Considerations for buffer filtration.
Buffers and cleaning chemicals are used throughout the production of monoclonal antibodies and other proteins produced through biotech processes. During the fermentation process, buffers can often by used for pH adjustment. A variety of buffers can be required for capture and purifying chromatography steps. In some processes, multiple chromatography purification steps may be required. Buffers used for chromatography can have a high, low or neutral pH. In addition to chromatography, buffers or low pH caustic solutions are also used for tangential flow filtration (TFF).
All additions to bioreactors require sterile filtration since the bioreactor must be protected from contamination. Although the chromatography and TFF applications may not be sterile processes, the buffers used in these unit operations are typically sterilized by 0.2 micron filters to provide bioburden control for the downstream process.
Buffers require 0.2 micron filtration. Sterilizing grade hydrophilic membrane filters are tested with a liquid bacterial suspension challenge, which is sensitive enough to detect the passage of any microorganisms, to establish of the micron rating of the filter. For sterilizing grade 02 [micro]m rated membrane filters, the test organism is Brevundimonas diminuta (ATCC 19146), The organism and minimum challenge level ([10.sup.7] CFU/[cm.sup.2] filter area) are described in PDA Technical Report 26. Since it is generally not possible to perform bacterial challenge tests in a production environment, a correlation is made to a physical integrity test by the filter manufacturer From that correlation, filter performance can be verified using a validated integrity test and every filter is integrity tested during manufacture. For example, a 0.2 micron filter is typically correlated to a water based Forward Flow integrity test. The Forward Flow test quantitatively measures the diffusive flow as well as flow through any open pores in a wetted membrane filter at a given pressure. Although the correlation is made with water, other compatible fluids can also be correlated to the bacterial challenge results.
In the validation of a 0.2 micron filter by the filter manufacturer, the correlation should be made with an adequate number of filters as well as with filters from at least three lots. The test pressure for the integrity test should be high enough to be a critical test, but not too close to the bubble point where the diffusive flow can be erratic.
The buffers used in the production of a biotechnology drug tend to range from very high to very low pH. In many processing operations, the use of a single filter that is compatible with all fluids simplifies the operation. A filter that is incompatible with a fluid, if used, could release a high level of extractables to the filtered fluid, and is not desirable for buffer filtration.
A number of hydrophilic membrane materials are available in 0.2 micron sterilizing filters; these include Nylon 66, PVDF, PTFE, and PES. A hydrophilic membrane is desired because water based fluids will wet the filter without the use of a wetting agent. Conversely, a hydrophobic material (such as PTFE) will repel water. In order to wet out a hydrophobic material, it is necessary to wet first with a low surface tension solvent such as isopropyl alcohol (IPA). Another requirement in many applications is the use of a single use capsule filter. A single use capsule does not require cleaning or cleaning validation and can be provided pre-sterilized by gamma irradiation, thus simplifying the process operations, since a pre-sterilized filter does not require assembly and does not require steam sterilization.
Based on Table 1, the best membrane for buffer filtration is PES, since it provides broad pH compatibility, is water wettable and can be gamma irradiated.
The filter membrane is assembled into a cartridge format and the material of construction for the supporting filter structure (cage, core, endcaps, and adaptors) are usually polypropylene, which is compatible with most acids and bases. The housing material used in capsule filters is also often polypropylene. For capsules that are gamma irradiated, the polypropylene used must be a resin that is gamma stable to ensure compatibility and low extractables.
A membrane filter, as well as all system components, can contribute extractables to the buffer. The extractables are measured as non-volatile residue after extraction for a set period in a solvent at a given temperature. Extractables can be determined for specific buffers.
Extractables are typically composed of the oligomers or additives of the plastic materials present in the filter element. The amount of extractables from a filter element, as well as any extractable from the rest of the process system, can be reduced by flushing the system prior to filtration. Extractables can be determined for specific buffers.
Based on compatibility, the membrane filter of choice for buffer filtration is PES. In terms of wettability of membrane material, Nylon 66 is the most easily wet, since it is naturally hydrophilic. PVDF is naturally hydrophobic and is modified to be hydrophilic. It is close to Nylon 66 in terms of wettability. PES membrane is generally less wettable than PVDE PTFE is not water wettable. However, it is important to note that all PES membranes do not have the same degree of wettability. The ability to wet a filter completely is important for the filtration, since the objective is to use the full filtration area. If a portion of the membrane is not wet properly the filtration flow can be impaired. In addition, it is necessary to have a fully wet membrane during the performance of an integrity test. If the filter membrane is not completely wet, then the filter could fail the integrity test causing process delays and re-testing. Most PES filters do not wet easily and usually require back pressure, making the test more difficult and time consuming. Some require rewetting after steaming to integrity test.
Filterability and system size
The objective is to size the smallest system for the application. Key factors to consider are filter flow rate, membrane porosity, filter construction and area.
Flow Rate- Required flow rate, maximum pressure drop and available pressure for the buffer filtration must be considered to size the system. The flow rate achievable through a filtration system is directly related to the applied pressure and inversely related to the resistance to flow. If the applied pressure is increased, the flow rate will increase. If the resistance to flow is increased, such as by membrane plugging, then the flow rate will decrease. Test conditions should match process conditions for pressure as closely as possible.
Porosity -The nature of microporous membrane filters is that they will tend to plug rapidly if they are subjected to a relatively high flow rate (directly related to a high applied pressure) during the start-up of filtration. For fluids with a significant particle loading, under conditions of initial high flow rate, the microporous membrane can become rapidly plugged, or fouled, the pressure drop will increase and the throughput, or filter life, will be reduced. Buffers are generally fluids with a low particle loading, however, and small scale filterability testing can be used confidently to calculate the size of the full scale system.
Construction-Lifetime (throughput) and flow rate at process scale are highly influenced by the construction of the membrane, Supor[R] EKV filters feature a serial layer construction where the upstream layer is an asymmetric, high porosity PES membrane using Pail's Supor[R] Mach V technology. The final layer is a symmetric sterilizing 02 micron membrane. The asymmetric pre-filter membrane has high retention efficiency for particulates to give long lifetime and protection of the final 0.2 micron layer
Supor EKV filter cartridges are made by the patented Ultipleat[R] membrane pleating technology to form a filter pack that allows up to 50% greater filtration area and better flow distribution in the membrane cartridge. The strength of the design means that a higher differential pressure (1 bar vs. 300 mbar for most fan pleated filters) can be reached during SIP operation without damaging the filter.
The scaled-down filterability test requires a relatively small representative sample of the process fluid (typically 2 ~ 4 liters). Membrane filter discs with a 47 mm diameter are typically used. If a variety of buffers are used, then the worst case fluid should be defined and tested. The filterability tests can be performed at either a constant pressure or a constant flow rate. Either test type can be used to design a system. However, it is recommended that the scaled-down test use the same conditions as the process scale. For example, if the process scale filtration is performed with a constant flow, then it is best to use constant flow for the scaled down test. Optimization parameters can include time (or throughput) to reach terminal (maximum allowable) differential pressure at a given flow rate or a minimum flow rate for a given applied pressure. The flow rate and throughput (total volume filtered) can be scaled linearly based on filtration area.
System scaling for a constant flow test can be based on the following ratios:
Full scale system area/Scaled down system area = Full scale system flow rate/Scaled down system flow rate
Full scale system area/ Scaled down system area = Full scale system throughput/Scaled down system throughput
Considerations for filterability and sizing
When performing a filterability test with a variety of membranes in 47 mm disc format, the disc with the highest flow rate may not size to the smallest system because filter area of the pleated cartridge is also important.
Table 2 shows the water flow rates for Supor EKV and SuporLife 92DP. The SuporLife[R] 92DP is a serial 0.8 over 0.2 micron construction with symmetric membranes. The 92DP has a higher flow rate that the EKV for the 47 mm disc. To scale up to a filter cartridge (10 inch), the total area of the filter needs to be considered.
The EKV filter uses Pall ultipleat technology and the filter area is greater than that of the 92DP. This translates into a higher flow rate for a 10 inch EKV filter, even though the scaled down test indicated that the flow rate for the 92PD filter was higher than the flow rate for the EKV filter.
If the required scale up flow rate for the system was 75 L/min. then two 92DP filters would be required to meet the flow rate requirement, whereas only one EKV filter would be needed.
For constant pressure tests, the set pressure can impact the maximum throughput possible
As mentioned above, if the initial pressure is too high for a microporous membrane, the membrane can foul and the capacity can be diminished. The example study to the right was a filterability test performed with a sodium phosphate buffer (pH 6.8) using a 47 mm PVDF membrane (0 002 [ft.sup.2]/2 [cm.sup.2]). The test was performed with a pressure of 10 psi and 5 psi. As shown, the disc with a 10 psi constant pressure started with a higher flow rate than the disc tested at 5 psi However, the flow rate for the disc at 10 psi decreased rapidly. It is possible to determine the maximum theoretical throughput for a membrane at a given pressure by plotting the time/volume vs. time and drawing a straight line through the data. The inverse of the slope of the line is the maximum volume (Vcap) for the filter under the test conditions. For the sodium phosphate test, the Vcap of the 5 psi test was higher than the Vcap of the 10 psi test.
Pressure response may be different for different membrane types
In the case above, the capacity for the PVDF membrane was found to be optimal at 5 psi. The test was also performed on a 47 mm Supor EKV membrane with sodium phosphate In this case, the capacity was greater for 10 psi than for 5 psi pressure. In terms of the membrane, the PES Supor EKV filter had greater capacity than the PVDF membrane.
Different buffers may have different filterability
Table 3 contains a variety of buffers that we subjected to testing with 47 mm Supor EKV discs at a pressure of 15 psi The results are presented as the approximate volume that can be filtered in one hour using different filter sizes. The results show almost a 20 fold difference in filter life between the best and worst filtered buffers, emphasizing the importance of small scale testing.
As illustrated in Table 3, having a variety of filter sizes available can help to provide an economical solution. For example, for filtration of 100 liters of a 6 M urea solution, a Pall Kleenpak KA3 capsule filter is available that could be used for this volume. If this filter were not available, than an unnecessarily large filter filer with 4 times the area (10 inch filter or capsule) would have to be used for the application. A variety of filter sizes also make it possible to scale up a process with the same materials of construction.
For filtration of fluids involved in manufacture of pharmaceutical products for human use, the biological safety of the filter cartridge or capsule needs to be demonstrated by the performance of the USP <88> Class VI (121[degrees]C) Plastics Test for Biological Reactivity. Information on biological safety tests and materials of construction for a membrane filter should be obtained from the filter manufacturer.
In addition, effluent quality tests should be performed on each filter lot. Standard tests are listed below.
Cleanliness: Per current USP limits under Particulate Matter in Injections and conformance with requirements for a non-fiber-releasing filter per Title 21 of the U.S. CFR Title 21 Part 211.72 and Part 210.3 (b) (6)
Total Oxidizable Carbon (TOC): Per current USP requirements under Purified Water after flushing.
pH: Per current USP requirements under Purified Water after flushing
Pyrogens: Per current USP requirements under Bacterial Endotoxins Test as determined using the Limulus Ameobocyte Lysate (LAL) reagent with an aliquot from a soak solution.
Filters from each lot are sampled for bacterial challenge to confirm the 0.2 micron rating and all filters in a lot are subjected to a Forward Flow integrity test.
The purpose of the filtration is to remove a particular contaminant or groups of contaminants from the process feed The filter system should not be adding anything significant (extractables) to the process fluid, nor should it remove a desired component from the fluid that is being filtered. In some cases, where low concentrations of proteins or preservatives are in the fluid, there may be a loss in product yield. PES is generally a low adsorptive material. The effluent should be tested to ensure that the amount of loss is acceptable. However, most buffers are water based and do not contain materials in very low concentrations that could be adsorbed by filters,
There are a number of important factors to consider in the selection of a buffer filter:
* The filter should be compatible with all buffers, with low extractables.
* The filter should have appropriate validation data, including biosafety and bacterial challenge data.
* The filter should be robust enough to meet all process requirements. For example, for applications that require SIP, a high capacity for differential pressure can help to avoid process upsets due to an out of control steam cycle.
* A variety of filter formats and sizes should be available to provide a choice of the most economical solution.
* When the sizing of the system, optimal conditions should be used for pressure and flow rate.
* The filter should be easy to wet for accurate integrity testing.
Table 1--Summary of general compatibilities for each membrane type. Membrane Compatible Compatible Hydrophilic Gamma with low with high irradiatable pH fluids pH fluid Nylon 66 NO YES YES YES PVDF YES NO YES YES PES YES YES YES YES PTFE YES YES NO NO Table 2--Water Flow Rates Membrane Water Flow rate Filter area for 10 Water Flow rate for for 47 mm discs inch cartridge (cm2) 10 inch cartridge at 10 psi EKV 190 ml/min 6040 86 L/min 92DP 205 ml/min 4650 68 L/min Table 3 Buffer 30 inch 20 inch 10 inch Kleenpak[TM]KA3 CEX elution buffer 27,000 18,000 9,000 2,000 50 mM Sodium Acetate 25,000 17,000 8,000 2,000 1.0 M Acetic Acid 12,000 8,000 4,000 1,000 HIC load conditioning buffer 2,500 1,700 800 200 6.0 M Urea 1,500 1,000 500 100 Buffer MiniKleenpak[TM] CEX elution buffer 300 50 mM Sodium Acetate 300 1.0 M Acetic Acid 100 HIC load conditioning buffer 30 6.0 M Urea 20 Pressure (psi) Vcap (mL) 5 1667 10 1430 Constant pressure filterability results on 47 mm PVDF membranes Pressure (psi) Vcap (mL) 5 2500 10 3333 Constant pressure filterability results on 47 mm PES Supor EKV
About the Author
Monica Cardona is a Product Manager for Pall Life Sciences, where she has Western Hemisphere responsibility for prefiltration and sterile liquid filtration product lines. Ms. Cardona has been with Pall Corporation for 7 years. She holds a Bachelor of Arts in Biology from Hofstra University and a Master of Science degree in Biology from Adelphi University. e-mail: firstname.lastname@example.org
Copyright 2004. Kleenpak, Supor, SuporLife, Ultipleat are trademarks of Pall Corporation[R] represents a Pall trademark registered in the USA
|Printer friendly Cite/link Email Feedback|
|Date:||Jun 1, 2004|
|Previous Article:||Lab roller compactor for research and testing.|
|Next Article:||High temperature hose with smooth interior maximizes flow.|