This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: WARM/”é: S/vppj
"Vulnru'TS we L . (\7; TECH REVIEW ﬂair/fan . "l/ x" M0173 W‘c L/m‘ 5%
#HZRﬁM/‘qfe Anya wills is inﬂict] Predicting Filter Size Using Vmax Testing By BALA RAGHUNATH,
MICHEL PAILHES, and
THOMAS MISTRETI'A he VmaxTM technique has been used extensively to estimate ﬁlter area require ments for normal ﬂow ﬁl tration (NFF) processes in
biopharmaceutical applications. The
beneﬁts that this technique presents
over conventional ﬂow decay methods
are the speed of testing, reduced vol
ume requirements for evaluation, and
competitive testing of varying ﬁlter
types/sizes — all of which present an
optimized ﬁlter screening strategy and
preliminary estimate of optimized ﬁl
ter size requirements. Filter size or
ﬁltration area requirements derived
using the Vmax technique consist of
contributions from both capacity and
ﬂow—time aspects of the ﬁltration pro
cess. This article examines the relative
contributions of these terms to overall
ﬁlter sizing visavis the ease of ﬂuid
ﬁlterability. Underlying Theory The Vmax technique is based on the
gradual pore plugging model, which
is characterized by gradual. controlled
blocking of the pores with increasing
throughput of ﬂuid (ﬁltrate) through
the ﬁlter. In this model, the particles
generally are considered smaller than
the ﬁlter pore size so that they enter
the complex porous network of the
ﬁlter where they deposit and progres
sively build up on the “walls” of the
ﬁlter medium (pore wall). This deposi
tion continues until the ﬂow channel WI‘TQ L/MI/h". :fkg Figure 1. The gradual pore plugging model used in Vmax technology. or “pore” is completely blocked. The
process is shown in Figure 1. The maximum volume of the ﬂuid
that can pass through the entire ﬁlter
before it is completely blocked is termed
as Vmax. Filters of various media com
positions and pore ratings can be evalu—
ated effectively with a wide range of
process fluids. The above ﬁltration behavior may be
expressed mathematically as: 3 2 _
d I = k[£)2  Equation 1
de dV which may be solved to obtain an inte
grated form of the equation as: t l =_+_._ ‘ — Equation 2 t
V VMA Q,A where Vis volumetric throughput (liters)
over a process time of t (h or min).
Vmax is ﬁlter capacity (L/mz), Q is nor
malized ﬁlter flow rate (LMi—l), and A is the ﬁltration area.
Implementation In a Vmax test, the ﬂuid is ﬁltered
at constant pressure through the ﬁltra
tion device while time of ﬁltration and
volume collected are recorded at regular
intervals. A plot of time/volume vs.
time is made on the yox axis. If the
results plot as a straight line, the gradual
pore plugging model may be considered
to ﬁt the ﬁltration behavior as described
in Equation 2. From Equation 2, it is evident that
Vmax and Q values may be obtained
from the slope and intercept of the plot of
IN vs. I, respectively. It also may be noted
that as the slope of the line becomes more
horizontal, the Vmax value becomes
increasingly high. Conversely, as the slope
of the line becomes more vertical, the
Vmax value becomes increasingly small.
These trends have important implica
tions when determining ﬁltration area. Equation 2 can be rearranged to esti
mate ﬁlter sizing as follows: Bala Raghunuth, Ph.D. ([email protected]) is a bioprocess engineering manager; Michel Pailhes is a technology man
ager; and Thomas Mistretm is a bioprocess engineer; BioPhtirmrrcetrtical Division, Millipore Corporation, Billerica, MA. 38 BioProcessing Journal  Fall 2006
b—J—.—_—__ , Supplied by The British Library  "The world's knowledge" 2d emphasize these differences with refer
a— ence to ﬂuids ﬁltered in biotechnology
1d processes.
ar
(5. Plugging Streams Have
1e Low Vmax Values
.31
:d For plugging streams, the capacity
:d term in Equation 3 (I/Vmax) ishigh com
pared with the flow term [I/(Qf‘tbﬂ.
at This implies that ﬁlter sizing is domi
:d
of
3d Vmax Result as a Function of Slope in
re NonPlugging Streams
es 1200
ac
1000
re
ll. too 1 smmhsropalzgaawgthm
a son
4100
i 200
0
0 0.005 0.010 0.015 0.020
Slope
.1
Figure 2.
,L ,, ,_ Table ‘1. Example of Vmax Sizing for Plugging Streams (Low Vmax Values). Vmax = 60 le2 , Qi = 20 L/m2*min. and Batch Volume = 1000 L Filter Size Requirements Batch Time Filter Area Rguirements 1 hour 4 hours 17.5 m2 (Equation 2)
16.9 m2 (Equation 2) Area based on VNmax = 16.7 m2 (Equation 3) AN = lleax + 1/Qitb — Equation 3 In Equation 3, the contribution to
sizing may be the result of a capac
ity term (1/Vmax) and ﬂowtime term
[ I/( thb )1. It might be useful to under
stand how both the capacity and flow
time affect ﬁlter sizing depending upon
the nature of the ﬂuid and/or process
application. Application Considerations Plugging characteristics of a ﬂuid
stream determine the relative contribu
tion of the capacity and ﬂow for ﬁlter
area estimation. The examples below nated by ﬁlter capacity for the ﬂuid
stream. Therefore Equation 3 can be
simpliﬁed to: A/V = le  Equation 4 The range of typical Vmax values for a
0.22—pm ﬁlter where capacity is dominant
is 50500 IJmZ. Examples of applications
in which this occurs are serum ﬁltration,
centrate ﬁltration, and bioreactor media
ﬁltration (Table 1). These examples also
show that for highly plugging streams,
the processing time has minimal effect
on ﬁlter area requirements. NonPlugging Streams
Have High Vmax Values Conversely, for nonplugging ﬂuids,
the capacity term in Equation 3 (1/
Vmax , will be small compared to the
ﬂow [1/(Qi*tb)]. Therefore, the sizing is
primarily a function of the ﬂow time for
clean ﬂuids. Equation 3 simpliﬁes to: A/V: I/Qitb — Equation 5 5
(ll
5
E
2
E.
'5
§
§
:2 0.025 Figure 3. Supplied by The British Library  "The world's knowledge" Typical Vmax values for a 0.22 pm ﬁl
ter where the ﬂow time is dominant can
be greater than 10,000 le2. Examples
of applications in which this occurs
would be for clean buffer streams, such
as sodium acetate and sodium phos
phate. The limiting case of a clean ﬂuid
may be represented by water for injec
tion (WFI), for which case, Vmax —> 0°
and the capacity contribution (1N max)
to sizing vanishes. It is worth noting that for nonplug
ging streams, the ﬁltration time can
have a signiﬁcant effect on the area
requirements. In the above examples,
increasing the ﬁltration time from one
hour to four hours reduces the filter
area requirements by 70%. Increasing
process time can be an effective way
to reduce surface area requirements
and total operating costs. For example,
if Vmax = 20,000/Lm2, Ql = 20/Lm2
min and batch volume = 1000 L, then,
using Equation 3 at one hour, ﬁlter
area requirement is 0.883 m2; whereas,
using Equation 3 at four hours, ﬁlter
area requirement is 0.258 m2. Using
Equation 5, ﬁlter area requirements are
0.833 m2 and 0.21 ml, respectiver In addition, the slope of t/V vs. t is
nearly horizontal and very sensitive to
scatter in the data set. Therefore, small
changes in the slope can produce large
differences in Vmax values, due to the
relationship between the slope and the
Vmax result. Since the area will be
primarily a function of the ﬂow time,
a large error results if the Vmax value
alone is used in sizing. The Vmax
results as a function of slope are illus
trated in Figure 2, where ﬂow decay is Percent Contribution to Sizing Based on Vmax Value oi: 100 Um! mm r1, = 60 minutes www.bioprocessingjournalcom . Fall 2006 39 measured over ten minutes time.
Intermediate Plugging Streams Most biopharmaceutical appli
cations are intermediate plugging
streams, in which both capacity and
ﬂow rate are important in sizing. For
such ﬂuids, Equation 3 more accurately
describes the area requirements. The
range of typical Vmax values for a
0.22pm ﬁlter for intermediate plug—
ging streams is LOGO—10,000 L/mz.
Examples of applications in which low
ﬂow rates dominate sizing include bio
burden reduction steps, such as before
a column puriﬁcation step or after a depth ﬁltration step. In addition, some
complex buffer and media applications
are within this Vrnax range. Figure 3 also shows the relative con
tribution to sizing from capacity and
ﬂow time for a range of Vmax values.
When conducting actual Vmax tri
als, each application is different. It is
important to observe what is occurring
during the test and understand if the
ﬂuid is plugging or nonplugging. For
most applications, both parameters are
important in determining sizing. Summary Vmax is an effective technique for optimized ﬁlter screening and esti
mating preliminary ﬁlter size require.
ments. It is based on the gradual,
uniform blocking of pores as a func.
tion of the amount of ﬂuid passing
through a ﬁlter and its mathematical
model representation. In the Vmax
sizing model, contributions to the ﬁlter
sizing arise from both the ﬁlter capac_
ity limitations (Vmax) and ﬂowtime
(Qi, r3) considerations. An apprecia—
tion of these effects provides a clearer
'understanding of the ﬁltration process,
as well as an assessment of the relative
importance of operating parameters
such as process time, flow rate, and
pressure. The Williamsburg BioProcessing Foundation The Most Trusted Source ofBioProcess Technology“ w. mum<3. Viral Vectors & Vaccines Conference 131/1 Annual illeering Process Development & Production Issues Fairmont Southampton ‘9 Southampton, Bermuda 9° November 6—8, 2006 Life Sciences Topics Include:  Vector Development  Product Characterization
 Raw Material Screening
 Viral Yield Optimization
 Downstream Processing
 Facility Design  Storage & Stability . Assay & Media Development ° Safety Testing
 QA/QC Issues WWWW Sponsored By:
GE Heulthcore Molecular Medlclne
BioServkes. In: SAFCBiosciences" ‘—‘ Academic Success ' Photo courtcg‘.‘nfllic Bermuda Dept. ('H'ouiim: To Register Now: Visit us at \mnmilbioxom, Callus MEI123.8823, or Email us at infoﬁwilbioxom 40 BioProcessing Journal . Fall 2006 Supplied by The British Library  "The world's knowledge" ...
View
Full Document
 '08
 staff

Click to edit the document details