Rejection Rate Testing and Analysis Methods for Membrane Filtration: A Technical Guide

Rejection rate testing and analysis methods are essential for evaluating the separation performance of reverse osmosis (RO), nanofiltration (NF), ultrafiltration (UF), and microfiltration (MF) membranes. This technical guide explains the definition and calculation of rejection rate, standard test procedures, applicable sample types, analytical instrumentation, and typical industrial applications. By following recognized standards such as ASTM and GB/T methods, manufacturers, researchers, and accredited third-party laboratories can obtain reproducible and comparable rejection rate data.

Technical Overview

The rejection rate (also called retention rate or salt rejection) is the most important performance indicator for a separation membrane. It quantifies the percentage of a specific solute, ion, particle, or microorganism that the membrane prevents from passing into the permeate stream. The observed rejection rate is calculated as follows:

R = (1 − Cp / Cf) × 100%

Where R is the rejection rate (%), Cp is the solute concentration in the permeate, and Cf is the solute concentration in the feed. For example, if a feed contains 2,000 mg/L of sodium chloride and the permeate contains 40 mg/L, the observed NaCl rejection rate is 98%.

In membrane science, a distinction is made between the observed rejection rate and the intrinsic (real) rejection rate. The observed value is based on the bulk feed concentration, while the intrinsic value is based on the solute concentration at the membrane surface (Cm). Because of concentration polarization, the surface concentration is higher than the bulk concentration, so the observed rejection rate is usually lower than the intrinsic rejection rate. Careful control of cross-flow velocity, turbulence promoters, and recovery is therefore required to obtain meaningful and comparable results.

Rejection rate testing is governed by several international and national standards, including ASTM D6161 (terminology for hollow fiber UF and MF membranes), ASTM E1343 (molecular weight cutoff evaluation of flat sheet membranes), ASTM F838 (bacterial retention of membrane filters), ASTM D1125 (electrical conductivity of water), GB/T 20103 (membrane separation terminology), and the relevant GB/T and HY/T membrane test method standards. Accredited third-party laboratories and in-house quality control departments routinely apply these methods for product qualification, incoming inspection, and research.

Testing Samples

Flat Sheet Membrane Coupons

Flat sheet samples cut from membrane sheets or removed from elements are the most common test specimens. Coupons are tested in dead-end stirred cells or small cross-flow cells with an active area of typically 10 to 50 cm². Flat sheet testing is fast, requires little sample material, and is ideal for research and development, material screening, and quality control of membrane production batches.

Hollow Fiber and Tubular Membranes

Hollow fiber membranes, capillary membranes, and tubular membranes are tested as single fibers or small fiber bundles potted into laboratory modules. The potting resin seals the fiber ends so that permeate can be collected separately from the retentate. These samples are typical for UF and MF products used in water purification, membrane bioreactors, and hemodialyzers.

Spiral Wound Elements and Complete Modules

Full-size spiral wound elements (for example, 2.5 inch, 4 inch, and 8 inch diameter) and complete UF or NF modules are tested on standardized cross-flow test stands. Element-level testing verifies the performance of glue lines, feed spacers, and the overall construction, and it produces the certified performance data (permeate flow and salt rejection) stated in product datasheets.

Ceramic and Specialty Membranes

Ceramic, polymeric composite, and other specialty membranes used in aggressive chemical or high-temperature service can be tested with the same methodology, provided that the test cell materials, seals, and feed solutions are compatible with the membrane chemistry and the intended operating conditions.

Testing Items

Depending on the membrane type and end use, the following rejection and retention parameters are typically measured:

• NaCl rejection rate: the key index for RO membranes and an important index for NF membranes.
• Divalent salt rejection rate: MgSO4, CaCl2, and Na2SO4 rejection, which is the primary index for NF membranes.
• Organic solute rejection: rejection of neutral solutes such as glucose, sucrose, glycerol, and ethylene glycol, used to characterize the pore size and free volume of NF and RO membranes.
• Dye and pigment rejection: measured by UV-Vis spectrophotometry for wastewater treatment and textile applications.
• Molecular weight cutoff (MWCO): the molecular weight of a reference solute (PEG, dextran, or protein) that is rejected at 90% or another defined level, used for UF membranes.
• Particle rejection: retention of monodisperse latex beads or natural particles, used for MF and UF membranes.
• Bacterial and viral retention: quantified as a log reduction value (LRV) for sterilizing-grade and virus filtration membranes.
• Companion measurements: pure water flux, permeability, recovery, and pressure drop are recorded together with rejection data because rejection depends strongly on operating conditions.

Testing Methods

Conductivity-Based Salt Rejection Method

For RO and NF membranes, the feed is prepared as a defined salt solution, and the electrical conductivity of the feed and permeate is measured with a calibrated conductivity meter following ASTM D1125. The rejection rate is calculated directly from the conductivity ratio. Typical industry-standard test conditions include 2,000 mg/L NaCl at 1.55 MPa (225 psi) and 25 °C with 15% recovery for brackish water RO membranes, and 32,000 mg/L NaCl at approximately 5.5 MPa (800 psi) for seawater RO membranes. NF membranes are commonly evaluated with 500 to 2,000 mg/L MgSO4 at 0.48 to 0.70 MPa. For accurate results, the temperature is held constant, the system is allowed to reach steady state before sampling, and the conductivity meter is calibrated with standard solutions.

Chromatographic and Spectrophotometric Methods for Organic Solutes

When the solute is non-ionic or the solution matrix is complex, conductivity is not suitable. Organic rejection is instead determined by quantifying the solute concentration in the feed and permeate using total organic carbon (TOC) analysis, high-performance liquid chromatography (HPLC), gel permeation chromatography (GPC), UV-Vis spectrophotometry, or gravimetric analysis. The same rejection formula applies, but the analytical technique is selected according to the solute chemistry, expected concentration range, and required detection limit.

Molecular Weight Cutoff Determination

MWCO testing follows the approach of ASTM E1343. A series of polyethylene glycol (PEG) or dextran standards spanning the expected cutoff range is filtered through the membrane, and the concentration of each fraction in the feed and permeate is measured by GPC with a refractive index detector or by TOC. A rejection curve is plotted as rejection rate versus molecular weight, and the MWCO is reported as the molecular weight at which the rejection reaches 90%. Protein standards such as bovine serum albumin, ovalbumin, and cytochrome c may also be used for bioprocess membranes.

Dead-End Stirred Cell Testing

Flat sheet coupons are clamped into a stirred pressure cell, pressurized with nitrogen or another inert gas, and operated in dead-end mode with magnetic stirring to reduce concentration polarization. This method is simple and reproducible and is widely used for membrane material research, MWCO determination, and quick screening of rejection performance. The stirring speed, applied pressure, and solute concentration must be reported together with the rejection result.

Cross-Flow Test Loop Evaluation

Cross-flow rigs circulate the feed across the membrane surface at a controlled flow rate and pressure, closely simulating real operating conditions. The loop is equipped with a high-pressure pump, flow meters, pressure gauges, temperature control, and sampling ports for feed, concentrate, and permeate. Cross-flow testing is the reference method for RO and NF elements and for hollow fiber modules, because it allows precise control of flux, recovery, and cross-flow velocity, which are the main variables governing concentration polarization and observed rejection.

Challenge Testing and Log Reduction Value

For MF membranes used in sterile filtration and biopharmaceutical virus removal, retention is verified by challenge testing rather than simple percentage rejection. According to ASTM F838, sterilizing-grade filters are challenged with a defined concentration of Brevundimonas diminuta, and retention is expressed as a log reduction value: LRV = log10(Cf / Cp). A minimum LRV of 7 is generally required for sterilizing-grade claims. Virus filters are similarly challenged with bacteriophages or model viruses, with LRVs of 3 to 6 or higher depending on the application.

Data Processing and Normalization

Rejection data must be normalized before comparison, because membrane performance varies with temperature, pressure, recovery, and feed concentration. Permeate flow is corrected to a reference temperature using a viscosity correction factor, and salt passage is corrected using standard normalization practice for RO systems, consistent with the principles of ASTM D4516. All test reports should state the membrane type, active area, pre-cleaning or compaction procedure, feed composition, applied pressure, cross-flow rate, temperature, recovery, steady-state criteria, and the analytical method used for concentration measurement.

Applications

Rejection rate testing and analysis methods are applied across the entire membrane value chain:

• Membrane manufacturing: routine quality control of production batches, certification of element performance, and release testing before shipment.
• Water treatment: verification of RO and NF performance for seawater desalination, brackish water purification, and boiler feed water production.
• Wastewater treatment and reuse: confirmation of retention of salts, heavy metals, dyes, and micropollutants in industrial effluent polishing systems.
• Food and beverage: evaluation of UF and NF membranes for dairy protein concentration, whey processing, juice clarification, and sugar concentration.
• Pharmaceutical and biotechnology: validation of sterile filtration, virus filtration, and tangential flow filtration steps required by regulatory filings.
• Research and development: characterization of new membrane materials, comparison of fouling behavior, and lifetime or chemical cleaning studies.
• Procurement and incoming inspection: independent verification by accredited third-party laboratories of manufacturer performance claims before installation.

FAQ

Q: What is the difference between rejection rate, retention rate, and desalination rate?

A: These terms are used interchangeably in membrane technology. All of them describe the fraction of a solute retained by the membrane, calculated as R = (1 − Cp/Cf) × 100%. "Desalination rate" and "salt rejection" are used specifically for salt solutions, while "retention" is more common in bioprocess and sterile filtration contexts.

Q: Why is my measured rejection rate lower than the datasheet value?

A: Common causes include test conditions that differ from the standard rating conditions (feed concentration, pressure, temperature, recovery), insufficient pre-flushing or compaction of the membrane, inaccurate calibration of the conductivity meter, air leakage in the test cell, and concentration polarization caused by inadequate cross-flow or stirring. Always compare data at normalized, steady-state conditions.

Q: How is molecular weight cutoff related to rejection rate?

A: MWCO is derived from a rejection curve. A membrane is challenged with a series of molecular weight standards, and the MWCO is defined as the molecular weight at which the rejection rate reaches 90%. A tighter membrane shows a given rejection level at a lower molecular weight, corresponding to a smaller MWCO value.

Q: Which concentration measurement technique should be selected?

A: Conductivity is the simplest and most accurate method for single-salt solutions. TOC and HPLC are preferred for organic solutes, GPC with refractive index detection for polymer standards in MWCO testing, UV-Vis for dyes and proteins, and plate counts or particle counters for microbial and particle retention studies.

Q: How is microbial retention reported?

A: Because sterilizing-grade membranes retain essentially all challenge organisms, the result is expressed as a log reduction value (LRV = log10 of the feed-to-permeate concentration ratio) rather than a percentage. A claim of sterilizing-grade filtration is typically supported by an LRV of at least 7 under the challenge conditions of ASTM F838.

Q: Can rejection rate testing be outsourced?

A: Yes. Accredited third-party laboratories perform rejection rate, MWCO, and challenge testing according to ASTM, ISO, GB/T, and other recognized standards, providing independent reports suitable for supplier qualification, regulatory submission, and dispute resolution.