If Membrane Cassettes From Different Manufacturers Are Labeled With The Same MWCO, Are There Actual Performance Differences?

We purchased membrane cassettes with the same MWCO from two different manufacturers. During diafiltration experiments, we observed a significant difference in product recovery. Further analysis showed that the permeation rate of the sample through one membrane cassette was much higher than through the other. This raises a question: For membrane cassettes from different manufacturers or different models that are labeled with the same MWCO (Molecular Weight Cut-Off), are the actual pore size distributions and retention behaviors different?

 

First, it should be clarified that MWCO (Molecular Weight Cut-Off) is essentially a nominal specification rather than an absolute pore size.

 

MWCO is typically defined as the molecular weight of a standard marker molecule at which approximately 90% rejection (90% retention) is achieved. This definition means that MWCO represents only one point on a full retention curve, rather than a complete description of membrane structure. In reality, membranes do not have a single uniform pore size. Instead, they exhibit a pore size distribution, or more precisely, a distribution of effective transport channel sizes. Many references also point out that there is no fully unified standard for MWCO determination, and different manufacturers may use different marker molecules, test conditions, and calculation methods. Therefore, the MWCO values reported by different suppliers are not directly comparable. As a result, membranes with the same labeled MWCO do not necessarily exhibit the same retention performance or separation behavior in actual applications.

 

Second, Why do membranes with the same MWCO (e.g., 10 kDa / 30 kDa) from different manufacturers perform differently?

 

A. Different marker molecules used for MWCO calibration
Different manufacturers may use different standard solutes for MWCO determination, such as PEG/PEO, dextran, or proteins (e.g., BSA). These markers have different molecular structures and hydrodynamic behaviors. For example, linear PEG and more "globular" dextran may have different hydrodynamic sizes even at the same molecular weight, which can lead to systematic differences in MWCO determination. When proteins are used as markers, additional effects such as adsorption and electrostatic interactions can further influence the measured "90% rejection point," making it less stable.

B. Different testing conditions
MWCO is not determined solely by pore size; it is also affected by operating conditions such as transmembrane pressure (TMP), concentration, temperature, crossflow rate, and concentration polarization. Changes in these parameters can alter the probability of solute passage through the membrane, causing shifts in the measured MWCO even for the same membrane. Many manufacturers also note that molecular shape, sample concentration, and operating conditions can significantly affect actual retention, so MWCO should only be considered a guideline.

C. Differences in membrane material and surface chemistry
Even if pore sizes are similar, retention behavior may still differ due to variations in membrane material (PES, PVDF, regenerated cellulose, polyamide, etc.), surface charge, hydrophilicity/hydrophobicity, and protein adsorption properties. These factors all influence the actual sieving performance.

D. Manufacturing variability and differences in pore size distribution width
Even with the same MWCO, the sharpness of the retention curve can vary significantly. MWCO represents only the 90% rejection point: two membranes may both achieve 90% retention at 10 kDa, but one may have a narrow pore size distribution (steep curve), while the other has a broader distribution (gentler curve). As a result, for molecules in the 3–20 kDa range, leakage or passage can differ greatly, leading to significant differences in real-world performance.

 

Third, What is the most accurate way to describe MWCO?

Even when different manufacturers label membrane cassettes with the same MWCO, actual performance differences may still exist. These differences arise from variations in calibration methods, marker molecules, testing conditions, membrane materials/surface chemistry, pore size distribution, and batch-to-batch variability.

 

Forth, What parameters should be focused on during selection, replacement, or incoming quality verification instead of relying solely on MWCO?

The following are recommended key parameters (as suggested by our purification specialists) that should be requested from suppliers:

MWCO determination method
Including:

• Type of marker molecule used (PEG/dextran/protein)
• How the 90% rejection point is calculated or interpolated
• Test setup (stirred cell or crossflow system)
• Testing conditions such as TMP, temperature, flow rate, and concentration

Full retention/sieving curve instead of a single MWCO value
The complete separation profile provides a much more reliable indication of actual performance than a single-point MWCO.

Water flux/permeability under defined test conditions
Including temperature, pressure, and other operating parameters, used to evaluate the "tightness or looseness" of membranes with the same MWCO.

Batch-to-batch consistency / Certificate of Analysis (COA) data

At minimum, this should include:

• Pure water permeability
• Integrity test results (if applicable)
• Typical retention or leakage performance indicators

Fifth, Practical recommendations for procurement and process scale-up

Do not rely on the same MWCO as the sole criterion for equivalence or replacement. At a minimum, the following key parameters should be aligned:

• MWCO calibration method
• Pure water permeability (flux)
• Full retention/sieving curve

 

If achieving high retention is the primary goal, many manufacturers recommend selecting a membrane with an MWCO that is approximately 2–6 times smaller than the target molecule size as an empirical rule, especially for proteins.

For nanofiltration or loose / near-NF MWCO ranges (low molecular weight applications), extra caution is required. Retention is not governed purely by size exclusion; factors such as surface charge and solution ionic strength can significantly amplify performance differences. Therefore, it is strongly recommended to perform testing under actual process conditions before final selection or scale-up.