The first thing to settle is the one that causes the most confusion on purchase orders: 0.2 µm and 0.22 µm describe the same filter. The two numbers are a historical artifact of different manufacturers rounding the same validation differently. What actually defines a sterilizing-grade membrane is the bacterial challenge: a filter earns the rating by retaining a challenge of at least 107 CFU per square centimetre of Brevundimonas diminuta under the method in ASTM F838, with zero passage. A membrane that passes that test is sterilizing-grade whether the label says 0.2 or 0.22. One that has not been challenged is not sterilizing-grade at any number printed on it.
With that out of the way, the real decision is format and membrane chemistry. This is where we see the most expensive mistakes, because both wrong answers look identical until the yield comes back low or the filter blinds halfway through a batch.
Syringe filter or capsule filter
The honest rule is volume per filtration event, and the limit is effective filtration area rather than anything printed on the housing. A 13 mm syringe filter has roughly 0.8 cm² of membrane, a 25 mm about 3.9 cm², and a 33 mm about 5.8 cm². A capsule filter starts around 300 cm² and runs past 1,200 cm². That is not a small difference in degree; it is a different operation.
| 13 mm syringe | 25 mm syringe | 33 mm syringe | Capsule | |
|---|---|---|---|---|
| Effective area | ~0.8 cm² | ~3.9 cm² | ~5.8 cm² | 300–1,200 cm² |
| Practical volume, clean solution | ≤10 mL | 10–100 mL | 100–200 mL | 1–50 L |
| Practical volume, viscous or particle-bearing | ≤2 mL | 10–25 mL | 25–50 mL | Depends on prefilter |
| Hold-up volume | ~10 µL | ~75 µL | ~150 µL | 10–60 mL |
| Connection | Female luer-lok inlet, male luer slip outlet | Same | Same | Luer, hose barb or sanitary flange |
| Driven by | Hand pressure, 10–20 mL syringe | Hand pressure | Hand or pump | Pump or pressure vessel |
| Typical use | Analytical samples, very small preparations | Single preparations, unit doses | Larger single preparations | Batch filtration into bags or bulk containers |
The hold-up row is the one buyers skip and then regret. A 33 mm filter retains roughly 150 µL in the housing and membrane. On a 2 mL preparation that is a seven percent loss before anything else goes wrong, and on a low-concentration preparation it is a potency problem rather than an inconvenience. Small-volume work belongs on the smallest housing that will pass the volume, not the largest one on the shelf.
Membrane chemistry decides your yield
We supply syringe filters in PES and PVDF, and the choice between them matters more than most stocking lists suggest. Polyethersulfone is hydrophilic, has the highest flow rate per unit area of the common membranes, and has low protein binding. It is the default for aqueous preparations and the right answer most of the time. Polyvinylidene fluoride has the lowest extractables of the two and the lowest protein binding of any membrane in routine use, which makes it the better choice for protein-containing and low-concentration preparations where losing a few micrograms to the membrane shifts the assay.
Nylon is the one to avoid in this setting. It is cheap, widely stocked and binds protein aggressively, so it belongs in solvent filtration and analytical work rather than anywhere near a biologic. PTFE is hydrophobic: it will not pass an aqueous solution at all until it is pre-wetted with alcohol, which is exactly why it is correct for venting and for organic solvents and wrong for everything else. We have had more than one customer order PTFE on price, find that nothing goes through, and conclude the filters were defective.
Extractables are the quieter issue. Every filter gives up some material to the stream that passes it, concentrated in the first few millilitres. For preparations where that matters, flush the filter with a few millilitres of diluent and discard before filtering the batch. On a capsule, flush with at least the hold-up volume.
A filter needle is not a sterilizing filter
This confusion is common enough to be worth stating plainly. A 5 µm filter needle exists to catch glass particulate when drawing from a snapped glass ampoule. Five micrometres is more than twenty times the sterilizing pore size; it removes glass, not organisms, and it is a particulate control step rather than a sterility step. The same goes for blunt fill needles, which exist to protect the technician's hands and the stopper, not the sterility of the contents.
Equally, a sterilizing filter is not a substitute for aseptic technique downstream of it. Filtration sterilizes the solution passing through the membrane; it does nothing for the container it lands in, the stopper that closes it, or the air it passes through on the way. If the receiving container is not sterile the filtration was decorative.
Integrity, and what blinding looks like before it happens
A sterilizing filtration is only as good as the evidence that the membrane was intact. For syringe filters in a compounding setting that evidence is usually a bubble point test after the filtration, performed per the manufacturer's stated value for that membrane and wetting fluid. A filter that fails bubble point after use invalidates the preparation that went through it, which is why the test comes after rather than only before.
The more common day-to-day failure is blinding — the filter clogging before the volume is through. It almost never arrives without warning. The pressure needed on the syringe plunger climbs, the flow slows, and somewhere in there a technician pushes harder, which is the moment the membrane is most likely to rupture. Three causes account for nearly all of it: filtering before a powder is fully dissolved, filtering a solution that has thrown a precipitate because it was prepared cold, and skipping a prefilter on a solution with visible particulate. Each is fixed upstream, not by a bigger filter.
The other mechanical failure we see is the connection. A luer slip connection under filtration pressure can separate, and when it does it does so suddenly and sprays the contents. Use luer-lok on both sides for anything beyond a few millilitres, and for pumped work use proper compounding tubing and transfer sets rather than improvising with syringe fittings.
What to stock
A pharmacy doing single preparations needs 25 mm PES as the workhorse, 13 mm PES for small volumes where hold-up matters, and PVDF in both sizes for protein-containing work. Add 5 µm filter needles if you draw from ampoules at all. A pharmacy doing batch work into bags needs capsules sized to the batch plus a prefilter, and should be buying them by the case rather than ordering against each batch — capsule lead times are the longest in this category and running out stops the room.
Filters are sterile, single-use and carry an expiry tied to their irradiation lot, so the beyond-use-date discipline in our cleanroom and garbing supplies guidance applies here too: order against consumption, not against a calendar. Where filtration sits in the wider stocking picture is covered in what to stock by room and on our compounding pharmacy supplies page.
Send us the preparations you filter, the volumes and whether anything is protein-containing, and we will come back with membrane and size recommendations and factory-direct pricing within one business day. Request a quote.
This article describes products and general industry practice. Your own SOPs, your filter manufacturer's validation data and the current text of USP General Chapter 797 govern how you filter and what you must document.