Electrodialysis and ion exchange processes; the case of milk whey


de Boer, R.; Robbertsen, T.

Progress in Food Engineering: 393-403

1983


Topics covered in this review are: manufacture of a demineralized whey by electrodialysis (process conditions and pre-treatment); recent developments in the demineralization of whey by ion exchange; processing of ultrafiltration permeate by ion exchange resins (manufacture of lactose and manufacture of a lactose-hydrolysed syrup); and whey protein recovery by ion exchange resins. Some new research is reported including: studies of the influence of process conditions and pretreatments on demineralization rate, 'power' efficiency and overall yield during electrodialysis of Gouda cheese whey; and studies on the decalcification and decolorization of the ultrafiltration permeate of Gouda whey during lactose manufacture.

Electrodialysis
and
Ion
Exchange
Processes:
The
Case
of
Milk
Whey
R.
de
Boer
and
T.
Robbertsen
Netherlands
Institute
for
Dairy
Research
(NIZO),
Ede,
The
Netherlands
1.
Introduction
The
production
of
cheese
is
one
of
the
main
activites
of
dairy
processing
and
it
results
in
the
formation
of
large
volumes
of
the
by-product
whey.
Mainly
due
to
environmental
problems
created
by
the
disposal
of
whey,
the
dairy
industry
has
to
process
this
whey
completely.
The
conversion
of
whey
into
powder
is
the
most
simple
operation,
however,
disappointingly
low
prices
make
it
necessary
to
look
for
alternatives.
One
of
the
alterna-
tives
is
the
manufacture
of
a
demineralized
whey
powder.
Due
to
the
re-
duction
of
the
mineral
content
this
product
is
suitable
for
human
con-
sumption;
it
can
be
used
in
infant-food
formulas,
ice-cream,
sherbets,
bakery
products
and
chocolate.
For
demineralization
of
whey
electro-
dialysis
and
ion
exchange
have
been
used
commercially
for
more
than
two
decades.
Dozens
of
installations,
for
electrodialysis
as
well
as
ion
exchange,
are
in
operation
in
the
dairy
industry
all
over
the
world.
As
waste
water
control
becomes
more
stringent,
it
is
expected
that
systems
involving
electrodialysis
and
ion
exchange
will
face
much
higher
costs
in
the
future.
Aspects
such
as
use
of
chemicals,
considerable
in
the
case
of
an
ion
exchange
process,
and
overall
yield,
therefore
play
an
important
role
in
the
choice
of
one
of
the
two
demineralization
processes.
In
some
regions
in
Holland
for
instance
the
amount
of
salt
disposal
is
limited
to
such
an
extent
that
for
the
manufacture
of
a
highly
demineralized
whey
powder
electrodialysis
is
the
only
possibility.
In
contrast
to
electrodialysis,
ion
exchange
can
be
applied
for
other
pur-
poses
than
demineralization.
In
the
beet
sugar
industry
ion
exchange
resins
are
also
used
for
decolorization,
inversion
of
saccharose,
separation
of
glucose
and
fructose
etc.
(Herve,
1974,
b).
Therefore
we
will
deal
with
a
wider
scope
than
desalting
and
also
go
into
subjects
as
decalcification,
decolorization
and
protein
recovery
as
far
as
whey
and
deproteinized
whey
is
concerned.
2.
Manufacture
of
a
highly
demineralized
whey
by
electrodialysis
Electrodialysis
is
an
electrochemical
process
for
reducing
the
salt
content
of
liquids
by
the
use
of
direct
current
electricity
and
anion-
and
cation-
permeable
membranes.
Under
the
influence
of
an
applied
voltage,
cations
(Na
+
,
K
+
)
will
migrate
towards
the
cathode,
anions
(Cl)
will
move
to
the
anode.
A
cation-permeable
membrane
will
allow
passage
of
cations
but
will
repel
the
approach
of
anions,
while
the
anion-permeable
membrane
per-
forms
the
opposite.
Between
the
membranes
spacers
are
placed,
which
avoid
contact
between
the
cation-
and
anion-permeable
membrane
and
393
which
act
as
turbulence
promotors
so
that
there
is
a
mixing
effect
as
the
solution
passes
between
the
membranes,
and
concentration
polarization
is
limited.
By
electrodialysis
the
whey
solids
are
split
into
two
fractions,
one
enriched
in
ionic
species
(concentrate)
and
the
other
depleted
(dilu-
ate-product
stream).
A
good
discussion
of
the
process
is
given
by
Mintz
&
Shaffer
(1966).
According
to
Ahlgren
(1972)
demineralization
of
whey
poses
problems
with
respect
to
fouling
of
membranes
such
as
deposition
of
inorganic
calcium
phosphate
salts
on
the
cathode
side
of
the
cation-permeable
mem-
branes
and
a
proteinaceous
layer
on
the
surface
of
the
anion-permeable
membranes.
Good
cleaning
procedures,
a
limitation
of
the
current
density
to
20-25
mA/cm
2
,
keeping
the
concentrate
at
low
pH
and
a
conductivity
of
about
55
mS
and
operating
above
critical
velocities
reduce
these
pro-
blems.
Another
problem,
which
is
inherent
in
electro-dialysis,
is
that
com-
plete
demineralization
is
impossible,
due
to
effects
such
as
concentration
polarization.
According
to
laconelli
(1973)
the
production
capacity
of
Ionics
equipment
is
doubled,
in
the
case
of
demineralization
to
75%
rather
than
90%
ash
removal
and
it
doubles
again,
in
the
case
of
demineralization
to
only
50%
ash
removal.
Recent
figures
by
Houldsworth
(1980)
indicate
that
the
capital
and
membrane
costs
are
considerable
at
a
90%
demin-
eralization
level.
In
the
following
sections
we
will
discuss
the
influence
of
the
process
conditions
and
pre-treatment
of
Gouda
cheese
whey
on
the
demineralization
rate,
"power"
efficiency
and
overall
yield
in
the
manu-
facture
of
a
90%
demineralized
whey
product.
2.1.
Process
conditions
In
our
experiments
(batch
system)
we
used
an
Asahi
Glass
DW-XYO
semi-technical
equipment
with
20
cell
pairs
and
a
membrane
surface
of
6.9
m
2
.
The
distance
of
the
membranes
(AMV
and
CMV
membranes)
at
the
diluate
side
is
0.75
mm
and
at
the
concentrate
side
0.56
mm,
the
flow
velocities
are
14
and
4
cm/s
respectively.
In
concentrate
as
well
as
diluate
cells,
which
are
provided
with
sheet-flow
spacers,
a
pressure
at
the
inlet
of
0.05
MPa
is
maintained.
The
concentrate
was
kept
at
pH
2.5
and
the
conductivity
was
about
55
mS.
In
the
beginning
of
an
exper-
iment
the
conductivity
of
the
whey
is
high
and
the
current
has
to
be
restricted
by
the
voltage,
while
at
the
lower
conductivities
the
voltage
is
allowed
to
reach
the
maximum
indicated
by
the
manufacturer.
In
Table
1
the
influence
of
temperature
and
voltage
per
cell
pair
on
the
demineralization
rate
and
"power"
efficiency
is
indicated
at
a
reduction
in
conductivity
of
80%,
which
can
be
achieved
by
various
types
of
elec-
trodialysis
equipment.
The
change
in
temperature
from
30
to
40
C
corresponds
with
an
increase
in
conductivity
and
therefore
with
a
higher
demineralization
rate
of
60
g
ash/m
2
.
h
and
a
positive
effect
on
"power"
efficiency
of
130
g
ash/kWh.
The
increase
of
voltage
per
cell
pair
from
1.5
to
2.2
the
latter
is
the
maximum
allowed
by
the
manufacturer
gave
a
higher
demineralization
rate
and
had
a
negative
influence
on
the
"power"
efficiency.
Obviously
the
electric
energy
is
used
less
economi-
cally
at
2.2
V
and
partially
dissipated
into
heat.
Our
experience
is
in
agree-
ment
with
that
of
Johnston
et
al.
(1976),
who
carried
out
experiments
with
Ionics
equipment,
even
at
higher
temperature
and
voltage.
394
Table
1.
The
influence
of
various
process
conditions
and
pretreatments
on
the
demineralization
rate
and
"power"
efficiency
during
electrodialysis
-
Asahi
Glass
DW-XYO
pilot
plant
-
of
Gouda
cheese
whey
(pH
6.6,
5,5
%
total
solids)
to
a
reduction
in
conductivity
of
80
%.
Concentrate:
55
mS,
pH
2.5
demineralization
rate
(9
ash/m
4
.
h)
"power"
efficiency
(g
ash/kWh)
process
conditions:
-
Gouda
whey/30
°
C
-
1.5
V
-
temperature:
30
40
°
C
-
voltage:
1.5
2.2
V
pretreatments:
-
Gouda
whey/40
°
C
-
2.2
V
-
whey
concentrated
whey
(22
%
TS)
-
whey
acidified
whey
(pH
4.6)
-
whey
decalcified
whey
-
whey
--deproteinized
whey
-
whey
simulated
milk
ultrafiltrate
70
+60
+
40
190
+
90
+110
+100
ns
+160
710
+130
-180
630
+290
+180
+150
ns
.
+160
no
significant
difference
2.2.
Pre-treatment
Demineralization
by
electrodialysis
is
carried
out
on
various
types
of
whey
and
its
derivatives.
Johnston
et
al.
(1976)
found
a
positive
effect
of
pre-concentration
due
to
an
increase
in
conductivity
of
the
whey.
However,
experiments
done
on
ultrafiltration-concentrate
were
less
successful,
which
was
attributed
to
the
protein
molecules
hindering
the
mobility
of
the
ionic
species
in
the
raw
whey.
D'Souza
et
al.
(1973)
de-
mineralized
unconcentrated
acid
cottage
cheese
whey
at
various
pH
values.
A
maximum
demineralization
rate
was
obtained
near
the
iso-
electric
point
(pH
4.6),
as
here
no
interference
should
be
experienced
in
the
transport
of
cations
and
anions
due
to
a
(net)
neutral
charge
on
the
protein
molecules.
Short
&
Doughty
(1977)
and
Higgins
&
Short
(1980)
investigated
demineralization
of
sweet
and
acid
ultrafiltration-permeate
with
SRTI
and
Morinaga
equipment.
Differences
in
demineralization
rate
and
energy
requirements
could
be
attributed
to
a
different
initial
ash
content
and
the
presence
of
acid
together
with
a
greater
content
of
calcium
and
phosphate
in
the
acid
permeate.
Table
1
gives
also
the
effects
of
various
pre-treatments
as
concentration,
acidification,
decalcification
and
deproteinization
carried
out
on
Gouda
cheese
whey,
which
make
comptrison
possible.
Deproteinized
whey
is
obtained
by
ultrafiltration
at
10
C.
A
decalcified
whey
is
prepared
by
treating
the
whey
with
a
weak
acid
ion
exchange
resin,
by
which
calcium
ions
are
replaced
by
sodium
ions.
The
simulated
milk
ultrafiltrate
(SMUF),
which
represents
the
salt
solution
of
the
whey,
is
prepared
following
the
procedure
of
Jenness
&
Koops
(1962).
According
to
Heting
(1972)
the
disappearance
rates
for
potassium
—,
chloride
and
sodium
ions
395
are
much
higher
than
those
for
calcium
and
phosphate
ions
in
cheese
whey.
This
phenomenon
relates
to
the
calcium
phosphate
and
calcium
citrate
salts,
which
are
only
partially
dissociated
at
pH
6.6.
By
acidifi-
cation
and
decalcification
the
salts
are
converted
into
the
ionized
form,
which
makes
them
more
feasible
for
removal
by
electrodialysis.
The
positive
influence
of
these
pre-treatments
on
the
demineralization
rate
is
of
the
same
magnitude
as
the
already
mentioned
influence
of
the
in-
crease
in
conductivity
by
pre-concentration.
The
best
results
were
ob-
tained
in
the
case
of
SMUF,
obviously
the
other
whey
constituents
(whey
proteins,
NPN,
lactose)
have
a
negative
effect on
the
demineraliz-
ation
rate.
Probably
due
to
mathematical
interactions
(concentrated
and
unconcentrated
liquids
were
used)
the
statistical
analysis
gives
at
a
re-
duction
in
conductivity
of
80%
no
demonstrable
influence
on
the
whey
proteins.
Furthermore,
the
"power"
efficiency
figures
of
Table
1
in-
dicate
that
alle
pre-treatments
had
a
positive
effect,
particularly
pre-
concentration
of
the
Gouda
whey.
Delbeke
(1975),
Hiraoka
et
al.
(1979)
and
laconelli
(1973)
de-ashed
with
90%,
using
various
equipment,
while
Okada
et
al.
(1977)
developed
for
this
purpose
specific
anion-permeable
membranes
with
larger
pores
to
remove
organic
acids
more
easily.
With
the
exception
of
pre-concen-
tration,
these
authors
did
not
consider
the
influence
of
the
pre-treatment.
Table
2
gives
results
for
normal,
decalcified
and
deproteinized
whey
and
SMUF
at
two
pH
values
for
two
concentration
factors.
All
products
were
demineralized
to
0.9%
ash
on
total
solids,
in
the
case
of
SMUF
the
con-
ductivity
was
reduced
by
96%.
Table
2
shows
that
a
pH
value
of
4.6
has
Table
2.
The
influence
of
various
pretreatments
on
the
demineralization
rate,
"power"
efficiency
and
overall
yield
during
etedrodialysis
-
Asahi
Glass
DW-XYO
pilot
plant
-
of
Gouda
cheese
whey
to
0.9
%
ash
on
total
solids.
Process
conditions:
40°C,
max.
2.2
V/cell
pair,
concentrate
55
mS
-
pH
2.5
pretreatment
initial
demineralization
"power"
efficiency
overall
pH
rate
19
ash/W.
hl
(g
ash/kWh)
yield
t%
simulated
milk
ultrafiltrate
normal
concentration
6.6
90
390
14.6
nal
600
four
times
concentrated
6.6
320
710
decalcified
whey
(after
ion
exchange)
normal
concentration
14.6
380
1
960
610
89
6.6
170
14.6
2001
580
86
four
times
concentrated
6.6
200
820
85
deproteinized
whey
lUF-permeate)
normal
concentration
14.6
3201
1080
240
87
87
6.6
90
14.6
1101
390
85
four
times
concentrated
6.6
150
580
86
14.6
180
640
86
Gouda
whey
normal
concentration
6,6
80
420
8o
lour
times
concentrated
0
b.6
460
580
85
88
14.6
1201
620
89
Reduction
in
conductivity
of
96
%
Difficult
to
obtain
0.9
ash
TS
396
a
positive
influence
on
the
demineralization
rate
in
all
cases.
At
this
pH
the
highest
concentration
factor
gave
the
best
results;
only
in
the
case
of
four-times-concentrated
Gouda
whey
was
there
no
important
dif-
ference.
Very
favourable
demineralization
rates
were
achieved
with
decalcified,
fourfold
concentrated
whey
at
4.6,
which
can
be
explained
by
the
replacement
of
calcium
by
sodium
ions.
Compared
with
the
normal
pro-
cedure
(concentrated
Gouda
whey
pH
6.6)
the
capacity
can
be
increased
by
a
factor
of
three,
at
least
in
short-run
experiments.
Moreover,
the
"power"
efficiency
can
be
reduced
by
a
factor
of
two.
However,
the
highest
demin-
eralization
rates
were
obtained
with
SMUF
(four
times
concentrated,
pH
4.6).
If
SMUF
was
brought
to
the
same
lactose
content
as
whey,
the
demineralization
rate
diminshed
from
380
to
240
g
ash/m`
.h.
The
ex-
planation
for
this
phenomenon
may
be
the
increase
in
viscosity,
which
decreases
the
mobility
of
(
the
ions
snsmuF
+
lactose
1
.
12
mPa.s
and
flsmuF
0.70
mPa.s
at
40
o
C).
The
demineralization
rate
was
further
re-
duced
in
the
case
of
concentrated
deproteinized
whey
at
pH
4.6
(180
g
ash/m
2
.h)
and
concentrated
whey
at
pH
4.6
(120
g
ash/m
2
.h)
demon-
strating
the
negative
influence
of
each
whey
constituent.
The
influence
of
the
whey
proteins
becomes
obviously
negative
after
concentration
of
the
whey.
In
addition
to
the
demineralization
rate
and
"power"
efficiency
figures
were
also
collected
concerning
the
overall
yield.
This
yield
is
a
percentage
of
the
initial
whey
solids,
which
remained
after
electrodialy-
sis.
The
average
value
of
87%
compared
with
other
equipment
is
quite
favourable
(Houldsworth,
1980).
3.
Recent
developments
in
the
demineralization
of
whey
by
ion
exchange
As
opposed
to
electrodialysis,
which
removes
ions
from
a
solution
on
an
electrochemical
basis,
ion
exchange
is
a
fixed-bed
technique
involving
use
of
ion
exchange
resins.
These
resins
are
macromolecular
composites,
which
can
be
considered
as
acids
or
bases.
Their
main
characteristic
is
their
ca-
pacity
to
exchange
the
mobile
ions
which
they
contain,
for
ions
of
the
same
charge
sign
contained
in
the
solution
being
treated.
Sulphonated
co-
polymers
of
styrene
and
divinyl
benzene
are
most
widely
used
as
cation
exchange
materials.
Quaternary
ammonium
derivatives
of
this
co-polymer
are
commonly
used
as
strong-base
anion
exchange
resins,
while
secondary
and
tertiary
amine
derivatives
serve
as
weak
anion
exchange
resins.
More
information
on
the
ion
exchange
process
itself
and
its
applications
is
given
by
Calmon
&
Gold
(1978).
For
the
treatment
of
whey
strong
acid
cation
and
weak
base
anion
ex-
change
resins
are
normally
used.
Delbeke
(1972)
investigated
various
ion
exchange
resins
for
the
90%
demineralization
of
cheese
whey.
Proper
resins
had
a
capacity
of
about
16
I
whey/I
resin,
while
the
overall
yield
of
the
whey
solids
was
86-88%
and
the
nitrogen
loss
about
15%.
For
the
mini-
mization
of
the
protein
°
loss
Delaney
(1976)
recommended
low
tem-
perature
operation
(5-12
C)
without
prior
pasteurization.
In
normal
in-
dustrial
installations
the
capacity
of
the
resin
is
about
10-11
I
and
in
the
case
of
acid
whey
even
lower
instead
of
the
mentioned
16
I
cheese
whey/I
resin.
This
lower
capacity
has
to
be
related
to
operating
conditions,
of
which
regeneration
is
the
most
important.
In
the
dairy
industry
the
exhausted
cation
exchange
resin
is
treated
normally
with
a
hydrochloric
397
acid
solution
and
the
anion
exchange
resin
with
a
caustic
soda
solution.
Minimization
of
the
quantities
of
these
chemicals
can
give
a
certain
loss
of
capacity.
According
to
Houldsworth
(1980)
and
Delbeke
(1975)
the
cost
price
of
treated
whey
depends
to
about
50
to
60%
on
the
consumption
of
chemi-
cals
used
for
regeneration:
therefore
a
lot
of
work
is
being
done
on
econ-
omize
these
chemicals.
Herve
(1974),
a)
points
out
that
weak
base
anion
exchange
resins
can
be
regenerated
easily;
normally
a
concurrent
system
is
applied
and
about
150%
reagent
in
relation
to
the
theoretical
amount
is
used.
Strong
acid
cation
exchange
resins
are
regenerated
with
more
dif-
ficulty.
In
a
concurrent
system
about
300%
of
hydrochloric
acid
is
required
in
relation
to
the
theoretical
amount.
According
to
Herve
(1974,
a)
the
countercurrent
technique
makes
it
possible
to
decrease
the
consumption
of
the
hydrochloric
acid
by
40%,
while
furthermore
the
final
product
has
a
superior
quality.
During
countercurrent
regeneration
the
acid
flows
in
a
direction
opposite
to
that
of
the
whey.
Normally
the
acid
solution
goes
from
bottom
to
top,
but
in
the
"Schwebebett
system"
described
by
Sie-
gers
and
Martinola
(1972)
the
regenerant
flows
from
top
to
bottom.
The
recovery
of
the
chemicals
needed
for
demineralization
of
whey
was
described
by
Jonsson
and
Forsman
(1978).
In
their
"ammonium
bicar-
bonate
process"
the
whey
cations
such
as
calcium,
sodium
and
potassium
were
exchanged
for
ammonium
ions
and
anions
such
as
chloride
and
phosphate
for
bicarbonate
ions.
At
the
subsequent
evaporation
of
the
whey
the
ammonium
bicarbonate
is
volatilized
as
ammonia,
carbon
dioxide
and
water.
By
taking
care
of
the
ammonia
and
carbon
dioxide
stripped
off
during
evaporation
and
by
distilling
the
surplus
of
ammonium
bicarbonate
in
the
used
regeneration
solution,
possibilities
exist
to
recover
ammonium
bicarbonate
for
regeneration
of
the
resins.
The
overall
recovery
of
the
chemicals
in
this
rather
complicated
process
is
at
least
70%.
The
authors
suggested
further
that
the
whey
salts
removed
can
be
used
in
fertilizers.
Bolto
et
al.
(1979)
described
a
thermally
regenerated
ion
exchange
process
for
water.
The
essence
of
this
process,
in
which
specially
developed
"plum
pudding"
resins
are
used,
is
the
discovery
that
the
contaminating
ions
from
the
water
are
adsorbed
at
ambient
temperature
and
made
to
re-
lease
by
hot
water
at
80-90
C.
In
this
process
the
need
for
acid
and
base
regenerants
required
for
the
conventional
ion
exchange
resins
is
obviated.
Parrish
et
al.
(1979)
showed
in
their
study
that
thermally
regenerable
ion
exchange
resins
can
be
used
to
demineralize Cheddar
whey
ultrafiltrate
to
such
an
extent,
that
the
lactose
yield
after
crystallization
is
improved.
The
system
is
unlikely
to
be
suitable
for
whey
because
of
the
precipitation
of
proteins
which
will
occur
at
high
temperatures.
The
Gryllus
deliming
process,
which
is
applied
in
the
beet
sugar
industry,
approaches
the
ideal
situation
(Felber,
1970).
The
thin
juice
is
decal-
cified
by
a
cation
exchanger
loaded
with
monovalent
ions
and
then
fed
to
an
evaporator.
Incrustations
on
the
heating
tubes
of
the
evaporator
are
avoided
by
decalcification.
The
thick
juice
is
then
used
for
regeneration
of
the
resins
in
countercurrent
flow.
This
simple
solution
is
based
on
the
fact
that
the
selectivity
of
ions
having
different
valencies
depends
on
their
concentration
in
the
solution.
At
low
concentration,
the
selectivity
in-
creases
considerably
in
favour
of
the
ions
with
higher
valency,
whereas
at
high
concentration
the
selectivity
is
displaced
towards
the
ions
of
low
valency.
Due
to
the
relatively
low
solubility
of
lactose,
this
process
may
be
less
suitable
for
whey.
398
4.
Processing
of
ultrafiltration
permeate
by
ion
exchange
resins
The
consequence
of
the
increasing
valorization
of
whey
proteins
either
in
cheese
or
in
whey
protein
concentrates,
is
the
partial
replacement
of
the
normal
by-product
whey
by
ultrafiltration
permeate
(UF-permeate).
In
this
section
we
will
discuss
the
manufacture
of
UF-permeate
(4.1%
lactose,
0.2%
non-protein
nitrogen,
0.5%
ash,
0.2%
citric
acid)
into
lactose
and
hydrolyzed
lactose
syrup.
Purification
of
the
UF-permeate
is
carried
out
by
means
of
ion
exchange
resins.
4.1.
Manufacture
of
lactose
Delbeke
(1979)
desalted
and
purified
UF-permeate
by
the
use
of
an ad-
sorbent
resin,
which
is
responsible
for
the
removal
of
riboflavin,
and
a
strong
acid
and
strong
base
ion
exchange
resin
which
removes
the
salts
and
a
great
deal
of
the
non-protein
nitrogen
(NPN)
fraction.
After
this
treatment
the
product
had
the
following
composition:
99.74%
lactose,
0.12%
ash
and
0.14%
"protein"
on
a
total
solids
basis.
It
may
be
ex-
pected
that
a
higher
purity
of
the
lactose
can
be
obtained
if
a
crystal-
lization
step
which
is
normal
in
the
manufacture
of
lactose
from
whey
is
introduced
into
the
system.
In
that
case
the
purification
of
the
UF-permeate
by
resins
may
be
less
intensive,
which
makes
the
waste
water
problem
less
severe.
At
our
institute
we
carried
out
experiments
with
UF-permeate,
in
which
the
calcium
was
exchanged
by
sodium
ions
to
prevent
the
precipitation
of
calcium
phosphate.
For
this
purpose
a
weak
acid
ion
exchange
resin
of
the
carboxylic
type
(Duolite
C464)
was
used.
Furthermore,
the
ribo-
flavin,
which
gives
the
yellow-green
colour
to
UF-permeate,
was
removed
by
means
of
a
resin
with
weak
and
strong
acid
groups
(Imac
Syn
106).
This
resin
can
be
regenerated
by
the
normal
chemicals
as
caustic
soda
and
hydrochloric
acid.
In
the
upper
part
of
Fig.
1
the
calcium-,
potas-
sium-
and
sodium
contents
are
shown
after
a
treatment
with
the
C464
resin.
The
calcium
content
of
the
UF-permeate
used
was
not
very
high,
which
can
be
attributed
to
the
pre-treatment
of
the
twofold
concen-
trated
whey
prior
to
ultrafiltration.
Under
these
circumstances
calcium
leakage
became
considerable
after
a
passage
of
60
I
UF-permeate
per
1
resin.
At
the
beginning
of
the
treatment
we
see
that
not
only
calcium
but
also
potassium
is
exchanged
by
sodium,
which
is
a
consequence
of
the
order
of
affinity
and
the
rather
high
concentration
of
the
potassium.
The
resin
can
also
be
loaded
by
hydrogen
ions
instead
of
sodium
ions;
in
that
case
the
calcium
content
after
passage
is
somewhat
higher,
the
capacity
somewhat
lower
and
the
UF-permeate
is
acid.
As
is
shown
in
the
lower
part
of
Fig.
1
the
riboflavin
content
in
the
UF-
permeate
decreases
linearly
with
the
number
of
bed
volumes.
If
the
con-
tent
of
the
riboflavin
in
the
UF-permeate
is
about
25
to
30%
of
the
initial
quantity,
then
the
Syn
106
resin
has
adsorbed
about
200
mg
riboflavin
per
I
resin,
the
liquid
is
used
for
the
manufacture
of
lactose
by
crystallization.
In
Table
3
the
composition
of
lactose
produced
with
decalcification
and
decolorization
is
compared
with
lactose
made
of
un-
treated
UF-permeate.
After
decalcification
the
ash
content
of
the
lactose
is
lower
due
to
the
absence
of
calcium
phosphate.
As
the
lactose
crystals
formed
in
decalcified
UF-permeate
are
larger,
the
washing
process
seems
to
be
more
efficient,
which
results
in
a
lower
"protein"
content.
The
399
L
N
A
570
kg
UF-permeate
TS
5.03
%
N
0.025
%
A
0.47
%
93
GL
GA
20
N
A
RO
cation
hydro
ysis
—6-
anion
TS
11.0%
80
I
resin
pH
1.20
95
1
resin
(adsorbing
150
°
C13.0
min
(adsorbing
strong
acid)
weak
base)
97
97
GL
GL
GI
GA
J.
a
pH-regulation
15.01
36
kg
-4--
activated
-4—
evaporation
-4—
anion
-4--
cation
syrup
charcoal
TS
62.5%
0.1%
N
0.022
%
narrow-pore
A
0.10%
filter
merry-go-round
system
4.2.
Manufacture
of
a
hydrolyzed
lactose
syrup
Hydrolyzed
lactose
syrups,
which
are
comparable
in
sweetness
and
func-
tionality
to
high
DE
(dextrose
equivalent)
glucose
syrups,
can
be
used
in
a
number
of
applications
in
the
food
industry
as
a
sweetener,
preservative,
fermentation
enhancer
and
browning
agent.
In
the
developed
processes
ion
exchange
resins
are
applied
very
often,
either
for
demineralization
or
for
hydrolysis.
According
to
Bernhard
and
Hammett
(1952),
for
the
acid
hydrolysis
of
esters
lightly
cross-linked
ion
exchange
resins
of
the
poly-
styrene
sulfonic
acid
type
are
preferred.
Haggett
(1976),
Demaimay
et
al.
(1978)
and
MacBean
et
al.
(1979)
hydrolyzed
the
lactose
in
demineralized
UF-permeate
by
means
of
ion
exchange
resins
as
catalysts.
Fora
high
degree
of
hydrolysis
a
reaction
time
of
some
hours
at
about
95
°
C
was
necessary.
An
additional
treatment
with
activated
charcoal
is
desirable
for
the
removal
of
off-flavours
and
colour,
which
are
formed
during
hydrolysis.
MacBean
et
al.
(1978)
and
De
Boer
and
Robbertsen
(1981)
investigated
hydrolysis
of
lactose
in
decationized
UF-permeate
at
high
temperatures.
The
high-temperature
system
acid
hydrolysis
which
was
developed
by
the
latter
authors
is
given
schematically
in
Fig.
2.
The
UF-permeate
is
concentrated
twice
by
reverse
osmosis
and
the
liquid
made
to
pass
through
Fig.
2.
Process
scheme
for
the
manufacture
of
a
purified
hydrolysed
lactose
syrup
from
UF-permeate.
The
lactose,
nitrogen
and
ash
contents
(indicated
with
L,
N
and
AI
are
expressed
as
percentages
of
the
starting
liquid.
TS:
total
solids;
RO:
reverse
osmosis;
GL:
glucose;
GA:
galactose.
a
strong
acid
cation
exchange
resin.
Positive
ions
such
as
potassium,
sodium
and
calcium
were
exchanged
for
hydrogen
ions
and
the
pH
was
lowered.
The
higher
the
total
solids
content,
resulting
in
a
higher
cation
401
content,
the
lower
the
pH
of
the
UF-permeate.
In
a
special
heat
exchanger
connected
by
a
holding
tube,
jhe
lactose
in
the
UF-permeate
(pH
1.2)
was
hydrolyzed
to
80%
at
150
C
for
3
min.
The
residual
ions,
NPN
and
the
brown
colour
developed
as
a
result
of
Maillard
reactions
are
removed
to
a
great
extent
by
the
adsorbing
strong
acid
and
weak
bast
phenol
formaldehyde
resins.
After
the
treatment
with
the
selected
ion
exchange
resins,
which
were
arranged
in
a
merry-go-round
system,
the
liquid
was
colourless
and
the
ash
and
nitrogen
content
were
reduced
to
1%
and
10%
of
the
initial
quantity
respectively.
Since
during
evaporation
some
light
brown
colour
reappeared,
a
treatment
with
0.1%
activated
charcoal
was
necessary
to
achieve
a
colourless
end-product.
5.
Whey
protein
recovery
by
ion
exchange
resins
Recently
interesting
new
ion
exchangers
were
developed
for
the
recovery
of
whey
proteins.
The
obtained
end-product
is
very
pure,
the
protein
con-
tent
is
90-95%
and
it
is
nearly
free
of
fat
and
lactose.
The
very
low
fat
content
means
good
foaming
characteristics,
which
make
it
probably
suitable
for
the
replacement
of
egg
white
in,
for
example,
confectionery
products.
Jones
(1974)
described
ion
exchange
resins
based
on
carboxy-methyl
cellulose
(CMC)
which
can
adsorb
proteins,
including
whey
proteins.
This
socalled
"Vistec
process"
is
operated
in
a
stirred
reactor
containing
the
CMC
and
whey
at
pH
3.2
for
0.5
h
at
50
°
C.
The
partially
deproteinated
whey
is
drained
off
and
the
protein
eluted
from
the
CMC
at
pH
9.
The
product
was
concentrated
tenfold
by
ultrafiltration
to
a
concentration
of
about
10%
total
solids
and
subsequently
dried.
Coton
(1979)
carne
to
the
conclusion
that
under
the
conditions
mentioned
the
costs
were
high
because
only
about
half
the
protein
in
the
whey
was
recovered
and
the
extra
ultrafiltration
step
was
expensive.
Mirabel
(1978,
1981)
described
the
adsorbtion
of
whey
proteins
with
ion
exchange
silicas.
Porous
silica
beads
have
ion
exchange
groups
attached
and
can
therefore
adsorb
proteins
which
are
either
positively
or
negatively
charged
according
to
their
pH.
In
the
case
of
cheese
whey,
the
protein
is
largely
negatively
charged
and
therefore
adsorbs
onto
a
column
loaded
with
an
anion
exchange
silica
(Spherosil
QMA).
If
desired
the
immuno-
globulines
can
be
adsorbed
on
a
cation
exchange
resin
(Spherosil
X015).
The
yield
figures
of
this
process
seem
to
be
quite
favourable.
According
to
Burgess
(1981)
90%
of
the
whey
proteins
is
adsorbed
on
the
QMA
ex-
changer.
After
washing
the
adsorbed
protein
is
removed
from
the
column
by
elution
with
hydrochloric
acid
and
recovered
as
an
eluate
containing
about
6%
protein.
This
eluate
can
be
evaporated
or
ultrafiltrated
and
subsequently
dried.
In
France
this
process
was
recently
put
into
operation
on
an
industrial
scale.
References
Ahlgren,
R.M.
in
Industrial
Processing
with
Membranes
(Lacey,
R.E.
&
Loeb,
S.,
eds.)
pp
57-69,
Wiley-lnterscience,
New
York.
1972
Bernhard,
S.A.
&
Hammett,
L.P.
J.
of
Am.
Chem.
Soc.
75,
5834-5835.
1953
Boer
de,
R.
&
Robbertsen,
T.
Neth.
Milk
Dairy
J.
in
press.
1981
Bolto,
B.A.,
Eppinger,
K.H.,
Jackson,
M.B.
&
Siudak,
R.V.
Desalination
34,
171-188.
1980
402
Burgess,
K.J.
Food
Flavourings,
Ingredients,
Packaging
and
Processing
3,
nr.
3,
22-41.
1981
Calmon,
C.
&
Gold,
H.
in
Ion
Exchange
for
Pollution
Control,
CRC
Press
Inc.
USA.
1979
Coton,
G.
New
Zealand
J.
of
Dairy
Sci.
Techn.
14,
111.
1979
Delaney,
R.A.M.
Austr.
J.
of
Dairy
Tcchn.
31,
12-17.
1976
Delbeke,
R.
Neth.
Milk
Dairy
J.
26,
155-167.
1972
Delbeke,
R.
in
Symposium
Separation
Processes
by
Membranes,
Ion-exchange
and
Freeze-concentration
in
Food
Industry
D20
pp
1-21
APRIA,
Paris,
March
13
an
14.
1975
Delbeke,
R.
Le
Lait
55,
76-94.
1975
Delbeke,
R.
Neth.
Milk
Dairy
J.
33,
181-192.
1979
Demaimay,
M.,
Le
Henaff,
Y.
&
Printemps,
P.
Proc.
Biochem.
13,
April,
3-4.
1978
Felber,
E.H.
in
16th
General
Meeting
of
the
American
Society
of
Sugar
Beet
Technologists,
Denver,
Colorado,
February.
1970
Haggett,
T.O.R.
New
Zealand
J.
of
Dairy
Sci.
Techn.
11,
176-179.
1976
HervO,
D.
Proc
Biochem.
9,
February/March,
16-17.
1974,
a
Nerve,
D.
Proc.
Biochem.
9,
May
and
June.
1974,
b
Higgins,
J.J.,
Short,
J.L.
New
Zealand
J.
of
Dairy
Sci.
Techn.
15,
277-288.
1980
Hiraoka,
Y.,
Itoh,
K.
&
Taneya,
S.
Milchwissenschaft
34,
397-400.
1979
Hoeting,
W.A.G.
in
Connaissance
en
technologie
laitiere
moderne,
Frcsnes-les-
Montauban,
February
28
to
March
4.
1972
Houldsworth,
D.W.
J.
of
Soc.
of
Dairy
Techn.
33,
nr.
2,
45-51.
1980
laconelli,
W.B.
in
IFT
Annual
Meeting.
June
10-13,
Miami
Beach,
USA.
1973
Jenness,
R.
&
Koops,
J.
Neth.
Milk
Dairy
J.
16,
153-164.
1962
Johnston,
K.T.,
Hill,
C.G.
&
Amundson,
C.H.
J.
Food
Sci.
41,
770-777.
1976
Jones,
D.T.
Proc.
Biochem.
9,
December,
17-19,
1974
Jonsson,
H.
&
Forsman,
B.
Nordisk
Mejeriindustri
5,
636-641.
MacBean,
R.D.,
Willman,
N.J.
&
Hall,
R.J.
20th
Internat.
Dairy
Congress,
Paris,
947.
1978
MacBean,
R.D.,
Hall,
R.J.
&
Willman,
N.J.
Austr.
J.
of
Dairy
Techn.
34,
53-
59.
1979
Mintz,
M.S.
&
Shaffer,
L.H.
Electrodialysis
in
Principles
of
Desalination,
Ed.
Spiegler,
K.S.,
Academic
Press,
New
York.
1966
Mirabel,
B.
Ann.
Nutr.
Alim.
32,
243-253.
1978
Mirabel,
B.
Techn.
Laitiere
nr.
952,
March,
37-40.
1981
Okada,
K.
Tomita,
M.
&
Tamura,
Y.
Milchwissenschaft
32,
1-8.
1977
Parrish,
F.W.,
Sharpies,
P.M.,
Hoagland,
P.D.
&
Woychik,
J.H.
J.
of
Food
Sci.
44,
555-557.
1979
Short,
J.L.
&
Doughty,
R.K.
New
Zealand
J.
of
Dairy
Sci.
Techn.
12,
156-
159.
1977
Siegers,
G.
&
Martinola,
F.
Wasser
39,
377-388.
1972
D'Souza,
S.V.,
Lund,
D.B.
&
Amundson,
C.H.
J.
of
Food
Sci.
38,
519-523.
1973
403