Jim Simmons is the Senior Director of Education at Spectrum Labs. He has over 30 years experience in membrane separations and spent 15 years as the Regional Sales Manager for the Central United States.
Characterizing a TFF process begins with membrane selection. Criteria include the membrane chemistry, selectivity, device geometry, channel, and pore size.
Membrane chemistry: The two main chemistries used in bioseparations are cellulosic and synthetic. Since most solutions used in biotech are aqueous, always consider the hydrophilicity or hydrophobicity of the membrane surface.
Membrane selectivity: How effective a membrane retains material of a given size.
Device geometry: Flat membrane or hollow fiber membrane. Most media chemistries are available in both formats, but some applications may favor one over the other.
Hydrophilicity or hydrophobicity: Most solutes used in biotech are aqueous. Polysulfone, including PS, is hydrophobic. Hydrophobic membrane must be ‘wetted’ prior to use – usually with an alcohol such as IPA – before they will pass aqueous solutions. In sterile processes, wetting agents must also be sterile, complicating the wetting process and limiting options for pre-sterilized devices. Proteins/antibodies tend to bind to hydrophobic surfaces. Bound proteins form a ‘boundary layer’ aka ‘gel layer’ on the membrane surface that can cause the pore size of the membrane to become tighter and slow the process and restrict the passage of proteins. Bound protein cannot be recovered, resulting in reduced yield.
Polyethersulfone, including PES and mPES, is hydrophilic. Hydrophilic membranes rinse easily with water eliminating the need for alcohol wetting and have little protein binding, often resulting in higher yields and faster processing.
Solute passage: The effectiveness with which the permeable species (solute) passes through a membrane. It is generally expressed as a percentage, ranging from 0% to 100%. Be aware that different solutes pass through different membrane materials with different efficiencies. While hydrophilic membranes freely pass aqueous solutes, agents may be present that can inhibit solute passage. Surfactants can coat membrane surfaces resulting in inhibited solute passage, so avoid their use if possible. Poor or non-optimal solute passage can lead to excessive buffer required to complete buffer exchange. In diafiltration, a rule of thumb is that with each diavolume processed, a 50% wash can be achieved. This rule only applies with 100% free solute passage, so as solute passage is inhibited, additional diavolumes are needed to achieve the wash. If less-than-ideal solute passage is unavoidable, consider a more open pore or increased membrane surface area to allow for acceptable processing times.
Pore Size: This chart gives us a look at the common size of various samples associated with tangential flow filtration.
Micro-Filtration Membranes: Membrane with a pore size range from 0.1mm to 10mm are typical for sterilization, contaminate and particle removal. Microfiltration membranes with pore sizes of 0.65µm, 0.45µm, 0.22µm, and 0.1 um are widely used for the separation of mammalian cells.
Ultra-Filtration Membranes: Membrane with a pore size range from 1KD to 1000KD NMWC (nominal molecular weight cutoff) are typical for protein and biological separation.
Virus-Removal Membranes: Membranes with a pore size range from 15nm to 50nm are typically used for virus clearance.
This illustration represents the commonly understood pore size selection criteria to select a pore that is 3 to 5 times smaller than the size of the species to be retained. It focuses on holding back the retained species.
The small car represents the sample and the potholes represent the pore structure of the membrane. The weight of the car represents applied transmembrane pressure, and shear rate is represented by the speed of the car driving down the road.
This is not recommended for deformable particles. It can be difficult to run high flow rates without damaging the sample. Processes using this pore size selection criteria must be carefully monitored for flux decline, damage to the sample, and lowered yields.
This illustration shows how critical pore size is to potential damage to the sample.
This illustration represents the pore size selection criteria to select the smallest pore that will freely pass the species to be passed. It focuses on the passage of the soluble species.
The illustration of a puppy on a grating shows a pore much smaller than the retained species. The grating represents the membrane pores and the puppy represents the sample.
Use of these criteria requires an understanding of the expected solute passage. For example, if a surfactant is present (which coats surfaces), or if a hydrophobic membrane is used (where protein binding can occur), this method will probably be less effective, but with highly selective hydrophilic membranes, this technique is growing in popularity. It is gentler on the sample, is less subject to membrane fouling, and results in a more forgiving process. If possible, use the smallest pore size that will freely pass the soluble species.
Here is a table of common membrane pore size and media chemistry selections for major biotech applications.
Rules of thumb are not absolute, so if possible, always confirm your assumptions with experimentation for a given sample and application.