A 0.22 micron membrane is the standard sterilising filter for a peptide solution because it is the pore size that bacterial challenge testing has validated as retaining the smallest routinely used test organism, and because it does the job without heating the peptide. Everything else about the procedure is a series of trade-offs: which membrane material you choose decides how much peptide reaches the other side, how dilute the solution is decides whether that loss matters, and the device's hold-up volume decides how much you never see again. This post covers why filtration rather than an autoclave, how to pick a membrane, where the material goes, and the cases where filtering is the wrong answer.
Why filter rather than autoclave?
Because an autoclave sterilises by doing chemistry to everything in the chamber, and a peptide is one of the things it does chemistry to.
Steam sterilisation is typically 121 °C under pressure. The degradation routes that peptide storage guidance exists to slow down — hydrolysis of the backbone, oxidation of methionine and cysteine, deamidation of asparagine and glutamine, and aggregation — are all accelerated by heat, and Manning and colleagues' review of protein stability is a catalogue of exactly this. Autoclaving a peptide solution reliably produces a sterile solution of something that is no longer entirely the peptide you started with, and nothing on the outside of the tube will tell you how much was lost.
Filtration is a physical separation instead. Organisms are larger than the pore; the peptide, at one to five kilodaltons for most of this catalogue, is thousands of times smaller than the pore and passes through in solution. Nothing is heated, nothing is oxidised by the process, and the molecule that goes in is the molecule that comes out — minus whatever stuck to the membrane, which is the subject of most of this post.
| Method | Sterilises | Effect on the peptide | Use |
|---|---|---|---|
| Steam autoclave, 121 °C | Yes | Accelerates hydrolysis, oxidation, deamidation, aggregation | Glassware, tips, media that tolerate it. Not peptide solutions |
| Dry heat, ~250 °C | Yes, and depyrogenates | Destroys the peptide | Glassware only |
| 0.22 µm filtration | Yes, for organisms | None chemically; loses material to adsorption | Peptide and protein solutions |
| Gamma or UV | Varies | Generates radicals, damages residues | Not for a peptide solution you intend to quantify |
Why 0.22 micron specifically?
Because that is where the validation sits. A sterilising-grade filter is not defined by a nominal pore size printed on the housing; it is defined by passing a bacterial challenge test, and the reference method for that is ASTM F838, which challenges the membrane with a high concentration of a small test bacterium and requires a sterile filtrate. USP General Chapter <1229.4> covers sterilising filtration of liquids as a process. Manufacturers rate the membranes that pass at 0.22 µm, and some write the same membrane as 0.2 µm — the two labels describe the same grade, and the difference is a rounding convention rather than a different product.
A 0.45 µm membrane is a clarifying filter, not a sterilising one. It will take out visible particulate and most fungal material and it will let small bacteria through. It has a legitimate use as a prefilter ahead of a 0.22 µm membrane when a solution is cloudy enough to blind the sterilising filter, but it is not a substitute for one.
Which membrane, and why not nylon?
This is the choice that decides your recovery, because a peptide in dilute solution will adsorb to whatever surface it meets, and membranes differ enormously in how much they hold.
| Membrane | Protein binding | Notes |
|---|---|---|
| PES (polyethersulfone) | Low | High flow, low extractables, the common default for aqueous protein and peptide solutions |
| PVDF, hydrophilic | Low | Filter manufacturers describe hydrophilic PVDF as their lowest protein-binding membrane; broad chemical compatibility |
| Cellulose acetate | Low | A long-standing low-binding option, less chemically tolerant than PES |
| Nylon | Binds protein | Avoid whenever recovery matters. Also a source of UV-absorbing extractables |
| Nitrocellulose | Binds protein strongly | It is used to immobilise protein on purpose. Do not filter a peptide through it |
| PTFE | Low, but hydrophobic | Needs pre-wetting with alcohol for aqueous solutions; better suited to organic solvents |
The practical rule is short: PES or hydrophilic PVDF for an aqueous peptide solution. Merck Millipore's product literature for its Millex range describes hydrophilic PVDF (Durapore) as its lowest protein-binding membrane and PES as its high-flow low-binding option, and other manufacturers say the same about their own equivalents. Nylon is cheap, ubiquitous in general-purpose filter drawers, and the wrong choice here for a reason that has nothing to do with sterility.
Two other things to check on the device before you use it: that it is sold sterile if you want a sterile filtrate into a sterile tube, and that the housing material tolerates whatever is in your diluent. A solution containing a substantial fraction of organic solvent — the kind sometimes needed for a stubbornly hydrophobic peptide, as the diluent post covers — can attack some housings and extract plasticiser into your sample.
Where the material actually goes
Three separate losses, and they add up differently depending on how dilute your solution is.
Adsorption to the membrane. Peptide molecules bind to the membrane surface until the available sites are occupied, so the absolute amount lost is roughly a property of the membrane area rather than of your concentration — which is precisely why it hurts dilute solutions most. The same few micrograms is a rounding error out of a 2.5 mg/mL stock and a large fraction of a nanogram-per-millilitre working solution. It is the same mechanism as the wall losses in the surface adsorption post, concentrated into a few square centimetres of deliberately high-surface-area material. We are not going to quote a percentage: it depends on the membrane, the area, the peptide's hydrophobicity and the buffer, and we do not have a measured value for any catalogue compound to cite.
Hold-up volume, also called dead volume. Every filter device retains some liquid in the membrane and the housing that never comes out. Manufacturers publish this figure per device, it scales with membrane diameter, and it is the number to look up rather than estimate. Filtering 200 µL through a device with a hold-up volume of the same order as your sample is how a solution disappears.
Transfer losses. Syringe, needle, tube walls, tip. Small individually, real when the volume is small.
The standard mitigations all work by making the membrane's appetite someone else's problem:
- Filter the largest volume you can at the highest concentration you can, then dilute afterwards. A concentrated stock saturates the membrane with a negligible fraction of itself.
- Filter before diluting, not after. Filtering the working solution puts the membrane in contact with exactly the material you cannot afford to lose.
- Discard the first fraction if volume allows, and collect the rest. The first millilitre through takes the worst of the adsorption.
- Pre-wet or passivate the membrane with buffer, or with a sacrificial carrier protein if the assay tolerates one — a common practice for very dilute protein stocks, though it introduces a component you then have to declare.
- Choose the smallest membrane diameter that will pass your volume, because less area means less binding and less hold-up.
- Do not filter what does not need filtering. A stock that will be diluted into non-sterile buffer for a binding assay does not need to be sterile.
For a protein rather than a short peptide the losses are worse, because the working concentrations are lower. IGF-1 LR3 is the catalogue example, and its own post covers why a 1 mg vial goes so much further than the number suggests — and therefore why every microgram left on a membrane is a larger share of the experiment.
When filtration is not the right move
When the problem is endotoxin. A 0.22 µm membrane removes bacteria, not lipopolysaccharide. Endotoxin passes straight through, and the endotoxin and sterility posts explain why a sterile filtrate can be pyrogenic. Filtering does not fix a water problem; buying certified water does.
When the problem is the preservative. Benzyl alcohol is dissolved, so it goes through. The only way to have an unpreserved stock is to start with unpreserved water, which is what the sterile water post is about.
When the solution is visibly cloudy or has particles. You can filter it, and you will get a clear sterile solution of unknown concentration, because what you removed was aggregated peptide. The aggregation post covers why that is a result rather than a fix: a clear filtrate after a cloudy stock means material was taken out, and nothing tells you how much.
When the volume is small and the concentration is low. Below a certain point the filter costs more peptide than the sterility is worth. Aliquoting into sterile tubes from a vial that was reconstituted with sterile technique, with no filtration step at all, is often the better answer.
When the peptide is supplied in solution already. PT-141 and the liquid sprays are formulated and filled as solutions; the PT-141 solution post covers why. Re-filtering a formulated solution adds an adsorption loss and removes nothing useful.
What to record
The membrane material, the pore size, the device, and the fact that you filtered at all — because a filtration step between the certificate and the assay is a place where concentration can change. The methods-section post has the wording. It is worth remembering what the certificate does and does not cover here: every certificate in the COA library reports purity by RP-HPLC at 214 nm on the powder as supplied. It says nothing about what your membrane kept.
Frequently asked questions
Is 0.2 micron different from 0.22 micron?
No. They are two labelling conventions for the same sterilising grade of membrane, and both describe a filter validated by bacterial challenge testing. What matters is whether the manufacturer rates the device as sterilising grade, not which of the two numbers is printed on it.
Do I need to filter if I reconstituted with sterile water and a sterile needle?
Usually not. If the powder was sterile-filled, the water is sterile, the stopper was wiped, and the solution goes straight into sterile tubes, you have not introduced anything a filter would remove — and you have avoided the adsorption loss. Filtration earns its place when the diluent or the technique is uncertain, or when the solution is going into long-term culture where a single organism has weeks to grow.
Can I autoclave the lyophilised powder instead?
No. Dry heat and steam both degrade the peptide, and the powder is already the most stable form it will be in. The sterility of the powder is the manufacturer's responsibility and is what the certificate's sterility line addresses where one is reported.
Does filtering change the concentration?
It can only lower it, and it does so invisibly. Adsorption takes a small and unknown amount, and if any aggregate was present, filtration removes that too. This is the main argument for filtering a concentrated stock rather than a dilute working solution, and for quantifying after filtration rather than before if the number has to be right.
Which is better, a syringe filter or a centrifugal one?
For a few hundred microlitres to a few millilitres, a small-diameter syringe filter with a PES or PVDF membrane is the usual choice and gives you control over how fast you push. Centrifugal devices avoid the syringe entirely and suit very small volumes, but check the membrane — some are supplied with materials chosen for speed rather than for protein recovery.
Can I reuse a filter?
No. A used membrane is partially blinded, already loaded with adsorbed peptide, and no longer has a validated retention claim. A device rated as sterilising grade is rated for one use.
References
- ASTM F838-20. Standard Test Method for Determining Bacterial Retention of Membrane Filters Utilized for Liquid Filtration. ASTM International, 2020. Read 29 September 2026. store.astm.org/f0838-20.html
- United States Pharmacopeia. General Chapter <1229.4> Sterilizing Filtration of Liquids. www.usp.org
- Merck Millipore. Millex syringe filters, product literature: Durapore (PVDF) described by the manufacturer as its lowest protein-binding membrane. Read 29 September 2026. www.merckmillipore.com/INTL/en/products/filtration/laboratory-syringe-filters/millex-syringe-filters
- Manning MC, Chou DK, Murphy BM, Payne RW, Katayama DS. Stability of protein pharmaceuticals: an update. Pharmaceutical Research 2010;27(4):544-575. doi.org/10.1007/s11095-009-0045-6
Every product mentioned is sold for laboratory research use only and is not for human or animal use. Nothing on this page describes or recommends use of the material sold here in humans or animals.
