Fouling of an ultrafiltration membrane by a dissolved whey protein concentrate and some whey proteins


Nilsson, J.L.

Journal of Membrane Science 36: 147-160

1988


The influence of adsorption on the pure water flux through a polysulphone membrane was determined using whey protein concentrate (LACPRODAN 80), beta -lactoglobulin and alpha -lactalbumin as solutes. Flux decreases as concentration of the solution in contact with the membrane increases; there is also a dependence on pH and buffer. The adsorption process followed the same pattern for the different proteins. The amount of beta -lactoglobulin adsorbed on the surface was determined using 14C-labelled protein. Flux of the commercial membrane sheets which were used varied between 0.007 and 0.220 g cm-2 min-1.

Journal
of
Membrane
Science,
36
(1988)
147-160
147
Elsevier
Science
Publishers
B.V.,
Amsterdam
Printed
in
The
Netherlands
FOULING
OF
AN
ULTRAFILTRATION
MEMBRANE
BY
A
DISSOLVED
WHEY
PROTEIN
CONCENTRATE
AND
SOME
WHEY
PROTEINS*
JAN
L.
NILSSON
Department
of
Food
Engineering,
Lund
University,
P.O.
Box
124,
S-221
01
Lund
(
Sweden)
(Received
September
4,
1986;
accepted
December
17,
1986)
Summary
The
influence
of
adsorption
on
the
pure
water
flux
through
a
polysulphone
membrane
has
been
determined
for
a
variety
of
conditions.
Whey
protein
concentrate
(
LACPRODAN
80),
fl-lactoglob-
ulin
and
a-lactalbumin
were
used
as
solutes.
The
flux
decreases
as
the
concentration
of
the
solu-
tion
in
contact
with
the
membrane
increases;
there
is
also
a
dependence
on
pH
and
buffer.
The
amount
of
/3-lactoglobulin
adsorbed
on
the
surface
was
determined
using
"C-radiotagged
protein.
The
flux
of
the
commercial
membrane
sheets
which
were
used
varied
within
the
interval
of
0.007
to
0.220
g/cm
2
-min.
Introduction
Ultrafiltration
is defined
as
"a
pressure
driven
operation
in
which
the
mem-
brane
fractionates
components
of
a
liquid
predominantly
according
to
their
size
and
shape.
It
is
used
for
the
separation
of
particles
with
a
molecular
weight
of
higher
than
500
daltons"
[
1
]
.
A
multitude
of
factors
influences
the
OF
performance
in
terms
of
flux
and
rejection
characteristics.
Fouling,
or
adsorp-
tion,
on
the
membrane
surface
and
in
the
pores
is
such
a
factor.
Reports
on
the
role
of
adsorption
in
membrane
processing
can
be
found
in
Table
1;
this
paper
deals
with
the
influence
of
adsorption
of
(
a
)
a
dissolved
whey
powder
(
LAC-
PRODAN
80)
,
(
b
)
fl-lactoglobulin,
and
(
c
)
a-lactalbumin
on
the
performance
of
a
polysulphone
membrane
(
molecular
weight
cut-off
20,000;
DDS
GR61
PP
)
.
The
ultrafiltration
of
a
liquid
solution
may
be
modelled
in
terms
of
a
force
and
several
resistances.
The
driving
force
is
the
pressure
drop
across
the
mem-
brane.
This
force
acts
on
the
resistances
of
the
membrane,
the
adsorbed
layer
and
the
liquid
solution
(
i.e.
the
concentration
polarization
layer)
.
The
con-
*Paper
presented
at
the
Fifth
International
Symposium
on
Synthetic
Membranes
in
Science
and
Industry,
Tubingen,
F.R.G.,
September
2-5,
1986.
0376-7388/88/$03.50
©
1988
Elsevier
Science
Publishers
B.V.
148
TABLE
1
Reports
on
adsorption
on
membranes
Author
Subject
Ref.
Matthiasson
Adsorption
of
bovine
serum
albumin
2,
3
on
different
membranes
Aimar
et
al.
Adsorption
of
bovine
serum
albumin
4
on
organic/inorganic
membranes
Zeman
Adsorption
of
fl-lactoglobulin
5
centratiqn
polarization
layer
is
dependent
on
the
transport
mechanisms
for
retained
solutes:
their
diffusion
out
of
the
layer
and
their
fluid-mechanical
transport
into
(permeate
flux)
and
out
of
(turbulence)
the
polarized
layer.
In
itself,
the
concentration
of
solute
in
the
polarized
layer
affects
the
osmotic
pressure,
thus
counteracting
the
driving
force.
For
instance,
the
transmem-
brane
flux
can,
according
to
the
three-parameter
theory,
be
expressed
as
J=L,
(P
—07r)
1
L
p
R.+
R.
where
e.
I=
permeate
flux,
it
=
osmotic
pressure
of
the
solution,
1
3
=
applied
pres-
sure,
a
=
rejection
coefficient,
L
p
.the
hydraulic
permeability
of
the
mem-
brane,
R
m
=
resistance
due
to
the
membrane,
and
R
a
=
resistance
due
to
the
adsorbed
layer.
In
the
following
experiments,
the
nature
of
the
term
R
a
is
studied,
as
speci-
fied
in
the
methodological
section.
For
this
purpose,
a
whey
powder
and
the
above-mentioned
proteins
were
used,
thus
connecting
the
experiments
to
the
common
practice
of
whey
ultrafiltration.
Properties
of
whey
and
whey
proteins
Milk
contains
about
0.6%
whey
proteins,
fl-lactoglobulin
(
0.32%
)
,
immu-
noglobulins
(
0.07%
)
and
a-lactalbumin
(0.04%)
being
the
most
common.
Whey
is
the
fraction
of
milk
left
after
the
removal
of
the
casein
fraction
in
cheese
making.
/3-Lactoglobulin,
the
most
abundant
single
whey
protein,
has
a
M
w
of
18,363.
The
spherical
monomer
(
diameter
3.6
nm
)
forms
a
dimer
(M,,,
36,700)
.
The
geometry
of
the
dimer
is
a
prolate
ellipsoid.
At
pH
<
3.5
the
dimer
dissociates
reversibly.
At
pH
3.5-5.2
there
is
an
association
into
an
octamer
structure.
In
alkaline
conditions,
the
dimer
dissociates,
and
a
conformational
change
occurs
around
pH
7.5.
The
solubility
of
fl-lactoglobulin
increases
with
ionic
strength.
Its
isoelectric
point
is
pH
5.2
[
6
]
.
149
a-Lactalbumin
has
a
molecular
weight
of
14,200.
It
is
a
globular
protein,
having
a
radius
of
gyration
of
1.67
nm
and
the
dimensions
2.2
x
4.4
x
5.7
nm
(
an
oblate
ellipsoid)
.
Its
isoelectric
point
is
4.8
[6
]
.
Materials
and
methods
Materials
Water
In
all
experiments
distilled
water
was
used,
which
was
also
treated
on
a
Milli-
Q
unit
(Millipore
Company)
.
Sodium
azide
was
added
to
the
water
(0.02%
)
to
prevent
bacterial
growth.
Whey
powder
The
product
LACPRODAN
80,
was
obtained
as
a
gift
from
Danmark
Protein
A.S.,
Denmark.
It
is
described
by
its
manufacturer
as
an
undenaturated
whey
protein
concentrate,
produced
from
whey
by
ultrafiltration
and
spray
drying.
Analytical
data
given
by
the
manufacturer
are
shown
in
Table
2.
fl-Lactoglobulin
This
protein
was
obtained
from
the
Sigma
Company,
product
specification
No.
L-0130,
three
times
crystallized
and
lyophilized.
Lot
No.
55F-8220
was
used.
It
contains
fl-lactoglobulin
A
and
B.
a-Lactalbumin
This
protein
was
obtained
from
the
Sigma
Company,
product
specification
No.
L-6010.
It
contains
approximately
85%
lactalbumin,
and
is
calcium
de-
pleted
(
<
0.3
mol
Ca/mol
a-lactalbumin
)
.
"Synthetic
milk
ultra
filtrate"
buffer
This
buffer
was
prepared
according
to
Ref.
[
7
]
at
a
pH
of
6.8.
The
SMUF
buffer
comprises
ions
in
the
following
concentrations:
Na,
18.3
m/V1/1;
K,
39.4
TABLE
2
Product
specification
of
LACPRODAN
80
Protein
78+4%
K
approx.
0.3%
Lactose
4+2%
Na
approx.
0.2%
Fat
7+2%
Ca
approx.
0.3%
Ash
max.
3%
P
approx.
0.3%
pH
(10%
solution)
5.5-6.5
Pb
max.
0.5
ppm
Cu
max.
2.0
ppm
Fe
max.
10.0
ppm
As
max.
0.1
ppm
150
mM/1;
Ca,
9.0
milf/1;
Mg,
3.2
mM/1;
P,
11.6
mM/1;
Cl,
32.6
mM/1;
citrate,
9.6
mM/1;
SO
4
,
1.0
mil//1;
and
CO
2
,
2.2
mM/1.
Experiments
on
relative
flux
reduction
Experiments
on
static
adsorption
were
done
in
a
multibatch
cell
unit,
con-
sisting
of
ten
identical
cells,
each
having
a
membrane
area
of
4.5
cm
2
and
a
volume
of
18
cm
3
.
The
applied
pressure
is
uniform
throughout
the
multibatch
cell
unit.
The
static
experiments
were
performed
as
follows.
The
flux
of
a
pure
buffer
solution
through
a
membrane
was
measured
until
it
reaches
a
constant
value.
Usually,
the
experiment
was
allowed
to
run
overnight,
and
at
the
follow-
ing
day
the
flux
was
measured
every
15
minutes
during
at
least
90
minutes.
All
fluxes
are
measured
in
grammes
per
minute.
After
steady
state
was
attained,
a
solution
of
a
determined
solute
concentration
was
applied
to
the
membrane
surface
for
a
set
time.
The
protein
solution
was
removed,
and
the
membrane
rinsed
with
buffer
solution
until
no
protein
was
found
in
the
rinsing
solution.
After
rinsing,
the
flux
of
a
pure
buffer
solution
was
again
determined
during
another
60
minutes
of
steady
state.
The
pure
buffer
solution
consisted
of
a
0.010
M
phosphate
buffer
at
pH
6.8.
It
must
be
observed
that
in
the
static
adsorption
experiments
there
is
no
flux
of
macromolecular
solution
at
or
through
the
membrane.
Terms
used
in
calculations
with
the
gathered
data
The
results
of
the
static
experiments
are
characterized
by
the
relative
flux
reduction
due
to
adsorption,
expressed
as
RFR
=
1
/J
o
The
term
RF
=
/J
0
is
the
relative
flux.
The
relative
resistance
is
the
term
RR=
(1/RF)
—1.
In
the
previous,
J
1
=pure
buffer
flux
after
adsorption,
J
o
=
pure
buffer
flux
before
adsorption.
Method
of
sampling
from
OF
membranes
The
membrane
sheets
used
(
DDS
GR61
PP)
were
20
cm
x
20
cm.
For
the
adsorption
experiments,
pieces
were
cut
and
numbered
according
to
Fig.
1.
Most
experiments
were
done
using
triplicates,
but
sometimes
duplicates
were
used.
Usually
a
series
of
triplicate
points
in
an
experiment
was
taken
from
one
of
the
combinations
of
numbered
membrane
pieces
shown
in
Table
3.
This
allowed
a
calculation
of
the
variation
in
flux
within
and
between
membrane
sheets.
For
each
set
of
experiments,
a
new
membrane
sheet
was
used.
Thus,
151
11
12
13
14
15
16
21
22
23
24
25
26 27
31
32
33
34
35
36
41
42
43
44
45
46
47
51
52
53
54
55
56
61
63
64
65
66
67
71
72
73
74
75
76
81
82
83
84
85
86
87
Fig.
1.
Numbering
of
pieces
cut
from
DDS
GR61
PP
membranes.
TABLE
3
Combinations
of
numbered
membrane
pieces
used
11,44,87
24,56,74
12,43,86
25,45,73
13,42,85
26,46,72
14,41,84
27,47,71
15,51,83
31,64,36
16,52,82
61,33,67
21,53,81
62,34,66
22,54,76
63,32,35
23,55,75
the
pure
water
flux
of
all
numbered
points
belonging
to
the
same
membrane
sheet
was
not
determined
necessarily.
Method
of
radioactive
labelling
[
14
0
]
Formaldehyde
was
used
to
label
fl-lactoglobulin.
The
method
is
de-
scribed
by
Dottavio-Martin
and
Ravel
[
8
]
.
500
,uCi
[
14
C
]
formaldehyde
was
dissolved
in
2
x
125
pl
buffer.
To
2.5
ml
of
a
2.5%
w/w
/3
-lactoglobulin
solution
was
added
the
[
14
C
]
formaldehyde
solution
and
1
ml
freshly
prepared
solution
152
of
NaBH
3
CN
(
6
mg/ml
)
.
The
same
buffer
was
used
for
all
solutions.
The
mixture
was
left
to
stand
for
6
hours
at
room
temperature,
and
for
24
hours
at
4
°
C.
Excess
reagents
were
dialyzed
for
a
3
day
period
at
4
°C.
Adsorption
experiments
with
radiotagged
/3-lactoglobulin
To
enable
a
comparison
to
be
made
between
the
flux
reduction
at
a
certain
protein
concentration
and
the
amount
of
protein
adsorbed,
the
adsorption
ex-
periments
for
fl-lactoglobulin
were
repeated
using
radiotagged
protein.
Mem-
brane
pieces
were
put
into
a
cell
unit
consisting
of
12
identical
cells,
which
had
the
same
geometrical
configuration
as
the
multibatch
cell
unit,
but
lacking
the
permeate
outlet.
The
membranes
were
brought
into
contact
with
a
protein
solution
containing
a
mixture
of
labelled
and
non-labelled/3-lactoglobulin.
The
contact
time
was
one
hour.
After
removing
the
protein
solution,
the
mem-
branes
were
rinsed
with
buffer
and
dried
at
room
temperature.
The
remaining
=
50'10
-'
ML
ADSORBED
ON
MEMBRANE
ID=
50'10
-3
ML
IN
AQUASOL-2
20-
18-
16-
14-
12_
10_
8_
6_
4-
2_
RELATIVE
CONCENTRATION
0.125
0.25
0.5
1:o
Fig.
2.
Disintegrations
per
minute
versus
relative
concentrations.
1
=0.53
X
10
-2
%
w/w
lactoglobulin.
DPM
153
radioactivity
of
the
adsorbed
layer
was
measured,
putting
each
membrane
piece
in
15
ml
of
Aquasoll-2
solution
and
using
an
LKB
1217
Rackbeta.
The
validity
of
this
technique
for
determining
all
of
the
adsorbed
protein
was
checked
by
putting
50
ul
samples
of
protein
solutions
of
different
concen-
trations
on
membrane
pieces,
drying
the
pieces
and
thus
adsorbing
a
known
amount
of
activity.
The
membrane
pieces
were
put
in
15
ml
Aquasoll-2
solution
and
compared
with
50
ill
samples
of
the
protein
solutions
used
put
directly
into
Aquasoll-2.
The
results
of
this
experiment
are
shown
in
Fig.
2.
A
maximum
of
five
percent
underestimate
of
the
amount
adsorbed
is
indicated.
Results
Variation
in
membrane
fluxes
The
flux
of
a
pure
buffer
solution
was
not
the
same
throughout
the
mem-
brane
samples,
but
varied
strongly.
The
variation
of
the
flux
through
samples
cut
from
each
membrane
sheet
can
be
seen
in
Fig.
3a.
The
figure
is
based
on
387
different
circular
samples,
each
having
a
diameter
of
2.6
cm,
cut
from
10
different
sheets,
but
all
belonging
to
the
same
batch
(
400-013)
of
DDS
GR61
PP
membranes.
In
Fig.
3b
the
number
of
membranes
with
a
certain
flux
has
160
150
140
130
NUMBER
OF
MEMBRANE
PIECES
FLOW
(G/cm
2
`miN.)
120
110
22_
100
20
_
90
18
80
16_
70
14
_
60
12
50
10
40
8
_
30
6
_
20
4
10
FLUX
(G
/cM
2
`MIN.)
2
FLUX(G/cm
2.
rim.)
,...-
0.1
0
2
0.3
0.1
0,2
0.3
Fig.
3a.
Variation
of
flux
through
samples
cut
from
each
membrane
sheet.
Fig.
3b.
Plot
of
number
of
membranes
with
a
certain
flux
times
that
flux
vs.
flux.
154
.00
.98
0.84 0.94
0.75
1.16
1.2
0.88
.9
b
.92
.92
0.8
0.88
0.94
0.8
.22
.21
.Z
0.7
.10
10.98
.14
.9
1.33
0.8
0
.31
0.670.65
0.
0.84
.25
94
.82
.94
1X31
.24
1.35
1.20
.47
0.96
08
1.04
0.64
0.8£
.71
1.22
0.78
0.8C
.69
0.:0
1.37
Fig.
4.
Variation
of
flux
in
a
DDS
GR61
PP
membrane
sheet.
The
fluxes
have
been
normalized,
i.e.
divided
by
the
flux
average
of
the
whole
membrane
sheet.
The
flux
average
of
0.11
g/cm
2
-min
thus
corresponds
to
a
normalized
value
of
1.00.
NUMBER
OF
MEMBRANE
PIECES
(MILLIPORE)
10
9
8
7
6
5
4
3
2
1
0.1
0.2
0.3
0,4
0.5
Fig.
5.
Flux
variation
pattern
of
Millipore
PTTK
06210
membranes.
been
multiplied
by
the
flux,
and
the
corresponding
result
has
been
plotted
versus
the
flux.
This
indicates
the
part
of
the
flux
originating
from
each
flux
interval.
In
Fig.
4,
the
flux
variation
in
a
selected
DDS
GR61
PP
membrane
sheet
is
shown.
For
the
sake
of
comparison,
the
pure
water
flux
was
determined
for
a
small
set
of
membranes
from
another
manufacturer.
Four
pieces
were
cut
from
each
of
nine
Millipore
PTTK
06210
membranes
(
lot
No.
C5K73239
).
These
mem-
branes
showed
the
same
pattern
of
flux
variation
as
do
the
DDS
membranes
(Fig.
5).
Flux
experiments
The
time
necessary
for
the
adsorption
to
reach
an
equilibrium
state
was
determined
for
three
different
solutions
of
whey
powder,
fl-lactoglobulin
and
155
0.5
0.025%w/w
WPC
=
0.25Zw/w
WPC
=
2.5%w/w
WPC
=
20%w/w
WPC
0.4
w
iL
0.3
0.2
0.1
v
10
30
60
120
TIME
Fig.
6.
Relative
flux
reduction
vs
time
for
various
whey
concentrations.
=
0.025%w/wp-LACTOGLOBULIN
=
2.5%w/w
)3-LACTOGLOBULIN
0.5
Q4
ta.
0.3
Q2
0.1
10
20
60
90
120
TIME
Fig.
7.
Relative
flux
reduction
vs
time
for
two
fl-lactoglobulin
concentrations.
a-lactalbumin
in
0.1
M
phosphate
buffer,
pH
6.8.
For
whey
it
was
determined
for
the
concentrations
of
0.025%,
0.25%,
2.5%
and
20%,
see
Fig.
6;
for
/3-lac-
toglobulin
and
a-lactalbumin
the
concentrations
were
0.025%
and
2.5%,
see
Figs.
7
and
8.
From
these
experiments
an
equilibrium
time
of
one
hour
was
decided
to
be
sufficient.
Q
156
=
0.025WWU-LACTALBUMIN
=
2,5Ww
CE-LACTALBUMIN
0.5—
0.4—
L.
T
.
O.
3
cc
0
2-
0.1_
10
30
60
90
120
TIME
Fig.
8.
Relative
flux
reduction
vs
time
for
two
a-lactalbumin
concentrations.
=
a-LACTALBUMIN
-
WPC
*
=j3-LACTOGLOBULIN
0.6
-
0.5
_
0,4
_
0.3
0.2
-
0.1
CONCENTRATION
%w/w
0.025
0.05
0.1
0.2
0.4
0.8
1.6
3.2
6.4
12.8
25.6
Fig.
9.
Relative
flux
reduction
vs
protein
concentration
for
whey
powder,
a-lactalbumin
and
)3-
lactoglob
ul
in.
Adsorption
isotherms
were
drawn
as
relative
flux
reduction
(
RFR
)
versus
protein
concentration
for
the
following
cases:
1,
whey
powder,
dissolved
in
0.1
M
phosphate
buffer,
pH
6.8,
Fig.
9;
2,
fl-lactoglobulin,
dissolved
in
0.1
M
phosphate
buffer,
pH
6.8,
Figs.
9
and
10;
3,
a-lactalbumin,
dissolved
in
0.1
M
phosphate
buffer,
pH
6.8,
Fig.
9;
157
=p-LACTOGLOBULIN,
0.1M
PHOSPHATE
BUFFER
=13-LACTOGLOBULIN,
0.1M
SMUF
BUFFER
0.5_
0.4_
cc
X
0.3_
0.2
0.1..
CONCENTRATION
%w/w
' I
I
I
025
05
1
4
8
1.6
3
'
.2
6
'
.4
12.8
'
25
.6
Fig.
10.
Relative
flux
reduction
vs
protein
concentration
for
fl
-
lactoglobulin
in
two
different
buffers.
=
WPC
=13-LACTOGLOBULIN
*
=f3
-LACTALBUM
I
N
0.6-
0.5-
0.4_
cc
w
l-L
0.3-
0.2_
0.1_
4
6
6
7
8
e
pH
Fig.
11.
Relative
flux
reduction
vs
pH
for
whey
powder,
a
-lactalbumin
and
/3-lactoglobulin.
4,
fl-lactoglobulin,
dissolved
in
SMUF
buffer,
pH
6.8,
Fig.
10.
The
pH
dependence
was
tested
for
1,
whey
powder,
0.8%
w/w,
Fig.
11;
2,
fl-lactoglobulin,
0.8%
w/w,
Fig.
11;
3,
a-lactalbumin,
0.8%
w/w,
Fig.
11.
Amount
of
fl-lactoglobulin
adsorbed
and
its
relation
to
flux
decrease
Using
the
technique
described
above,
the
amount
of
/3-lactoglobulin
ad-
sorbed
on
DDS
membranes
from
a
solution
of
0.1
M
phosphate
buffer,
pH
6.8,
158
(a)
G/CM
2
100-
50.
CONCENTRATION
%w/v4
lo
20
30
40
(b)
GiCM
2
100
50.
CONCENTRATION
%w/v4
0.25
0
.
5
1
2
4
8
1.6
3.2
6
.
4
12.8
25.6
Fig.
12.
Adsorption
of
fl-lactoglobulin
vs
concentration.
159
0.6
0.5
cc
0.4
cr
0.3
Q2
0.1
G
C
M
2
0
50
100
150
Fig.
13.
Relative
resistance
of
adsorbed
$-lactoglobulin
vs
amount
of
protein
adsorbed.
was
determined
for
a
range
of
protein
concentrations
of
0.025%
w/w
to
40%
w/w.
These
amounts
are
given
in
Figs.
12a
and
b.
Figure
13
shows
the
relative
resistance
of
adsorbed
/3-lactoglobulin
as
a
function
of
the
amount
of
protein
adsorbed.
Discussion
The
following
conclusions
may
be
drawn
from
the
gathered
material;
The
adsorption
process
follows
the
same
pattern
for
the
different
proteins
used.
There
is
a
strong
dependence
on
the
pH
of
the
protein
solution;
also
the
type
of
buffer
is
an
influence.
Different
buffers
give
the
same
pattern
of
different
plateaus
in
the
adsorption
isotherms,
but
the
extent
of
each
such
region
is
displaced.
There
is
a
very
early
inflexion
point
in
the
dependence
of
the
relative
resistance
on
the
amount
of
protein
adsorbed.
Also,
one
should
note
the
ho-
mogeneity
of
adsorption
versus
the
non-homogeneity
of
the
flux
distribution
of
the
membrane.
References
1
V.
Gekas,
Terminology
for
Pressure
Driven
Membrane
Operations,
approved
by
the
European
Society
of
Membrane
Science
and
Technology,
June
1986.
2
E.
Matthiasson,
The
role
of
macromolecular
adsorption
in
fouling
of
ultrafiltration
mem-
branes,
J.
Membrane
Sci.,
16
(1983)
23-26.
160
3
E.
Matthiasson,
B.
Hallstrom
and
B.
Sivik,
Adsorption
phenomena
in
fouling
UF-membranes,
in:
Engineering
and
Food,
Vol.
1,
Proc.
Third
International
Congress
on
Engineering
and
Food,
Dublin,
Ireland,
September
1983.
4
P.
Aimar,
J.P.
Lafaille
and
V.
Sanchez,
Influence
of
adsorption
on
protein
ultrafiltration
using
organic/inorganic
membranes,
Proc.
2nd
International
Conference
on
Fouling
and
Cleaning,
Madison,
WI,
July
14-17,1985,
pp.
486-495.
5
L.
Zeman,
Adsorption
effects
in
rejection
of
macromolecules
by
ultrafiltration
membranes,
J.
Membrane
Sci.,
15
(1983)
213-230.
6
P.F.
Fox
(Ed.)
,
Development
in
Dairy
Chemistry,
Vol.
1,
Proteins,
Applied
Science
Publish-
ers,
Barking,
Great
Britain,
1982.
7
R.
Jenness
and
J.
Koops,
Preparation
and
properties
of
a
salt
solution
which
simulates
milk
ultrafiltrate,
Neth.
Milk
Dairy
J.,
16
(3)
(1962)
153-164.
8
D.
Dottavio-Martin
and
J.M.
Ravel,
Radiolabeling
of
proteins
by
reductive
alkylation
with
(14C)
formaldehyde
and
sodium
cyanoborohydride,
Anal.
Biochem.,
87
(1978)
562-565.