution can cost a noticeable share. Filtering early and diluting afterwards into clean vessels is almost always the cheaper route.
The caveat is that everything downstream of the filter has to stay clean, which is a handling question rather than a chemistry one. Sterile consumables and a disciplined sequence do more here than a second filtration would.
Can the loss be measured rather than estimated?
Yes, and it is worth doing once for any workflow that will be repeated. Measure the concentration before and after filtration on the same instrument, under the same conditions, and the difference is the loss for that membrane at that concentration. Quantitation against a standard is the reliable route.
Doing it once characterises the step permanently for that combination of peptide, membrane and concentration, which converts an unknown into a correction factor. For a compound used routinely, that measurement pays for itself the first time a result needs defending.
Does the solvent change which filter to use?
Yes, and getting it wrong destroys the filter rather than the sample. Cellulose acetate tolerates water and very little else; an organic co-solvent will attack it. PTFE handles organics well but is hydrophobic, so an aqueous solution will not pass through an unwetted one at all without priming.
Where a co-solvent is in play, check the membrane's compatibility before assuming the usual choice works. A filter that partially dissolves contaminates the filtrate with whatever leached out of it, which is a harder problem to detect than a lost fraction of peptide.
Should a filtered solution be re-measured?
For anything quantitative, yes, at least once per workflow. The concentration after filtration is the concentration the experiment actually used, and it is not the one written on the tube unless someone checked. The alternative is to characterise the loss once and apply it as a known factor, which is cheaper over many runs.
What does not work is assuming the loss is negligible because the solution looked the same coming out. Adsorption is invisible, and a clear filtrate at the expected volume tells you nothing about how much peptide is in it.
Does filtration change anything other than concentration?
It can remove aggregates preferentially, which sounds helpful and is worth thinking about carefully. A solution containing both monomer and aggregate, filtered, comes out enriched in monomer — so the filtered solution is not a diluted version of the original, it is a different composition.
For most work that is an improvement. For anything studying aggregation it quietly removes the thing being studied, and the measurement afterwards describes what survived the filter rather than what was in the vial.
Is a smaller pore size ever worth using?
Rarely, and it costs more peptide. A 0.1 µm filter is used where mycoplasma is a concern, and the smaller pore means more membrane surface and a larger adsorptive loss for the same volume. Unless there is a specific reason, 0.22 µm is the standard because it removes bacteria while costing less material.
Going the other way, a 0.45 µm filter is a clarifying step rather than a sterilising one. It removes particulate and will not reliably remove organisms, so using one and calling the result sterile is a category error.
All products referenced here are supplied for laboratory and research use only. They are not drugs, foods, supplements or cosmetics, and are not for human or veterinary use.
