Abstract image of a woman sitting on books and studying on a laptop

Selection, sizing, and operation of bioprocess filtration trains for optimal performance

MilliporeSigma, profile picture
MilliporeSigma

This document discusses the selection, sizing, and operation of bioprocess filtration trains for optimal performance. It covers background on membrane types used in biopharma processing including ultrafiltration, virus retentive, and microfiltration membranes. It then discusses filtration mechanisms, series filtration, and modeling approaches to predict filter performance and sizing including classical fouling models and combined blocking and adsorption models. Experimental data is presented to validate the predictive capability of the combined blocking and adsorption fouling model for sizing filters in a series configuration.

Healthcare
1 of 40
Downloaded 60 times
1
2
Most read
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Merck KGaA Darmstadt, Germany Sal Giglia January 24, 2019 Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance
2 Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019 The life science business of Merck KGaA, Darmstadt, Germany operates as MilliporeSigma in the U.S. and Canada.
Agenda Background Filtration mechanisms Series filtration Summary Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.20193
Background
Membranes in Biopharma Processing Bacteria Mycoplasma Product Conc. Virus Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.20195
Membrane Types in Biopharma Processing Mycoplasma Product Conc. 3. Microfiltration Membranes 2. Virus Retentive Membranes 1. Ultrafiltration Membranes Prefiltration Protection for final filter Virus Bacteria Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.20196
Sterilizing grade membranes What is required for optimum performance? • High, predictable retention  Sterility assurance • High process flux  Speed of unit operation • High filtration capacity  Economy • High mechanical strength  Robustness at scale • Ease of wetting  Ease of integrity testing • Low extractables  No added contamination • Chemical compatibility  Robust Morphology Chemistry Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.20197
Filtration mechanisms
How sterilizing-membrane filters work Sterilizing-grade filter definition: ASTM® F 838-15 is a standard method against which all sterilizing grade membranes can be compared  Removal of a standard test organism (Brevundimonas diminuta) at minimum concentration of 107 cfu/cm2 Sterilizing-grade filters function primarily on the basis of size exclusion “A filter that reproducibly removes test microorganisms from the process stream, producing a sterile filtrate.” PDA® Technical report N°26, 2008 fluid flow 9
Throughput Flux Pores that prevent passage of bacteria or viruses can also become plugged in streams containing foulants. Observed as decline in flux (constant pressure) or increase in pressure (constant flow) with time. Membrane plugging (fouling) Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201910
Filtration Operating Modes Constant Pressure Constant Flow Models can give insight into fouling mechanisms Theoretical basis for prediction of volume at extended durations  Save time, process volume Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201911
Classical filter fouling models Standard pore constriction Complete pore blocking Intermediate Cake deposit V Pore constriction Pore blocking surface tion ng cept urface surface ion Physical concept Formation of a surface deposit Pore blocking + surface deposit formation of a surface deposit pore blocking + surface deposit 1 2 43 2 0 2 1        VK J J s VK i e J J   0        00 1 J VK J J b 1 1 00   VJKJ J c • Based on Darcy’s law • Uniform particle size • Uniform pore size • Macroscopic description P.H. Hermans and H.L. Bredee, J. Soc. Chem. Ind., 55T:1-4 (1936) J. Hermia Trans. IChemE. 60:183-187 (1982) R PA J  Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201912
Classical filter fouling model limitations Standard pore constriction Complete pore blocking Intermediate Cake deposit V Pore constriction Pore blocking surface tion ng cept urface surface ion Physical concept Formation of a surface deposit Pore blocking + surface deposit formation of a surface deposit pore blocking + surface deposit 1 2 43 2 0 2 1        VK J J s VK i e J J   0        00 1 J VK J J b 1 1 00   VJKJ J c Flux decay is a function of only throughput volume and is independent of time Assumes that only one filtration mechanism is occurring Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201913
0.1 g/l Hy-Soy T, 0.45 m PES membrane Classic filter fouling models Classic models predict that ultimate filter capacity is independent of pressure and velocity True for some streams…. Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 200 400 600 800 1000 Throughput (l/m 2 ) J/Jo 2 psig 10 psig 1g/l EMD Soy, 0.45 m PES membrane Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1000 2000 3000 4000 Throughput (l/m2 ) J/Jo 10.5 psig 5.2 psig 2.2 psig But not always! Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201914
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.01 0.1 1 10 100 1000 NormalizedParticleVolume Particle Diameter (um) Hy-Soy T EMD Soy Filter sizing estimates are affected by stream/filter characteristics Filter pore size distribution 0.45 µm PES Non-classic behavior Many small particles. small enough to stick to pore or particle surface. Classic behavior Most particles large enough to be easily retained by size exclusion. Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201915 (m)
Velocity dependent “Adsorption” Fouling Model Accounting for filtration velocity  4 0 1 tK J J a Particle deposits inside membrane pores depend on fluid velocity Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201916
Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1000 2000 3000 4000 Throughput (l/m2 ) J/Jo 10.5 psig 5.2 psig 2.2 psig Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1000 2000 3000 4000 5000 6000 Throughput (l/m 2 ) J/Jo 10.5 psig 5.2 psig 2.2 psig Adsorption model 1g/l EMD Soy, 0.45 m PES membrane Thin solid lines represent model fit to data Blocking model Adsorption model Adsorption model over predicts velocity sensitivity Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201917
Combined blocking and adsorption model Combined intermediate 1. Glen R. Bolton, Austin W. Boesch, Matthew J. Lazzara, The effects of flow rate on membrane capacity: Development and application of adsorptive membrane fouling models, Journal of Membrane Science, Volume 279, Issues 1-2, 1 August 2006, Pages 625-634  4 0 1 tKe J J a VKi   -adsorption1 factors in both blocking effects and filtration velocity Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201918
Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1000 2000 3000 4000 Throughput (l/m2 ) J/Jo 10.5 psig 5.2 psig 2.2 psig Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1000 2000 3000 4000 5000 6000 Throughput (l/m 2 ) J/Jo 10.5 psig 5.2 psig 2.2 psig Combined model shows superior fit to data 1g/l EMD Soy, 0.45 PES membrane Thin solid lines represent model fit to data Blocking model Adsorption model Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1000 2000 3000 4000 5000 Throughput (l/m2 ) J/Jo 10.5 psig 5.2 psig 2.2 psig Combined model Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201919
Series Filtration
Why series filtration? 1. Redundant filtration – FDA 2004 aseptic guidelines: “Use of redundant filtration should be considered in many cases.” – EMEA 2008: “A second filtration via a further sterilized micro-organism retaining filter… may be advisable.” Used as a back-up filter and typically at 1:1 area ratio Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201921
Why series filtration? 2. Prefilter increases throughput of final filter Throughput Time Final filter only Final filter with prefilter Prefilter Final Filter Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201922
Series filtration design ▪ Designing a series filtration process (for example, a prefilter and final filter) is more complicated ▪ Need to determine optimal prefilter to final filter area ratio ▪ Although the inlet pressure to the prefilter is constant, neither filter in the train is operating at constant pressure due to differing rates of membrane plugging ▪ The performance of one filter in the train can affect the performance of the other filter ▪ Modeling tools can simplify filter sizing Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201926
Test methods for sizing filters in series Pre-filter Final filterPre-filter Final filter P1 P2 • Direct measure of serial filtration throughput • Added predictive flexibility: but may have limited extrapolation accuracy • Independent performance data for each filter • High predictive flexibility: but only if each filter is tested at two different pressures Pre-filter Final filter P1 • Simple direct measure of serial filtration throughput • No knowledge of each filter’s performance: does not inform on capacity or protective value of prefilter Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019 24
Experimental verification of fouling model Experiment • 10 psig constant pressure • 500 LMH constant flow • Test at 5 different area ratios in duplicate • Small particle stream (EMD Soy) that challenges both prefilter and final filter •OptiScale® 25 devices (3.5 cm2) •[Milligard® PES 1.2/0.2 µm nominal membrane]/[Millipore Express® SHF (PES) membrane] •[Milligard® PES 1.2/0.2 µm nominal membrane]/[Durapore® 0.22 µm(PVDF) membrane] Analysis • Obtain fouling constants from 1:1 area ratio test • Use these constants to predict throughput at the other area ratios 0:1 Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019 25
Serial filtration throughput data fitting • Both filters plugging • Good fit of model of each filter • Moderate advantage of high area ratio expected EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Millipore Express® SHF (PES) membrane] 1:1 area ratio Constant Pressure Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019 26
Serial filtration throughput data fitting • Both filters plugging • Good fit of model of each filter • Moderate advantage of high area ratio expected EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Durapore® 0.22 µm (PVDF) membrane] 1:1 area ratio Constant Pressure Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019 27
Model predicts throughput as function of area ratio ▪ Model prediction based on data collected at 1:1 area ratio ▪ Excellent predictive accuracy over realistic range of area ratios Throughput at 30 minutes 0 500 1000 1500 2000 2500 3000 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 Throughput(l/m2finalfilter) Area Ratio (Pre/Final Filter) Model Data Millipore® Express SHF only EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Millipore Express® SHF (PES) membrane] Constant Pressure 0 500 1000 1500 2000 2500 3000 3500 0 20 40 60 80 100 Throughput(l/m2offinalfilter) Time (min) Combined Throughput Milligard PES/SHF SHF Only 0:1 0.5:1 1:1 2:1 3:1Milligard® PES/SHF Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201928
Model predicts throughput as function of area ratio ▪ Model prediction based on data collected at 1:1 area ratio ▪ Good predictive accuracy over realistic range of area ratios Throughput at 60 minutes 0 500 1000 1500 2000 2500 3000 0,0 1,0 2,0 3,0 4,0 Throughput(l/m2finalfilter) Area Ratio (Pre/Final Filter) Model Data Durapore® 0.22 µm only EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Durapore® 0.22 µm (PVDF) membrane] Constant Pressure 0 500 1000 1500 2000 2500 3000 0 20 40 60 80 100 Throughput(l/m2offinalfilter) Time (min) Combined Throughput 0:1 0.5:1 1:1 3:1 2:1 Milligard® PES/Durapore® Durapore® Only Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201929
Economic optimum area ratio • Utilizing cost information for the prefilter and final filter, the economic optimum area ratio can be determined. • Milligard® PES prefilter enables >70% reduction in the cost of filtration. *Cost normalized to final filter only Prefilter/final filter cost ratio = 0.5 Constant Pressure EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Millipore Express® SHF (PES) membrane] Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201930
Serial filtration in constant flow operation • Both filters plugging • Good fit of model of each filter 0 5 10 15 20 25 30 0 50 100 150 200 250 TMP(psid) Time (min) Interstage Pressure Drop Total, Data Prefilter Data Final Filter Data PreFilter Model Final Filter Model Total, Model 0 100 200 300 400 500 600 0 5 10 15 20 25 30 0 500 1000 1500 2000 2500 FinalFilterFlux(LMH) Presuure(psi) Throughput (l/m2 of final filter) Combined Throughput Model Pressure Data Flux Data 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 Permeability(LMH/psi) Throughput (l/m2 of final flter) Interstage Permeability Prefilter Data Final Filter Data Prefilter Model Final Filter Model y = 3.2184x + 0.0889 0 5 10 15 20 25 30 35 0 5 10 15 Volume(ml). Time (min) Water/Buffer Flux EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Durapore® 0.22 µm (PVDF) membrane] 1:1 area ratio Constant Flow 0 5 10 15 20 25 30 0 50 100 150 200 250 TMP(psid) Time (min) Interstage Pressure Drop Total, Data Prefilter Data Final Filter Data PreFilter Model Final Filter Model Total, Model 0 100 200 300 400 500 600 0 5 10 15 20 25 30 0 500 1000 1500 2000 2500 FinalFilterFlux(LMH) Presuure(psi) Throughput (l/m2 of final filter) Combined Throughput Model Pressure Data Flux Data 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 Permeability(LMH/psi) Throughput (l/m2 of final flter) Interstage Permeability Prefilter Data Final Filter Data Prefilter Model Final Filter Model y = 3.2184x + 0.0889 0 5 10 15 20 25 30 35 0 5 10 15 Volume(ml). Time (min) Water/Buffer Flux Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019 31
0 500 1000 1500 2000 2500 3000 3500 4000 0,0 1,0 2,0 3,0 4,0 Throughput(l/m2finalfilter) Area Ratio (Pre/Final Filter) Data Model Durapore® 0.22 µm only * Model predicts throughput as function of area ratio ▪ Model prediction based on data collected at 1:1 area ratio ▪ Good predictive accuracy over realistic range of area ratios Throughput at 20 psid *Test at area ratio of 3 was terminated when TMP was about 10-12 psid. Data is projected to 20 psid. EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Durapore® 0.22 µm (PVDF) membrane] Constant Flow Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201932
0 5 10 15 20 25 30 35 0 500 1000 1500 2000 2500 Pressure(psig) Throughput Volume (l/m2) 100 LMH 250 LMH 1000 LMH 2000 LMH 5000 LMH 500 LMH Determination of optimum flux in constant flow operation LMH = L/m2-hr In constant flow filtration, the filtration endpoint is typically a maximum allowable pressure.  Membrane area requirement depends on flux, and if fouling rate is flux dependent, there will be an optimum (minimum filter area) flux.  At low flux, membrane will suffer increased adsorptive plugging, minimizing the capacity of the filter.  At high flux, the initial pressure drop across the filters will be near the allowable maximum, leaving little room for pressure rise.  An optimum flux will exist that balances the two factors above. Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201933
Experimental verification of model for optimum flux Experiment • 3 area ratios (0:1, 1:1, 2:1) • 4 constant flux conditions (500, 1000, 1500, 2000 LMH) • Small particle stream (soy peptone ~ 200 nm) that challenges both prefilter and final filter • OptiScale® 25 devices (3.5 cm2) • [Milligard® PES 1.2/0.2 µm nominal]/[Durapore® 0.22 µm (PVDF) membrane] 34 Analysis • Obtain combined intermediate- adsorption fouling constants from one area ratio (1:1) and at one flux condition (1000 LMH). • Use these constants to predict throughput at the other area ratios and fluxes. 1000 LMH 1000 LMH 500 LMH1000 LMH 1000 LMH 500 LMH 0:1 1:1 2:1 Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019
Comparison between measured data and model predictions • Good agreement between model and data • Using filter cost data, the economic optimum process conditions (flux and area ratio) can be determined 35 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 0 500 1000 1500 2000 2500 FinalFilterThroughput(l/m2) Final Filter Flux (LMH) Data 2:1 Area Ratio Model 2:1 Area Ratio Data 1:1 Area Ratio Model 1:1 Area Ratio Data 0:1 Area Ratio Model 0:1 Area Ratio 20 psi filtration endpoint Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019
Economic optimum constant flow filtration conditions • Including prefilter upstream of sterile filter reduces cost of filtration by 6X in this example. • Further 25% cost improvement by optimization of flux and area ratio 36 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 Durapore® 0.22 µm Only (0:1 area ratio) 500 LMH, 1:1 Area Ratio [Milligard® PES 1.2/0.2 µm Nominal]/[Durapore® 0.22 µm] 1500 LMH, 1.9:1 Area Ratio (optimum) [Milligard® PES 1.2/0.2 µm Nominal]/[Durapore® 0.22 µm] Normalized Process Cost per Volume Filtered Prefilter/final filter cost ratio = 0.5 Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019
Summary
Summary Understanding of filtration mechanisms is essential for optimal design and operation of membrane filters. Determination of an optimum series filtration process design can be readily achieved using an efficient test methodology along with an appropriate fouling model that accounts for both blocking and adsorptive fouling. Dramatic improvements in sterile filtration economics can be realized with a suitable prefilter installed into an optimally designed filtration train. w/o prefilter w/ prefilter Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201938
Acknowledgments Shannon Cleveland Sherry Ashby-Leon Songhua Liu Trish Greenhalgh Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201939
Questions Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201940 © 2019 Merck KGaA, Darmstadt, Germany and/or its affiliates. All Rights Reserved. The vibrant M, Vmax, Durapore, Milligard, Millipore Express and OptiScale are trademarks of Merck KGaA, Darmstadt, Germany or its affiliates. All other trademarks are the property of their respective owners. Detailed information on trademarks is available via publicly accessible resources.

More Related Content

PDF
Introduction to Tangential Flow Filtration (TFF)
MilliporeSigma
 
PDF
Selection, sizing, and operation of bioprocess filtration trains for optimal ...
Merck Life Sciences
 
PDF
Normal Flow Filtration: Design and Scale Up
MilliporeSigma
 
PDF
Find your filter. What’s best for your process?
Merck Life Sciences
 
PDF
An Efficient and cGMP-friendly Solution to Diafiltration for Intensified or C...
Merck Life Sciences
 
PDF
Single-Use Tangential Flow Filtration for Closed Processing
Merck Life Sciences
 
PDF
Introduction to Tangential Flow Filtration (TFF)
Merck Life Sciences
 
PDF
Find your filter. What’s best for your process?
MilliporeSigma
 
Introduction to Tangential Flow Filtration (TFF)
MilliporeSigma
 
Selection, sizing, and operation of bioprocess filtration trains for optimal ...
Merck Life Sciences
 
Normal Flow Filtration: Design and Scale Up
MilliporeSigma
 
Find your filter. What’s best for your process?
Merck Life Sciences
 
An Efficient and cGMP-friendly Solution to Diafiltration for Intensified or C...
Merck Life Sciences
 
Single-Use Tangential Flow Filtration for Closed Processing
Merck Life Sciences
 
Introduction to Tangential Flow Filtration (TFF)
Merck Life Sciences
 
Find your filter. What’s best for your process?
MilliporeSigma
 

What's hot (20)

PDF
Normal Flow Filtration: Design and Scale Up
Merck Life Sciences
 
PDF
Validation of Tangential Flow Filtration in Biotech Processes
Merck Life Sciences
 
PPTX
66 membrane filter integrity test
Kethana Nellore
 
PDF
Optimization of Tangential Flow Filtration Applications and Scale Up Consider...
Merck Life Sciences
 
PDF
Integrity Test Troubleshooting – Beyond Rewet and Retest
Merck Life Sciences
 
PPTX
Bioburden control: Strategies to address bioburden control in downstream proc...
Merck Life Sciences
 
PDF
Sterile filtration of complex injectables by Partha Banerjee
Merck Life Sciences
 
PDF
QBD for Downstream Virus Filtration
MilliporeSigma
 
PDF
Parvovirus Filtration Best Practices - 25 Years of Hands-On Experience
Merck Life Sciences
 
PDF
Process Scale Chromatography
Merck Life Sciences
 
PDF
Addressing Downstream Challenges with Complex Injectables
MilliporeSigma
 
PPT
Filter Validation and Regulatory aspects
bala murugan
 
PDF
Single-Pass Tangential Flow Filtration (SPTFF) Theory and Practice
Merck Life Sciences
 
PDF
Membrane Chromatography Solutions for Single-Use, Intensified mAb Purification
Merck Life Sciences
 
PDF
Compressed Timelines for Breakthrough Therapies: Impact on Process Characteri...
KBI Biopharma
 
PPTX
Particle Technology- Filtration
The Engineering Centre for Excellence in Teaching and Learning
 
PPTX
IVT Presentation Batch vs Continuous - 45min_REV3
Eric Sipe
 
PPT
Rheology (1)
specialimroz
 
PPTX
regulatory requirement for validation in pharma industry
Synchron Research Services Pvt. Ltd.
 
PPTX
Water for Pharmaceutical Purposes - New USP Requriements
GMP EDUCATION : Not for Profit Organization
 
Normal Flow Filtration: Design and Scale Up
Merck Life Sciences
 
Validation of Tangential Flow Filtration in Biotech Processes
Merck Life Sciences
 
66 membrane filter integrity test
Kethana Nellore
 
Optimization of Tangential Flow Filtration Applications and Scale Up Consider...
Merck Life Sciences
 
Integrity Test Troubleshooting – Beyond Rewet and Retest
Merck Life Sciences
 
Bioburden control: Strategies to address bioburden control in downstream proc...
Merck Life Sciences
 
Sterile filtration of complex injectables by Partha Banerjee
Merck Life Sciences
 
QBD for Downstream Virus Filtration
MilliporeSigma
 
Parvovirus Filtration Best Practices - 25 Years of Hands-On Experience
Merck Life Sciences
 
Process Scale Chromatography
Merck Life Sciences
 
Addressing Downstream Challenges with Complex Injectables
MilliporeSigma
 
Filter Validation and Regulatory aspects
bala murugan
 
Single-Pass Tangential Flow Filtration (SPTFF) Theory and Practice
Merck Life Sciences
 
Membrane Chromatography Solutions for Single-Use, Intensified mAb Purification
Merck Life Sciences
 
Compressed Timelines for Breakthrough Therapies: Impact on Process Characteri...
KBI Biopharma
 
Particle Technology- Filtration
The Engineering Centre for Excellence in Teaching and Learning
 
IVT Presentation Batch vs Continuous - 45min_REV3
Eric Sipe
 
Rheology (1)
specialimroz
 
regulatory requirement for validation in pharma industry
Synchron Research Services Pvt. Ltd.
 
Water for Pharmaceutical Purposes - New USP Requriements
GMP EDUCATION : Not for Profit Organization
 
Ad

Similar to Selection, sizing, and operation of bioprocess filtration trains for optimal performance (20)

PDF
03.pdf
Carki1
 
PPTX
Presentation1 tff
chauhan suraj
 
PPTX
Designing of mechanical process such as separation process of filtration.pptx
salehin1907047
 
PPTX
Filtration
Tapeshwar Yadav
 
PDF
Cell Culture Media Filtration – Filter Selection and Sizing
MilliporeSigma
 
PPT
Separation-Customer-Training_3-2007 Malaysia.
KCLee53
 
PDF
Optimization of Tangential Flow Filtration Applications and Scale Up Consider...
MilliporeSigma
 
PPTX
Membrane based water purification technology(ultra filteration,dialysis and e...
Sanjeev Singh
 
PPTX
011 PP Foundation Training TFF- new brand EU.pptx
MuhammadMedhat8
 
PDF
Filtration Strategies for Optimal Development and Purification of a FMD Virus...
Merck Life Sciences
 
PDF
Filtration Strategies for Optimal Development and Purification of a FMD Virus...
MilliporeSigma
 
PDF
Filteration selection guide
Arash Abed
 
PPTX
Membrane filtration by Akram Hossain, Food and Process Engineering, HSTU
Akram Hossain
 
PPTX
Nipun
Nipun
 
PPTX
FILTRATION AND TYPES OF FILTRATION . pptx
VISHALI SELVAM
 
PDF
Pharmaceutical Engineering: Filtration
Parag Jain
 
PPT
Fitration
kavithaaut
 
PDF
An overview on_biofiltration_modeling
Avi Kumar
 
PPTX
Membrane Processing Technology in dairy industry
shrustimanjunath78
 
PDF
Parvovirus Filtration Best Practices - 25 Years of Hands-On Experience
MilliporeSigma
 
03.pdf
Carki1
 
Presentation1 tff
chauhan suraj
 
Designing of mechanical process such as separation process of filtration.pptx
salehin1907047
 
Filtration
Tapeshwar Yadav
 
Cell Culture Media Filtration – Filter Selection and Sizing
MilliporeSigma
 
Separation-Customer-Training_3-2007 Malaysia.
KCLee53
 
Optimization of Tangential Flow Filtration Applications and Scale Up Consider...
MilliporeSigma
 
Membrane based water purification technology(ultra filteration,dialysis and e...
Sanjeev Singh
 
011 PP Foundation Training TFF- new brand EU.pptx
MuhammadMedhat8
 
Filtration Strategies for Optimal Development and Purification of a FMD Virus...
Merck Life Sciences
 
Filtration Strategies for Optimal Development and Purification of a FMD Virus...
MilliporeSigma
 
Filteration selection guide
Arash Abed
 
Membrane filtration by Akram Hossain, Food and Process Engineering, HSTU
Akram Hossain
 
Nipun
Nipun
 
FILTRATION AND TYPES OF FILTRATION . pptx
VISHALI SELVAM
 
Pharmaceutical Engineering: Filtration
Parag Jain
 
Fitration
kavithaaut
 
An overview on_biofiltration_modeling
Avi Kumar
 
Membrane Processing Technology in dairy industry
shrustimanjunath78
 
Parvovirus Filtration Best Practices - 25 Years of Hands-On Experience
MilliporeSigma
 
Ad

More from MilliporeSigma (20)

PDF
Launch of our new Titanium Dioxide Alternative
MilliporeSigma
 
PDF
The Viscosity Reduction Platform: Viscosity-reducing excipients for improveme...
MilliporeSigma
 
PDF
Use of Excipients in Downstream Processing to Improve Protein Purification
MilliporeSigma
 
PDF
Exploring the protein stabilizing capability of surfactants against agitation...
MilliporeSigma
 
PDF
The Viscosity Reduction Platform: Viscosity Reducing Excipients for Protein F...
MilliporeSigma
 
PDF
Characterization of monoclonal antibodies and Antibody drug conjugates by Sur...
MilliporeSigma
 
PDF
The Role of BioPhorum Extractables Data in the Effective Adoption of Single-U...
MilliporeSigma
 
PDF
The Future of Pharma- and Biopharmaceutical Audits
MilliporeSigma
 
PDF
Moving your Gene Therapy from R&D to IND: How to navigate the Regulatory Land...
MilliporeSigma
 
PDF
Identity testing by NGS as a means of risk mitigation for viral gene therapies
MilliporeSigma
 
PDF
Latest advancements of melt based 3D printing technologies for oral drug deli...
MilliporeSigma
 
PDF
CAR-T Manufacturing Innovations that Work - Automating Low Volume Processes a...
MilliporeSigma
 
PDF
How does the ICH Q5A revision impact viral safety strategies for biologics?
MilliporeSigma
 
PDF
Improve Operational Efficiency by Over 30% with Product, Process, & Systems A...
MilliporeSigma
 
PDF
Insights from a Global Collaboration Accelerating Vaccine Development with an...
MilliporeSigma
 
PDF
Risk-Based Qualification of X-Ray Sterilization for Single-Use Systems
MilliporeSigma
 
PDF
Rapid Replication Competent Adenovirus (rRCA) Detection: Accelerate your Lot ...
MilliporeSigma
 
PDF
The High Intensity Sweeteners Neotame and Sucralose: 2 Ways to ace the Patien...
MilliporeSigma
 
PDF
The Developability Classification System (DCS): Enabling an Optimized Approac...
MilliporeSigma
 
PDF
How to Accelerate and Enhance ADC Therapies
MilliporeSigma
 
Launch of our new Titanium Dioxide Alternative
MilliporeSigma
 
The Viscosity Reduction Platform: Viscosity-reducing excipients for improveme...
MilliporeSigma
 
Use of Excipients in Downstream Processing to Improve Protein Purification
MilliporeSigma
 
Exploring the protein stabilizing capability of surfactants against agitation...
MilliporeSigma
 
The Viscosity Reduction Platform: Viscosity Reducing Excipients for Protein F...
MilliporeSigma
 
Characterization of monoclonal antibodies and Antibody drug conjugates by Sur...
MilliporeSigma
 
The Role of BioPhorum Extractables Data in the Effective Adoption of Single-U...
MilliporeSigma
 
The Future of Pharma- and Biopharmaceutical Audits
MilliporeSigma
 
Moving your Gene Therapy from R&D to IND: How to navigate the Regulatory Land...
MilliporeSigma
 
Identity testing by NGS as a means of risk mitigation for viral gene therapies
MilliporeSigma
 
Latest advancements of melt based 3D printing technologies for oral drug deli...
MilliporeSigma
 
CAR-T Manufacturing Innovations that Work - Automating Low Volume Processes a...
MilliporeSigma
 
How does the ICH Q5A revision impact viral safety strategies for biologics?
MilliporeSigma
 
Improve Operational Efficiency by Over 30% with Product, Process, & Systems A...
MilliporeSigma
 
Insights from a Global Collaboration Accelerating Vaccine Development with an...
MilliporeSigma
 
Risk-Based Qualification of X-Ray Sterilization for Single-Use Systems
MilliporeSigma
 
Rapid Replication Competent Adenovirus (rRCA) Detection: Accelerate your Lot ...
MilliporeSigma
 
The High Intensity Sweeteners Neotame and Sucralose: 2 Ways to ace the Patien...
MilliporeSigma
 
The Developability Classification System (DCS): Enabling an Optimized Approac...
MilliporeSigma
 
How to Accelerate and Enhance ADC Therapies
MilliporeSigma
 

Recently uploaded (20)

PPTX
osteoporosis in menopause...............
Vikram Khanna
 
PPTX
ANALGESIC AND ANTI-INFLAMMssssssATORY DRUGS.pptx
khaalidmohamed6
 
PDF
Back node with known primary managementt
LokeshDahiya11
 
PPTX
Full Slide Deck - SY CF Talk Adelaide 10June.pptx
sarahychurchillfello
 
PPTX
Hospital Services healthcare management in india
BUBBLESFlair
 
PPTX
ACUTE CALCULAR CHOLECYSTITIS: A CASE STUDY
OmarShurafa
 
PPTX
Tracheostomy Care: A Comprehensive Guide
chiiicherryy17
 
PPTX
1. FAMILY PLANNING-1-2, nursing students
lizmukade
 
PDF
cerebral aneurysm.. neurosurgery , anaesthesia
cheryljyoti
 
PPTX
Fever and skin rash - Approach.pptxBy Dr Gururaja R , Paediatrician. An usef...
GururajaRamaiah1
 
PPTX
Acute renal failure.pptx for BNs 2nd year
neelamjoshi0922
 
DOCX
PT10 continues to explose your mind right after reading
LinhL765970
 
PDF
ENT MedMap you can study for the exam with this.pdf
Azhar856819
 
PPTX
Carcinoma of the breastfgdvfgbddbdtr.pptx
pierresemeko1989
 
PDF
mycobacterial infection tuberculosis (TB)
TheophilusBaidoo3
 
PPT
12.08.2025 Dr. Amrita Ghosh_Stocks Standards_ Smart_Inventory Management_GCLP...
Dr. Debjyoti Halder
 
PPT
heartap-240428112119-ec76d6fb.pp for studentt
anup222212
 
PPTX
MEDICAL NURSING. Endocrine Disorder.pptx
MutegekiAdolf1
 
PDF
health promotion and maintenance of elderly
prasana8
 
PPTX
Mortality rate in a teritiary care center of infia july stats sncu picu
Ramkumar84712
 
osteoporosis in menopause...............
Vikram Khanna
 
ANALGESIC AND ANTI-INFLAMMssssssATORY DRUGS.pptx
khaalidmohamed6
 
Back node with known primary managementt
LokeshDahiya11
 
Full Slide Deck - SY CF Talk Adelaide 10June.pptx
sarahychurchillfello
 
Hospital Services healthcare management in india
BUBBLESFlair
 
ACUTE CALCULAR CHOLECYSTITIS: A CASE STUDY
OmarShurafa
 
Tracheostomy Care: A Comprehensive Guide
chiiicherryy17
 
1. FAMILY PLANNING-1-2, nursing students
lizmukade
 
cerebral aneurysm.. neurosurgery , anaesthesia
cheryljyoti
 
Fever and skin rash - Approach.pptxBy Dr Gururaja R , Paediatrician. An usef...
GururajaRamaiah1
 
Acute renal failure.pptx for BNs 2nd year
neelamjoshi0922
 
PT10 continues to explose your mind right after reading
LinhL765970
 
ENT MedMap you can study for the exam with this.pdf
Azhar856819
 
Carcinoma of the breastfgdvfgbddbdtr.pptx
pierresemeko1989
 
mycobacterial infection tuberculosis (TB)
TheophilusBaidoo3
 
12.08.2025 Dr. Amrita Ghosh_Stocks Standards_ Smart_Inventory Management_GCLP...
Dr. Debjyoti Halder
 
heartap-240428112119-ec76d6fb.pp for studentt
anup222212
 
MEDICAL NURSING. Endocrine Disorder.pptx
MutegekiAdolf1
 
health promotion and maintenance of elderly
prasana8
 
Mortality rate in a teritiary care center of infia july stats sncu picu
Ramkumar84712
 

Selection, sizing, and operation of bioprocess filtration trains for optimal performance

  • 1. Merck KGaA Darmstadt, Germany Sal Giglia January 24, 2019 Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance
  • 2. 2 Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019 The life science business of Merck KGaA, Darmstadt, Germany operates as MilliporeSigma in the U.S. and Canada.
  • 3. Agenda Background Filtration mechanisms Series filtration Summary Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.20193
  • 5. Membranes in Biopharma Processing Bacteria Mycoplasma Product Conc. Virus Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.20195
  • 6. Membrane Types in Biopharma Processing Mycoplasma Product Conc. 3. Microfiltration Membranes 2. Virus Retentive Membranes 1. Ultrafiltration Membranes Prefiltration Protection for final filter Virus Bacteria Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.20196
  • 7. Sterilizing grade membranes What is required for optimum performance? • High, predictable retention  Sterility assurance • High process flux  Speed of unit operation • High filtration capacity  Economy • High mechanical strength  Robustness at scale • Ease of wetting  Ease of integrity testing • Low extractables  No added contamination • Chemical compatibility  Robust Morphology Chemistry Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.20197
  • 9. How sterilizing-membrane filters work Sterilizing-grade filter definition: ASTM® F 838-15 is a standard method against which all sterilizing grade membranes can be compared  Removal of a standard test organism (Brevundimonas diminuta) at minimum concentration of 107 cfu/cm2 Sterilizing-grade filters function primarily on the basis of size exclusion “A filter that reproducibly removes test microorganisms from the process stream, producing a sterile filtrate.” PDA® Technical report N°26, 2008 fluid flow 9
  • 10. Throughput Flux Pores that prevent passage of bacteria or viruses can also become plugged in streams containing foulants. Observed as decline in flux (constant pressure) or increase in pressure (constant flow) with time. Membrane plugging (fouling) Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201910
  • 11. Filtration Operating Modes Constant Pressure Constant Flow Models can give insight into fouling mechanisms Theoretical basis for prediction of volume at extended durations  Save time, process volume Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201911
  • 12. Classical filter fouling models Standard pore constriction Complete pore blocking Intermediate Cake deposit V Pore constriction Pore blocking surface tion ng cept urface surface ion Physical concept Formation of a surface deposit Pore blocking + surface deposit formation of a surface deposit pore blocking + surface deposit 1 2 43 2 0 2 1        VK J J s VK i e J J   0        00 1 J VK J J b 1 1 00   VJKJ J c • Based on Darcy’s law • Uniform particle size • Uniform pore size • Macroscopic description P.H. Hermans and H.L. Bredee, J. Soc. Chem. Ind., 55T:1-4 (1936) J. Hermia Trans. IChemE. 60:183-187 (1982) R PA J  Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201912
  • 13. Classical filter fouling model limitations Standard pore constriction Complete pore blocking Intermediate Cake deposit V Pore constriction Pore blocking surface tion ng cept urface surface ion Physical concept Formation of a surface deposit Pore blocking + surface deposit formation of a surface deposit pore blocking + surface deposit 1 2 43 2 0 2 1        VK J J s VK i e J J   0        00 1 J VK J J b 1 1 00   VJKJ J c Flux decay is a function of only throughput volume and is independent of time Assumes that only one filtration mechanism is occurring Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201913
  • 14. 0.1 g/l Hy-Soy T, 0.45 m PES membrane Classic filter fouling models Classic models predict that ultimate filter capacity is independent of pressure and velocity True for some streams…. Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 200 400 600 800 1000 Throughput (l/m 2 ) J/Jo 2 psig 10 psig 1g/l EMD Soy, 0.45 m PES membrane Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1000 2000 3000 4000 Throughput (l/m2 ) J/Jo 10.5 psig 5.2 psig 2.2 psig But not always! Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201914
  • 15. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.01 0.1 1 10 100 1000 NormalizedParticleVolume Particle Diameter (um) Hy-Soy T EMD Soy Filter sizing estimates are affected by stream/filter characteristics Filter pore size distribution 0.45 µm PES Non-classic behavior Many small particles. small enough to stick to pore or particle surface. Classic behavior Most particles large enough to be easily retained by size exclusion. Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201915 (m)
  • 16. Velocity dependent “Adsorption” Fouling Model Accounting for filtration velocity  4 0 1 tK J J a Particle deposits inside membrane pores depend on fluid velocity Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201916
  • 17. Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1000 2000 3000 4000 Throughput (l/m2 ) J/Jo 10.5 psig 5.2 psig 2.2 psig Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1000 2000 3000 4000 5000 6000 Throughput (l/m 2 ) J/Jo 10.5 psig 5.2 psig 2.2 psig Adsorption model 1g/l EMD Soy, 0.45 m PES membrane Thin solid lines represent model fit to data Blocking model Adsorption model Adsorption model over predicts velocity sensitivity Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201917
  • 18. Combined blocking and adsorption model Combined intermediate 1. Glen R. Bolton, Austin W. Boesch, Matthew J. Lazzara, The effects of flow rate on membrane capacity: Development and application of adsorptive membrane fouling models, Journal of Membrane Science, Volume 279, Issues 1-2, 1 August 2006, Pages 625-634  4 0 1 tKe J J a VKi   -adsorption1 factors in both blocking effects and filtration velocity Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201918
  • 19. Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1000 2000 3000 4000 Throughput (l/m2 ) J/Jo 10.5 psig 5.2 psig 2.2 psig Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1000 2000 3000 4000 5000 6000 Throughput (l/m 2 ) J/Jo 10.5 psig 5.2 psig 2.2 psig Combined model shows superior fit to data 1g/l EMD Soy, 0.45 PES membrane Thin solid lines represent model fit to data Blocking model Adsorption model Flux Decay 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1000 2000 3000 4000 5000 Throughput (l/m2 ) J/Jo 10.5 psig 5.2 psig 2.2 psig Combined model Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201919
  • 21. Why series filtration? 1. Redundant filtration – FDA 2004 aseptic guidelines: “Use of redundant filtration should be considered in many cases.” – EMEA 2008: “A second filtration via a further sterilized micro-organism retaining filter… may be advisable.” Used as a back-up filter and typically at 1:1 area ratio Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201921
  • 22. Why series filtration? 2. Prefilter increases throughput of final filter Throughput Time Final filter only Final filter with prefilter Prefilter Final Filter Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201922
  • 23. Series filtration design ▪ Designing a series filtration process (for example, a prefilter and final filter) is more complicated ▪ Need to determine optimal prefilter to final filter area ratio ▪ Although the inlet pressure to the prefilter is constant, neither filter in the train is operating at constant pressure due to differing rates of membrane plugging ▪ The performance of one filter in the train can affect the performance of the other filter ▪ Modeling tools can simplify filter sizing Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201926
  • 24. Test methods for sizing filters in series Pre-filter Final filterPre-filter Final filter P1 P2 • Direct measure of serial filtration throughput • Added predictive flexibility: but may have limited extrapolation accuracy • Independent performance data for each filter • High predictive flexibility: but only if each filter is tested at two different pressures Pre-filter Final filter P1 • Simple direct measure of serial filtration throughput • No knowledge of each filter’s performance: does not inform on capacity or protective value of prefilter Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019 24
  • 25. Experimental verification of fouling model Experiment • 10 psig constant pressure • 500 LMH constant flow • Test at 5 different area ratios in duplicate • Small particle stream (EMD Soy) that challenges both prefilter and final filter •OptiScale® 25 devices (3.5 cm2) •[Milligard® PES 1.2/0.2 µm nominal membrane]/[Millipore Express® SHF (PES) membrane] •[Milligard® PES 1.2/0.2 µm nominal membrane]/[Durapore® 0.22 µm(PVDF) membrane] Analysis • Obtain fouling constants from 1:1 area ratio test • Use these constants to predict throughput at the other area ratios 0:1 Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019 25
  • 26. Serial filtration throughput data fitting • Both filters plugging • Good fit of model of each filter • Moderate advantage of high area ratio expected EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Millipore Express® SHF (PES) membrane] 1:1 area ratio Constant Pressure Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019 26
  • 27. Serial filtration throughput data fitting • Both filters plugging • Good fit of model of each filter • Moderate advantage of high area ratio expected EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Durapore® 0.22 µm (PVDF) membrane] 1:1 area ratio Constant Pressure Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019 27
  • 28. Model predicts throughput as function of area ratio ▪ Model prediction based on data collected at 1:1 area ratio ▪ Excellent predictive accuracy over realistic range of area ratios Throughput at 30 minutes 0 500 1000 1500 2000 2500 3000 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 Throughput(l/m2finalfilter) Area Ratio (Pre/Final Filter) Model Data Millipore® Express SHF only EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Millipore Express® SHF (PES) membrane] Constant Pressure 0 500 1000 1500 2000 2500 3000 3500 0 20 40 60 80 100 Throughput(l/m2offinalfilter) Time (min) Combined Throughput Milligard PES/SHF SHF Only 0:1 0.5:1 1:1 2:1 3:1Milligard® PES/SHF Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201928
  • 29. Model predicts throughput as function of area ratio ▪ Model prediction based on data collected at 1:1 area ratio ▪ Good predictive accuracy over realistic range of area ratios Throughput at 60 minutes 0 500 1000 1500 2000 2500 3000 0,0 1,0 2,0 3,0 4,0 Throughput(l/m2finalfilter) Area Ratio (Pre/Final Filter) Model Data Durapore® 0.22 µm only EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Durapore® 0.22 µm (PVDF) membrane] Constant Pressure 0 500 1000 1500 2000 2500 3000 0 20 40 60 80 100 Throughput(l/m2offinalfilter) Time (min) Combined Throughput 0:1 0.5:1 1:1 3:1 2:1 Milligard® PES/Durapore® Durapore® Only Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201929
  • 30. Economic optimum area ratio • Utilizing cost information for the prefilter and final filter, the economic optimum area ratio can be determined. • Milligard® PES prefilter enables >70% reduction in the cost of filtration. *Cost normalized to final filter only Prefilter/final filter cost ratio = 0.5 Constant Pressure EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Millipore Express® SHF (PES) membrane] Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201930
  • 31. Serial filtration in constant flow operation • Both filters plugging • Good fit of model of each filter 0 5 10 15 20 25 30 0 50 100 150 200 250 TMP(psid) Time (min) Interstage Pressure Drop Total, Data Prefilter Data Final Filter Data PreFilter Model Final Filter Model Total, Model 0 100 200 300 400 500 600 0 5 10 15 20 25 30 0 500 1000 1500 2000 2500 FinalFilterFlux(LMH) Presuure(psi) Throughput (l/m2 of final filter) Combined Throughput Model Pressure Data Flux Data 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 Permeability(LMH/psi) Throughput (l/m2 of final flter) Interstage Permeability Prefilter Data Final Filter Data Prefilter Model Final Filter Model y = 3.2184x + 0.0889 0 5 10 15 20 25 30 35 0 5 10 15 Volume(ml). Time (min) Water/Buffer Flux EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Durapore® 0.22 µm (PVDF) membrane] 1:1 area ratio Constant Flow 0 5 10 15 20 25 30 0 50 100 150 200 250 TMP(psid) Time (min) Interstage Pressure Drop Total, Data Prefilter Data Final Filter Data PreFilter Model Final Filter Model Total, Model 0 100 200 300 400 500 600 0 5 10 15 20 25 30 0 500 1000 1500 2000 2500 FinalFilterFlux(LMH) Presuure(psi) Throughput (l/m2 of final filter) Combined Throughput Model Pressure Data Flux Data 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 Permeability(LMH/psi) Throughput (l/m2 of final flter) Interstage Permeability Prefilter Data Final Filter Data Prefilter Model Final Filter Model y = 3.2184x + 0.0889 0 5 10 15 20 25 30 35 0 5 10 15 Volume(ml). Time (min) Water/Buffer Flux Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019 31
  • 32. 0 500 1000 1500 2000 2500 3000 3500 4000 0,0 1,0 2,0 3,0 4,0 Throughput(l/m2finalfilter) Area Ratio (Pre/Final Filter) Data Model Durapore® 0.22 µm only * Model predicts throughput as function of area ratio ▪ Model prediction based on data collected at 1:1 area ratio ▪ Good predictive accuracy over realistic range of area ratios Throughput at 20 psid *Test at area ratio of 3 was terminated when TMP was about 10-12 psid. Data is projected to 20 psid. EMD Soy [Milligard® PES 1.2/0.2 µm nominal membrane] /[Durapore® 0.22 µm (PVDF) membrane] Constant Flow Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201932
  • 33. 0 5 10 15 20 25 30 35 0 500 1000 1500 2000 2500 Pressure(psig) Throughput Volume (l/m2) 100 LMH 250 LMH 1000 LMH 2000 LMH 5000 LMH 500 LMH Determination of optimum flux in constant flow operation LMH = L/m2-hr In constant flow filtration, the filtration endpoint is typically a maximum allowable pressure.  Membrane area requirement depends on flux, and if fouling rate is flux dependent, there will be an optimum (minimum filter area) flux.  At low flux, membrane will suffer increased adsorptive plugging, minimizing the capacity of the filter.  At high flux, the initial pressure drop across the filters will be near the allowable maximum, leaving little room for pressure rise.  An optimum flux will exist that balances the two factors above. Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201933
  • 34. Experimental verification of model for optimum flux Experiment • 3 area ratios (0:1, 1:1, 2:1) • 4 constant flux conditions (500, 1000, 1500, 2000 LMH) • Small particle stream (soy peptone ~ 200 nm) that challenges both prefilter and final filter • OptiScale® 25 devices (3.5 cm2) • [Milligard® PES 1.2/0.2 µm nominal]/[Durapore® 0.22 µm (PVDF) membrane] 34 Analysis • Obtain combined intermediate- adsorption fouling constants from one area ratio (1:1) and at one flux condition (1000 LMH). • Use these constants to predict throughput at the other area ratios and fluxes. 1000 LMH 1000 LMH 500 LMH1000 LMH 1000 LMH 500 LMH 0:1 1:1 2:1 Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019
  • 35. Comparison between measured data and model predictions • Good agreement between model and data • Using filter cost data, the economic optimum process conditions (flux and area ratio) can be determined 35 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 0 500 1000 1500 2000 2500 FinalFilterThroughput(l/m2) Final Filter Flux (LMH) Data 2:1 Area Ratio Model 2:1 Area Ratio Data 1:1 Area Ratio Model 1:1 Area Ratio Data 0:1 Area Ratio Model 0:1 Area Ratio 20 psi filtration endpoint Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019
  • 36. Economic optimum constant flow filtration conditions • Including prefilter upstream of sterile filter reduces cost of filtration by 6X in this example. • Further 25% cost improvement by optimization of flux and area ratio 36 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 Durapore® 0.22 µm Only (0:1 area ratio) 500 LMH, 1:1 Area Ratio [Milligard® PES 1.2/0.2 µm Nominal]/[Durapore® 0.22 µm] 1500 LMH, 1.9:1 Area Ratio (optimum) [Milligard® PES 1.2/0.2 µm Nominal]/[Durapore® 0.22 µm] Normalized Process Cost per Volume Filtered Prefilter/final filter cost ratio = 0.5 Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.2019
  • 38. Summary Understanding of filtration mechanisms is essential for optimal design and operation of membrane filters. Determination of an optimum series filtration process design can be readily achieved using an efficient test methodology along with an appropriate fouling model that accounts for both blocking and adsorptive fouling. Dramatic improvements in sterile filtration economics can be realized with a suitable prefilter installed into an optimally designed filtration train. w/o prefilter w/ prefilter Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201938
  • 39. Acknowledgments Shannon Cleveland Sherry Ashby-Leon Songhua Liu Trish Greenhalgh Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201939
  • 40. Questions Selection, Sizing, and Operation of Bioprocess Filtration Trains for Optimal Performance | 24.01.201940 © 2019 Merck KGaA, Darmstadt, Germany and/or its affiliates. All Rights Reserved. The vibrant M, Vmax, Durapore, Milligard, Millipore Express and OptiScale are trademarks of Merck KGaA, Darmstadt, Germany or its affiliates. All other trademarks are the property of their respective owners. Detailed information on trademarks is available via publicly accessible resources.