PHARMA LAB · PL-03-014
Syringe filters: adsorption, compatibility and analyte loss

In this article
Filtration is often the last step before measurement, but it is not analytically neutral. Alongside removing particles, a filter can retain analyte or introduce substances that alter response. The useful question is not which membrane is generally “best”: it is whether the chosen system preserves the result under the actual method conditions.
Define the purpose before choosing pore size
Clarify why filtration is needed: protecting the instrument flow path, removing a precipitate or separating a defined fraction. Identify matrix, particles, viscosity, available quantity and target analytes. If analyte is associated with removed particles, the filtrate does not automatically represent total content. Define the fraction of interest before selecting a membrane.
Nominal pore size must suit the method and downstream equipment; simply matching a pore size to an instrument acronym is insufficient. This article concerns analytical preparation, not equivalence between clarification, sterilising filtration and sterility testing. A filter supplied sterile does not demonstrate absence of adsorption or interference in your analysis.
Assess the complete sample-contact system
Consider membrane, effective area, wettability, any prefilter, support, housing, connections and syringe. The polymer name does not describe all these elements. Document the exact configuration, including dimensions and declared surface treatments. A chemically compatible material can still interact with the analyte; a “low binding” claim requires understanding which substances and conditions were tested.
Check compatibility with solvent composition, relevant pH, temperature and contact duration using configuration-specific technical information and method testing. Do not assume a hydrophobic filter wets like a hydrophilic one, or that a wetting treatment is harmless: introduced liquid can alter composition and concentration. Match area to particle load and available volume rather than simply choosing the largest diameter.
Distinguish adsorption from retained liquid
Adsorption involves analyte interaction with a surface. Relative loss can be particularly important at low concentration or with small samples. Li and colleagues’ study of hydrophobic drugs demonstrates dependence on the drug–filter combination. Its rankings do not justify choosing the same material for every molecule.
Hold-up volume is liquid remaining in the device. For a homogeneous solution without adsorption, dilution or evaporation, retaining some solution reduces the amount collected but does not, by itself, change concentration. It can nevertheless compromise quantitative transfer or a calculation assuming recovery of the entire initial amount. Do not attribute every low concentration to “dead volume”.
The original Reid study, 2026, confirms effects of material, surface area, concentration and matrix for PFAS. This illustrates system dependence; it is not a universal membrane recommendation for pharmaceutical QC.
Investigate what the filter may add
A solvent blank passed through syringe and filter helps reveal substances introduced by that path. Compare it with the same unfiltered solvent and, when locating the source, a syringe-only control. Keep receiving vessel, times and conditions comparable. An instrument blank alone does not pass through the device and does not answer the same question.
Puah and colleagues observed extractable impurities from fluoropolymer filters, including after conditioning in some configurations. The finding concerns the tested conditions, not every filter made from that polymer. Assess interfering signals with the detector and method actually used: material suitable for one determination may be unsuitable for another that is more sensitive or less selective.
Build an interpretable comparative assessment
Define criteria, analytes, levels, matrices and independent replicates before testing. Include the relevant low level, not only a concentrated standard. An unfiltered reference is appropriate only if it can be measured without instrument damage, instability or particle-related bias. Do not deliberately inject an incompatible suspension to obtain a comparison.
Where needed, assess a justified alternative separation, such as centrifugation already demonstrated suitable, or representative material with an appropriate reference. The alternative may also change the measured fraction and requires justification. Compare concentrations on equivalent bases, collected amounts where relevant, blanks and method response. Addition studies follow the cautions discussed for recovery and matrix effects.
| Sample or condition | Risk to distinguish | Useful control |
|---|---|---|
| Low analyte concentration | Relatively significant adsorption | Recovery at relevant levels and comparison of collected fractions under a plan. |
| Different solvent or pH | Changed wetting, compatibility or response | Verify the complete configuration in the new composition. |
| Small available volume | Retained liquid and insufficient quantity | Collected volume and a balance consistent with the final calculation. |
| Particle-rich matrix | Clogging or removal of solid-associated analyte | Assess the required fraction and separation path. |
| Impurity testing or weak signals | Device-derived substances and interference | Comparable blanks and selectivity checks with the intended method. |
Manage pressure, conditioning and initial fractions
Use compatible connections and respect the limits of the syringe–filter combination. Unexpected resistance calls for stopping and assessment, not greater force, improvised tools or reuse of a single-use device. Follow containment and protection measures specified for the sample and solvent; detachment can cause splashes and loss of traceability.
Conditioning, rinsing or initial discard must follow evidence. There is no “first mL” valid for every analyte, concentration and filter. Check whether treatment changes dilution, blanks, recovery or available volume. When different fractions are collected during development, keep them identified: pooling too early can conceal a trend. The final procedure must clearly define the fraction to use.
Control changes and unexpected results
Simulated case. A concentrated control passes, but a low-level preparation shows reduced response after filtration. Collected volume is sufficient, and blanks do not show the interference being investigated. This weakens some explanations but does not yet prove adsorption. Planned comparisons using an alternative configuration, fractions and equivalent samples can strengthen or challenge the hypothesis; stability and response must remain controlled.
Retain original results and handle the anomaly under the procedure. Do not automatically correct the result using average recovery or introduce an unapproved additional discard. Record lot, configuration, quantity, conditioning and collected fraction in the sample preparation workflow.
A change in supply, material, diameter or housing requires impact assessment even when pore size is unchanged. ICH Q14 supports understanding of preparation parameters; EU GMP Chapter 6 provides the QC and records framework. Connect this assessment with other steps in the PL-03 area.
For the related steps, see: Sample concentration and evaporation: losses and recovery.
Sources and limitations
Consulted on 30 September 2026. Li et al., 2019, and Reid et al., 2026: indexed PubMed abstracts; direct access limited. Puah et al., RSC Advances 2019: publisher’s PDF. ICH Q14, EMA Step 5 revision 1; EU GMP Chapter 6, 2014. Evidence on specific molecules does not establish general rankings. Matrix and case are original; the case is simulated and is not a validated procedure.
Continue exploring
PL-03-015
Sample concentration and evaporation: losses and recovery
Reducing volume does not guarantee proportional enrichment. Control must follow the analyte through to the final solution actually measured.
Read the articlePL-03-013
Sample extraction: recovery, matrix effects and reproducibility
Good recovery of an added analyte does not by itself demonstrate extraction efficiency. Controls must distinguish loss, response effects and representativeness.
Read the articlePL-03-012
Serial dilutions: calculations, execution and error control
A dilution series must work on the bench and in the calculation: explicit volumes, links between solutions and controls for shared errors.
Read the article


