Accepted answer
Filtering is correct in exactly one scenario and counterproductive in most others. Work out the cost first, because it is larger than people expect, then work out what you are buying.
What it costs, in milligrams
Two loss mechanisms: hold-up volume (solution physically retained in the filter housing and membrane after you stop pushing) and adsorption (peptide sticking to the membrane and housing). Manufacturer hold-up figures for common syringe filters:
| Filter diameter | Typical hold-up volume | Loss at 5 mg/mL | % of a 10 mg vial |
|---|
| 4 mm | about 10 µL | 0.05 mg | 0.5 % |
| 13 mm | under 25 µL | up to 0.13 mg | 1.3 % |
| 25 mm | about 75–100 µL | 0.38–0.50 mg | 3.8–5.0 % |
| 33 mm | about 100 µL or more | 0.50 mg or more | 5 % or more |
Worked, for the 25 mm case most people buy because it is the cheapest per unit: 100 µL retained out of 2000 µL is 5.0 % of the volume, and 5.0 % of 10 mg is 0.50 mg. On top of that, adsorption. Membrane chemistry matters a great deal here:
| Membrane | Protein binding | Suitable for aqueous peptide |
|---|
| PES (polyethersulfone) | low | yes — usual first choice |
| PVDF, hydrophilic | low | yes |
| Cellulose acetate | low | yes |
| Nylon | high | no — avoid for peptide |
| PTFE, hydrophobic | low but will not wet | no — needs solvent prewet |
Choose nylon because it was on the shelf and you can lose a meaningful additional fraction to the membrane on top of the hold-up. With PES the adsorptive loss on a single small pass is modest but not zero, and it is worst at low concentrations, which is the opposite of intuition — a dilute solution loses a larger proportion because the binding sites on the membrane are a fixed quantity.
So a realistic all-in figure for a 25 mm PES filter on a 2 mL reconstitution is somewhere around 5 to 8 % of the vial. Call it 0.5 to 0.8 mg of a 10 mg vial, gone.
What it removes, and what it does not
- Bacteria and fungi: yes. 0.22 µm is the conventional sterilising grade and retains vegetative organisms reliably.
- Rubber coring fragments and fibres: yes, trivially — they are orders of magnitude larger.
- Endotoxin: no. This is the important one. Lipopolysaccharide is a molecular species and passes straight through a 0.22 µm membrane. A vial with a dead Gram-negative population filters to a beautifully clear solution with essentially unchanged pyrogenic load. Filtration therefore removes the evidence without removing the hazard, which is worse than not filtering.
- Mycoplasma and viruses: no. Not relevant to most of what you are doing, but for completeness, 0.22 µm does not retain them.
- Soluble aggregates: no. Large insoluble aggregates and visible particles yes; soluble oligomers pass.
The receiving container problem
You identified this yourself and it is the decisive objection. Filtering is a sterilising step only if it discharges into a sterile container through a sterile path, under conditions that keep the critical surfaces protected. Pushing filtered solution into a vial you rinsed, or back into the vial it came from through the same non-sterile septum, or into an open container on a bench, discards the benefit at the last step. You have then spent 0.5 mg and gained a clear solution of unknown sterility.
When it is actually right
One clear case: you have visible non-biological particulate — a coring fragment, a fibre, glass from a chipped neck — in material that is otherwise sound and expensive, and you want to remove the particle. There the filter is doing a job it can genuinely do, the target is mechanical, and no sterility claim is being made or needed. That is a legitimate use.
The case where it is clearly wrong: you suspect microbial growth. Then the only defensible action is to discard, because the filter cannot address the endotoxin and cannot tell you what grew.
3The point about dilute solutions losing a larger proportion to adsorption is right and is the reason people who filter 1 mg/mL solutions report bigger losses than they expect. – h_villanueva 7 months ago 4Filtration removing the evidence but not the hazard is the sentence that should be at the top of every thread about this. – lane_transit 8 months ago add a comment