Ten Common Mistakes in Molecular Weight Cut-Off (MWCO) Testing, and How to Avoid Them

Molecular weight cut-off is the most quoted number for UF and NF membranes, and one of the easiest to get wrong. There is no active ASTM method for it (ASTM E1343 was withdrawn in 2010 without replacement), so labs and membrane makers use their own procedures. These are the mistakes we see most often.
1. Marker concentration too high
High marker concentrations build up at the membrane surface and form a gel or cake layer that changes retention. A 2024 review notes that tracer concentrations of 0.3–3.6 g/L used in earlier studies can cause artefacts and underestimate the MWCO; its authors worked at about 10 mg of carbon per litre. Keep markers dilute.
2. Flux and pressure not controlled or not reported
Retention depends on flux. The same membrane gives different curves at different pressures and crossflow velocities. Fix the flux (the review used about 20 L/m²·h for loose NF), keep it constant across tests and report it.
3. Comparing results from different marker families
PEG, dextran and proteins give different MWCO values for the same membrane. Flexible PEG chains pass loose NF membranes more easily than compact sugars, so PEG overestimates their MWCO. Compare like with like; see MWCO marker molecules by kDa.
4. Using charged proteins as size markers
Proteins carry charge. Bovine serum albumin (about 69 kDa) has been retained above 90% by a 300 kDa membrane at low ionic strength, through electrostatic repulsion. That says nothing about pore size. Use neutral markers for MWCO.
5. Treating a polydisperse marker as one molecular weight
A "PEG 10,000" or "dextran 70" sample contains a range of molecular weights. Analyse feed and permeate by gel permeation chromatography and calculate rejection across the distribution, rather than assuming one nominal value.
6. Mixing too many markers at once
Mixtures save time, but the review warns of solute–solute interactions and cake layers in which large markers hinder small ones. Validate a mixture against single-marker tests before relying on it.
7. Sampling before steady state
Markers adsorb onto membranes and rig surfaces, and some, such as large PEO on open membranes, deposit and inflate retention. Discard early permeate, wait for stable flux and permeate concentration, and run a clean-water flux check before and after.
8. Letting the feed concentrate
In a dead-end stirred cell, the feed gets stronger as permeate is removed, so later samples see a different feed. A 2008 study compared dead-end stirred cell and crossflow results for proteins and found the set-up matters. Keep recovery low, or use crossflow with the permeate returned to the feed tank.
9. Forgetting to rinse new membranes
New membranes often carry preservatives and wetting agents. They show up in total organic carbon and refractive index signals and distort the first results. Flush thoroughly with clean water first.
10. Reporting a single number without the method
MWCO is conventionally the molecular weight retained at 90%, but not every maker uses the same definition. Report the definition, the marker family, concentration, flux, pressure, temperature, cell type and analysis method, and include the full rejection curve.
Equipment that removes the variables
Our MWCO Testing System holds pressure, crossflow and temperature constant and logs flux throughout. Stirred cells and crossflow cells suit coupon work, and our MWCO testing service runs the test for you. Full method: how to measure MWCO.
Write to mail@techincresearch.com or call +1 (289) 622-1032.
Molecular weights are typical values; check the certificate of each marker lot. Equipment specifications may change; results depend on the membranes, markers and conditions, how the equipment is used and operation in line with the operation manual.
Sources
Separation and Purification Technology 354 (2025) 128612, Imbrogno, Calvo, Breida, Schwaiger and Schäfer, Molecular weight cut off (MWCO) determination in ultra- and nanofiltration: Review of methods and implications on organic matter removal, via KIT.
Separation and Purification Technology 62 (2008) 228–239, Becht, Malik and Tarleton, Evaluation and comparison of protein ultrafiltration test results: dead-end stirred cell compared with a crossflow system, via CORE.
Membranes 14 (2024) 26, Suryawirawan, Janssen, Boom and van der Padt, Bovine Serum Albumin Rejection by an Open Ultrafiltration Membrane: Characterization and Modeling.
ASTM International, ASTM E1343-90(2001) Standard Test Method for Molecular Weight Cutoff Evaluation of Flat Sheet Ultrafiltration Membranes (withdrawn 2010).





Comments