Effect of pH and Neutral Salts upon the Swelling of

Cellulose Gels
J. GRIGNON and A. M. SCALLAN, P u l p and Paper Research Institute of
Canada, Pointe Claire, Quebec, Canada

Synopsis
The swelling of cellulose gel is considered to be caused by an osmotic pressure differential resulting
from a difference in concentration of mobile ions between the interior of the gel and the exterior
solution. The fundamental feature within the gel, bringing about the unequal distribution, is the
presence of ionizing acidic groups attached to the macromolecular network. Using the Donnan theory
to describe the distribution of ions, it is shown that the experimentally observed decreases in swelling
brought about by extremes of pH and by salt addition may be explained in a semiquantitative way.
The swelling behavior of superabsorbent gels appears to be almost completely governed by the
proposed mechanism. For conventional wood pulps, the mechanism describes the smaller variations
in swelling caused by pH and salt addition. For these gels, electrolytic effects are only responsible
for a fraction of the total swelling-the remainder being associated with the macromolecular network
itself. The unequal distribution of hydrogen ions between a gel and an exterior solution is described
briefly, along with the effect that this has on the potentiometric titration of the gel.

INTRODUCTION
In a recent article, we have reviewed some of the extensive literature on the
observed effects of electrolytes upon the swelling of cellulosic gels. In accom-
panying experimental studies, we confirmed the important role played by the
acidic groups in the gels and the nature of their counter i0ns.l Most particularly,
we showed that if the mixture of counter ions normally present in a wood pulp
was removed and replaced by a single species, the pulp was increasingly swollen
in the order AP+ < H+ < Mg2+ < Ca2+< Li+ < Na+. In the present article, we
examine the effect of pH upon the swelling of cellulose gels. Among other
phenomena, a change in pH can bring about the conversion of the acidic groups
from the free acid to salt form (and vice versa) with accompanying changes in
swelling. In addition, because of the important role it plays during pH changes,
we study the simultaneous presence of a neutral salt.
The most systematic study of the effect of pH and added neutral salt on the
swelling of wood pulps was recently carried out by Lindstrom and Carlsson2:
these authors showed clearly that the swelling of wood pulps is maximum at about
neutrality and is depressed a t both low and high pH values. A t all pH values,
swelling was further reduced by the presence of salt. The effects are considerably
more exaggerated in the superabsorbent pulp^.^^^ Lepoutre, Hui, and Robertson?
increased the degree of swelling of a bleached kraft pulp some thirtyfold by
grafting onto the fibers a polymer rich in acid groups. They noted that while
the swelling remained constantly high in the pH range of 9-5, it dropped dras-
tically when pH was lowered further, reaching the swelling level of the unmodified
pulp a t pH 3.5. The swelling was also reduced to about half when the neutral
product was placed in 1%sodium chloride solution rather than water.

Journal of Applied Polymer Science, Vol. 25,2829-2843 (1980)

0 1980 John Wiley & Sons, Inc. 002 1-8995/80/0025-2829$01.50
2830 GRIGNON AND SCALLAN

A detailed explanation of this is possible, we feel, using the concepts on the

mechanism of swelling of gels originally proposed by P r ~ c t o r .He ~ suggested
that the important characteristic of many gels is the presence of ionizable groups
attached to the macromolecular network. When such a gel is placed in an ionic
solution, an exchange of mobile ions takes place between the gel and the solution,
but because of the presence of the bound groups, the mobile ions are unevenly
distributed-the gel contains the higher concentration. Proctor suggested that
the gel swells as a result of the entry of water in an attempt to reduce the osmotic
pressure differential resulting from the difference in concentration. Swelling,
he proposed, continues until the osmotic pressure differential is equal to the
resistance to further expansion brought about by the cohesive forces of the
macromolecular network. Using the theory of D ~ n n a nto ~ .describe
~ the dis-
tribution of ions, Proctor and Wilsons explained the swelling behavior of gelation
as a function of pH, accurately pinpointing the pH of maximum swelling.
Application of Proctor's theory to cellulose is largely confined to the work of
Neale.gJ0 To explain the swelling of cellulose in sodium hydroxide solution,
Neale considered the hydroxyl groups as very weak acids capable of forming
sodium salts. In this way, he was able to account for the swelling behavior on
a semiquantitative basis. Because of the very low dissociation constant of the
hydroxyl groups, conversion to the sodium salt. and subsequent swelling neces-
sarily takes place at very high alkalinities. Farrar and NealelO also observed
variations in the swelling of cellophane and cotton placed in ionic solutions of
different concentrations at near neutral pH. Here they considered the carboxylic
acid groups to be the active species. Farrar and Neale, however, only tested their
hypothesis so far as to confirm that in dilute solutions, the distribution of ions
between the gel and external solution was in accord with the Donnan theory for
the number of carboxylic acid groups present.
In the present article, we extend the early work of Proctor and Neale using
essentially their type of approach. As a result of the concentrated effort that
has more recently been put into the field of synthetic polymers, more sophisti-
cated treatments of the swelling of polyelectrolyte gels now exist. However, as
stated by Pennings and Prins in a recent critical appraisal of Neale's work, the
older approach does bring out the salient features of swelling behavior without
a great loss of rigor." In addition, the approach is susceptible to easy compre-
hension and relates to readily measured parameters. In the absence of previous
theoretical studies of the swelling of cellulose in the pH range where the ionizing
groups are provided by the acid groups rather than the hydroxyl groups, we thus
feel this further application of the technique is justifiable.

THEORY
Our theoretical treatment follows the manner of those of Donnan, Proctor,
and subsequent workers. In brief, we consider a gel containing bound acidic
groups that has swollen to equilibrium in an aqueous solution of given pH and
neutral salt concentration. As the acidic groups cannot move out of the gel, the
solution within the gel is regarded as separated from the external solution by a
semipermeable membrane which confines the acidic groups but gives free passage
to water and all simple ions. All mobile ions are assumed to be present to some
extent on either side of the membrane and, by taking into account the various
 EFFECTS OF pH/NEUTRAL SALTS ON CELLULOSE GELS 2831

chemical equilibriums that must exist on either side and between the two sides,
the interrelationships of concentration are derived. From these, the excess
concentration of mobile ions within the gel is calculated. This excess of con-
centration E is directly proportional to the osmotic pressure differential between
the two sides of the membrane, which, in turn, we assume is directly proportional
to the degree of swelling achieved. Thus, the net result of the theoretical cal-
culations is E as a function of pH and neutral salt concentration; and this will
subsequently be compared with experimental data of swelling as a function of
the same parameters.
Model
The treatment will be restricted to the case where any given set of conditions
is achieved by an acid, base, and salt derived from a particular monovalent cation
and a particular monovalent anion. As the ions used in experimental work were
sodium and chloride, these ions will be used in the theory. In Figure 1,the hy-
pothetical membrane is represented as a broken line separating the solution
within the gel from the external solution; and expressions have been given to the
concentrations of the various ions in the following sequence. Inside the gel, the
acid groups are assumed to have a concentration c and a degree of dissociation
N , thus yielding a concentration c (1 - a ) in undissociated form (RH) and CCY in
dissociated form (R-). Symbols have arbitrarily been chosen to represent the
concentration of H+ and C1- on either side of the membrane. Having defined
the concentrations of hydrogen ions, the hydroxyl-ion concentrations immedi-
ately follow from the hydrolytic constant of water. Electrical neutrality must
exist in each solution and in order to satisfy this condition, the concentration
term given the sodium ion is that necessary to ensure that there are equal con-
centrations of positive and negative ions in each solution.
We will now suppose that the acid groups behave similarly to a soluble acid
and that their degree of dissociation is governed by a dissociation constant de-
fined in the usual way:

Introducing the appropriate concentration terms from Figure 1and rearranging,

we obtain the concentration of ionized groups:

GEL EXTERNAL
PHASE SOLUTION

I
RH c(1-a) I
R- cu I
H+ Y 1 H+ X

OH- KW/Y I OH- Kw/x

Na+ K,/y+n+ca-y I Na+ K,/x+m-x

CI - n I CI- rn

Fig. 1. Representation of the equilibriums existing when a gel is placed in contact with a solution.
RH and R- refer to the undissociated and dissociated acid groups, respectively.
2832 GRIGNON AND SCALLAN

Having thus satisfied in the model the various equilibriums which must exist
on either side of the membrane, we may now consider the equilibriums between
the two sides.

Distribution of Ions

According to the Donnan theory, the presence of a nondiffusible ion on one

side of a membrane affects the partition of all diffusible ions. This is effectively
the case with an electrolytic gel and an exterior solution. Using as an approxi-
mation the concentrations of the various ions rather than the activities, the ap-
propriate Donnan equilibrium for the model is:

where g and s refer to the gel and the external solution, respectively. X is a dis-
tribution constant defined by the ratios. Substituting from Figure 1:

A = - Y= K,ly + n + C N - y - K,Jx - m
(2)
x KJx + m - x K,ly n
Adding the first two ratios:

A=
KJy + +
n CCY
KJx +m
and substituting Ax for y , and An for m , we have
Cff
(3)
K,,ly +n
This equation, in conjunction with the expression for c a [eq. (l)], gives the dis-
tribution constant in terms of the hydrogen-ion and chloride-ion concentrations
within the gel. The derivation is in the manner used by Hitchcock in his con-
siderations of membrane equilibrium in general12;and for n = 0, eq. (3) reduces
to his’equation for the equilibrium between a nondiffusible acid and a base in
the absence of added salt.
It will be noted that, in the absence of chloride ions (n = 0), X decreases with
increasing hydroxyl concentration (K,,/y ) tending from an initially very high
value to unity in the limit. At any hydroxyl concentration, concentrations of
chloride approaching the hydroxyl concentration bring about a drastic reduction
in A.
Since X is always greater than unity, positive ions are more concentrated in
the gel and negative ions more concentrated in the exterior solution. This of
course also applies to the hydrogen-ion concentration and we have by taking
logarithms of the relationship X = y / x :

That is, the pH in the solution is always greater than that in the gel and as X can
be very large, the difference in pH can be several units.
 EFFECTS OF pH/NEUTRAL SALTS ON CELLULOSE GELS 2833

Calculation of the Degree of Swelling

The difference in concentration of all diffusible ions E between the gel and
the external solution is given by:
E = [H+],: + [OH-], + “a+], + [Cl-1,
- [H+Is - [OH-], - “a+], - [Cl-1,
From this, by introducing terms from Figure 1 and using X to convert external
concentrations, we obtain
E =CN+ 2(Kw/y + n ) ( l - A)
which by use of eq. (3) to eliminate ( K w l y + n ) reduces to

Our research has shown that this simple relationship has not been previously
reported. E is directly proportional to the osmotic pressure differential as stated
in van’t Hoff equation; and, following Proctor’s original proposal, we will also
accept E as in turn proportional to the degree of swelling. I t will be noted that
E approximates to C N when X is high.
Swelling due to electrolytic effects is a manifestation primarily brought about
by the concentration of nondiffusible ions within the gel which, in turn, is a
function of the internal pH. The equations governing swelling, therefore, most
naturally relate to the conditions within the gel. In using the equations, we
specified the concentration of acid groups, the acid dissociation constant, and
the concentration of chloride ions within the gel, and then for a series of postu-
lated internal pH values calculated successively a,A, and E. External concen-
trations yielding the set internal concentrations were calculated once X was
known. The equations may be rearranged to express swelling directly in terms
of conditions external to the gel; but the resultant equations are quite complex.
Table I shows sets of calculations for the absence and presence of the chloride
ion. The chloride-ion concentration can be taken as a measure of the sodium
chloride concentration a t all but the lowest pH values where a small part of the
chloride is associated with the hydrochloric acid necessary to achieve the set pH.
A concentration of acid groups of 0.1 equivll. was chosen for the calculations as
this corresponds to the experimental values for the swollen pulps.

Swelling in the Absence of Salt

Figures 2 and 3 shows the variation in swelling with pH for gels containing the
same concentration of acid groups but different dissociation constants. The
shapes of the curves may be understood by considering the principal ions present
under any given conditions.
First, consider a gel in free-acid form allowed to equilibrate with water. The
most abundant ions in the system are the hydrogen ions dissociated by the acid
groups and these will remain in the gel to preserve electrical neutrality. There
being no ions in the external ‘solution’ other than hydrogen and hydroxyl re-
sulting from the dissociation of water, the external solution will exhibit a pH of
7.0, and its total ion concentration will be very low. The concentration of ions
within the gel in excess of the external concentration E will thus approximate
2834 GRIGNON AND SCALLAN

TABLE I
Sample Calculations of the Degree of Swelling of a Gel Containing Weak Acid Groups in the
Absence and Presence of Salt”

c = 0,1000, K, = 1 X [Cl-1, = 0 (no salt case)

2.5 0.0031 0.0000 31343.00 7.0 0.0000 0.0000 0.0030
3.0 0.0091 0.0081 30151.OO 7.5 0.0000 10-7 0.0091
4.0 0.0500 0.0499 22360.00 8.3 0.0000 10-6 0.0500
5.0 0.0909 0.0910 9534.00 9.0 0.0000 10-5 0.0909
6.0 0.0990 0.0990 3147.00 9.5 0.0000 10-5 0.0989
7.0 0.0999 0.0999 999.00 10.0 0.0000 0.0001 0.0997
8.0 0.1000 0.1000 316.00 10.5 0.0000 0.0003 0.0994
9.0 0.1000 0.1000 100.00 11.0 0.0000 0.0010 0.0980
10.0 0.1000 0.1001 31.64 11.5 0.0000 0.0032 0.0939
11.0 0.1000 0.1010 10.04 12.0 0.0000 0.0101 0.0819
12.0 0.1000 0.1100 3.32 13.0 0.0000 0.0332 0.0537
13.0 0.1000 0.2000 1.41 13.2 0.0000 0.1414 0.0172

c = 0.1000, = 1 x 10-4, C1-1, = 0.01 (salt case)

2.5 0.0031 0.0099 1.14 2.6 0.0114 0.0087 0.0002
3.0 0.0500 0.0181 1.38 3.1 0.0138 0.0131 0.0015
4.0 0.0909 0.0599 2.45 4.4 0.0245 0.0245 0.0210
5.0 0.0990 0.1008 3.18 5.5 0.0318 0.0318 0.0474
6.0 0.0999 0.1090 3.30 6.5 0.0330 0.0330 0.0530
7.0 0.1000 0.1099 3.31 7.5 0.0332 0.0331 0.0536
8.0 0.1000 0.1100 3.32 8.5 0.0332 0.0332 0.0536
9.0 0.1000 0.1100 3.31 9.5 0.0332 0.0332 0.0536
10.0 0.1000 0.1100 3.30 10.5 0.0330 0.0333 0.0535
11.0 0.1000 0.1110 3.18 11.5 0.0318 0.0349 0.0521
12.0 0.1000 0.1200 2.45 12.4 0.0245 0.0490 0.0420
13.0 0.1000 0.2100 1.38 13.1 0.0138 0.1520 0.0160
a Numbers rounded off from computer calculations.
to C N or [H+],, which in turn will be determined by the acid dissociation constant.
Thus, if the acid groups are strong, the gel will exhibit high swelling at low pH,.
If the acid groups are weak, the gel will have a low degree of swelling at high
PH,.
With the progressive neutralization of the acid groups, it can be seen, from
eq. (3) and the example given in Table I, that X is reduced but still remains high.
Most of the sodium of the added caustic soda, therefore, enters the gel and E
continues to approximate to C N ,the only appreciable quantity of ions affecting
swelling are the sodium and hydrogen counter ions of the dissociated acid groups.
Maximum swelling is achieved upon complete ionization.
Beyond a pH, of 10, X is reduced rapidly. There is an increasingly even dis-
tribution of ions with an appreciable quantity of ions accumulating outside the
gel. Swelling is, therefore, reduced and, as all acid groups are now dissociated,
the decrease is independent of the dissociation constant.

Swelling in the Presence of Salt

Figures 2 and 3 also show the degree of swelling as a function of pH, in solutions
of different concentrations of salt.
In the presence of small concentrations of salt, the rising section of the swelling
 EFFECTS OF pH/NEUTRAL SALTS ON CELLULOSE GELS 2835

0.10

0.08

0033M NaCl

ui

0.02

014 M NoCl

0 L I I I I I I ,

0 2 4 6 8 lo 12 14

SOLUTION pH

Fig. 2. Theoretical plots of E (which is taken as proportional to the degree of swelling) as a function
of pH and external salt concentration. The gel contains 0.1 equiv/l. of strong acid groups ( K , =
0.1 mol/l.).

curve, as the gel is neutralized, is displaced to lower external pH values; however,

swelling rises to almost the same maximum. As can be seen from eq. (3), the
internal concentration of chloride ions need only be of the order of the hydroxyl
concentration ( K J y ) to bring about an appreciable reduction in A. The dis-
tribution constant at low chloride concentrations, however, remains sufficientiy
high that E still approximates c a y . Thus, the swelling for a given internal pH

0.10

0.02

0
0 2 4 6 8 10 12 14

SOLUTION pH
Fig. 3. Theoretical plot as for Figure 2 but for a gel containing weak acid groups ( K , = 1 X
mol/l.).
2836 GRIGNON AND SCALLAN

is still close to that in the absence of chloride, however, the external pH a t which
the swelling is observed is reduced considerably [eq. (4)]. In qualitative terms,
the salt may be regarded as permitting the transport of hydrogen ions out of the
gel.
The addition of high concentrations of salt causes X to approach unity and,
in addition to the internal and external pH values approaching equality, swelling
is much reduced. A t a certain level of salt addition, the chloride ions actually
have more control on swelling than pH. Thus from eq. (3):

A=++= n

where n >> K J y . The condition where the chloride concentration exceeds the
hydroxyl concentration is readily achieved in practice with salt solutions which
are close to neutrality and quite dilute.

Titration of Gels
That the pH within a gel is not necessarily equal to that of the external solution
is an interesting manifestation of the Donnan theory. The effect that the phe-
nomenon has on the potentiometric titration of a gel comes directly from the
calculations of swelling as a function of pH. Figure 4 shows the theoretical curves
for the titration of a weak acid in the presence of various concentrations of salt.
Taking logarithms of eq. (l),we have the usual relationship between pH of a
soluble acid of low molecular weight and its degree of neutralization

However, the pH which would be measured in the titration of a gel would be

that of exterior solution. Introducing this expression into eq. (4):
12
I

10

Ip
6

0 0.2 0.4 0.6 0.8 1.0

DEGREE OF DISSOCIATION, o(
Fig. 4. Theoretical plot of the internal pH and the accompanying external pH values in salt so-
lutions of different concentrations as a gel is progressively neutralized. The curves are for a gel
containing weakly acidic groups ( K O =
 EFFECTS OF pH/NEUTRAL SALTS ON CELLULOSE GELS 2837

Thus, if the pK, of a cellulose gel is estimated (as is usual for a soluble acid) from
the pH of half neutralization, the pK, would be too high by an amount loglo
X-an appreciable amount in the absence of salt.

EXPERIMENTAL

Materials
The superabsorbent pulp was a carboxymethylated cotton crosslinked in the
wet state with epi~hlorohydrin.~I t was received in dry sodium form from
Buckeye Cellulose Corp. Wood pulps were prepared from red sprucewood (Picea
rubens) by conventional pulping procedures; variations in pulp yield being ob-
tained by varying the cooking time. The pulps reported in this study were kraft
pulp of 49% yield and sulfite pulp of 77% yield. The high yield in the latter case
was adopted to obtain a pulp which contained a high concentration of acid groups.
Wood pulps were kept in a “never-dried” condition prior to treatment.
Chemicals used throughout were of reagent grade and all water used was
deionized to a conductivity of less than 0.3 X mho cm-’.

Swelling Procedures
The “as-received” superabsorbent pulp was soaked at 0.5% consistency in
either water or a salt solution of the set concentration, and the pH was adjusted
with either sodium hydroxide or hydrochloric acid.
Wood pulps were first freed of their usual mixture of cations and converted
to the acid form by repeated washing with 0.1M hydrochloric acid and were then
washed with deionized water until free of the chloride i0n.l The pulps were then
made to 0.5% consistency in water or salt solution and the pH adjusted by the
addition of sodium hydroxide or hydrochloric acid.
All pulps were allowed to equilibrate overnight in sealed containers and the
final pH values noted before the measurement of swelling.

Measurement of Swelling
The degrees of swelling reported are the water retention values determined
in the swelling medium. The technique used and its validity as a measure of
swelling has been previously reported.13
For a number of samples the degree of swelling was also determined by the
more absolute solute-exclusion technique.14. This procedure was also carried
out in the swelling medium using a solution of dextran of molecular weight 2 X
lo6 adjusted to the same pH and salt concentration as the swelling medium.
Changes in dextran concentration were monitored by precision polarimetry. The
degrees of swelling thus obtained agreed well with the water retention values.
Both techniques give the degree of the swelling of a pulp expressed as the ratio
by weight of water to solid material in the swollen gel.
2838 GRIGNON AND SCALLAN

Titration of Pulps
Pulp, 2.0 g converted to its acid form, as described earlier, was dispersed in
200 ml of water or salt solution. The suspension was then titrated slowly under
a nitrogen atmosphere using standardized sodium hydroxide solution. The ti-
tration was followed potentiometrically.
The acid contents of the pulps were: superabsorbent 3.8 mequivlg, sulfite
pulp 0.30 mequivlg, kraft pulp 0.09 mequiv/g.

RESULTS AND DISCUSSION

Swelling Behavior of Pulps

The cellulosic fibers most clearly demonstrating the effects discussed are the
superabsorbent pulps. The high swelling potential of these pulps appears at-
tributable to their augmented content of acid groups.
Figure 5 shows the swelling behavior of a carboxymethylated cotton as a
function of pH and added salt. The results follow closely predicted trends. The
rise in swelling occurs at the pH values expected of weakly acidic groups, a plateau
region occurs around neutrality and the decrease in swelling commences a t pH
10, as predicted for acid groups of any strength. A lowering of the whole curve
occurs with added salt. Not coinciding with expectation is the curve for swelling
in the'nominal absence of salt which appears to correspond to a theoretical curve
for the presence of a small amount of salt. It is probable that this is caused by
the experimental difficulty of reducing salt concentrations to the very low levels
required by the theory.
The data that we have gathered on wood pulps together with that published
by Lindstrom and Carlsson2 show that these pulps behave as expected in the
theory (Fig. 6). However, the effects are slighter than those observed on the

40

30 .
W
m
YI

3
s
Z
20.

0
I-
f
I-
w
LL

OL 10.
c
w

2
0
0 2 4 6 8 10 12 14

SOLUTION pH

Fig. 5. Experimentally determined degree of swelling of a superabsorbent pulp as a function of

the pH and salt concentration of the external solution.
 EFFECTS OF pH/NEUTRAL SALTS ON CELLULOSE GELS 2839

w 20 -
w
\

w
3
3
z
0
15 -
I-
Z
w
c i I

CL 1.0 -
I-w

s
0.51 I I I c 1 I I I , 1 1 1

0 2 4 6 8 10 12 14

SOLUTION pH
Fig. 6. Experimentally determined degrees of swelling of sulfite and kraft wood pulps as a function
of pH in the presence and absence of salt. ( 0 )Kraft with water; ( 0 )kraft with 0.25M NaCI; ( W )
sulfite with water; (0)sulfite with 0.25M NaCl.

superabsorbents and appear superimposed upon a basic degree of swelling which

is independent of electrolytic conditions. It is possible that such a basic swelling
exists in the case of the superabsorbents but is only a small fraction of the total
swelling. In considerations of the swelling of ion-exchange resins, Gregor has
found that these too appear to have a “rest volume” when the osmotic differential
due to electrolytic effects is 2 e r 0 . ~ ~The
J ~ volume may be interpretated as the
intrinsic swelling ability of the macrmolecular network itself once free of acidic
groups.16
At neutral pH and in the nominal absence of salt, the swelling of the super-
absorbent pulp is 34 X l./g and, as the pulp contains 4 X 10+ equivlg of acid
groups, c a is of the order 0.1 equivh. Conventional pulps swell to a lesser extent,
and have fewer acid groups, but c a is of the same order. Since, for neutral pH
and the absence of salt, we have equated ca with the degree of swelling, clearly
there is some other property which must be taken into account when comparing
different gels. This has been termed the elastic modulus of the The first
water entering a dry gel placed in water may be imagined as forming an extremely
concentrated solution of the available ions; and hence, a high driving pressure
for swelling. However, as the gel swells, the interior solution will become more
dilute and the osmotic pressure less. At the same time, if we consider the gel
matrix to behave elastically in the manner of Hooke’s law, the resistance of the
gel to further expansion will increase with swelling. The position of equilibrium
between the two forces will then depend upon the elastic modulus. For ion-
exchange resins, Gregor and Frederick equated the elastic modulus with the
degree of crosslinking; the less the crosslinking, the greater the swelling of a gel
under a given osmotic pressure.16 A difference in modulus due to crosslinking
might well be responsible for the difference in swelling behavior between the
conventional and superabsorbent pulps. The volume of the swollen cell walls
2840 GRIGNON AND SCALLAN

of pulp fibers is close to that of the original wood fibers,17and it is possible that
much of the original cohesive strength remains, the source of this strength being
undisturbed intermicrofibrillar contacts.Is On the other hand, during the
preparation of the superabsorbent pulps, a forceful expansion could take place
such that many of these crosslinks could be irreversibly broken and the fibers
henceforth be more deformable structures. The lack of cohesive strength in the
superabsorbents is apparent in the ease with which the fibers disintegrate under
mechanical action.lg
While the theoretical plots bring out the main features of swelling behavior,
the predicted reductions in swelling for the extremes of pH or added salt are
somewhat more drastic than experimentally observed. This is apparent in the
quantitative comparison of theory and experiment in Figure 7. The major cause
of this discrepancy is probably that, in common with Neale in his studies at high
pH, we have assumed that the concentration c a remains constant throughout
all swelling and contraction. The fixed value we have chosen is that of the
swollen gel. However, ca is itself a function of the degree of swelling and may
be expected to increase with contraction of the gel. The effect will be to increase
both X and E over what we have calculated. While realizing this, we have not
yet found it possible to introduce the necessary correction into usable equa-
tions.

Titration Curves of Pulps

In Figures 8 and 9, the experimental titration curves of a superabsorbent pulp
and a wood pulp carried out in the presence and nominal absence of salt are given.
Similar curves also have been reported with synthetic ion-exchange resins where
the ideal case for no salt is not observed.20 With the superabsorbent, the ap-

1.0

(3 0.8
z

Ly
>
-
I-
4
Y 0.2

EXTERNAL CONC. NoCl (moledl)

Fig. 7. A comparison of theory and experiment for the fractional reduction in swelling of the su-
perabsorbent pulp brought about by increasing concentrations of salt a t near neutral pH. The
lractional reductions are expressed as EIEo and WRVIWRVo where Eo (theoretical) and WRVo
(experimental) refer to the degrees of swelling in water.
 EFFECTS OF pH/NEUTRAL SALTS ON CELLULOSE GELS 2841

12, , , , , , , , , , , , ,

0 0.2 0.4 0.6 0.8 1.0 1.2

(NOOHADDED)/ (NOOH FOR NEUTRALlZATlONl

Fig. 8. The potentiometric titration curves of the superabsorbent pulp in the presence and absence
of salt.

parent pK, drops from over 6.0 to 3.75 with the addition of 0.25M NaC1; 3.75
being a more reasonable value for acidic groups of carboxymethylated cellulose.21
Chowdhury and Neale have reported a value of 3.42 for the soluble material.21
Kraft pulps in which the acid groups are also carboxylic exhibit a similar pK,
in the presence of sufficient salt, while high-yield sulfite pulps have a lower pK,
in keeping with the greater strength of the sulfonic acid groups present.
More rigorous treatments of the potentiometric titration curves of polyelec-
” ~it~has been shown that taking
trolytes exist22and continue to be r e f i r ~ e d ~but
into account the Donnan effects explains a large fraction of the anomalous be-
havior of these gels.22
12

10

(NOOHADDED) / (NoCm F O R NEUTRALIZATION)

Fig. 9. The potentiometric titration curve of the sulfite pulp in the presence and ahsence of
salt.
2842 GRIGNON AND SCALLAN

Implications of These Studies

The results presented in this article compliment existing studies in showing
the profound effect of pH and salt upon the swelling of the superabsorbent pulps.
I t has been known for some time that such pulps must be in neutralized form for
maximum absorption, and the reduced absorption at low pH has recently been
exploited to produce an economical drying process.25 That the absorption of
these pulps for 1%sodium chloride solution is a fraction of that for water was
probably a surprising and somewhat disappointing initial finding, since such a
solution simulates the physiological fluids to which certain absorbency products
are exposed. All these effects are, however, completely in keeping with the
mechanism we have proposed.
Although conventional wood pulps show less response to the effects of pH and
added salt, the small changes observed are of industrial significance. Small
increases in swelling bring about appreciable increases in the plasticity of the
fibers and their ability to bond extensively during sheet formation. We have
already shown that converting the acid groups of sulfite and kraft pulps from
the free acid to the sodium salt results in 20-30% increase in the tensile strength
of the paper made from them. Conversion to salts of divalent and trivalent
metals, however, bring about less swelling and lower paper strength. Part of
the raison d’Btre for the current work was to investigate what additional effects
we may expect if, after neutralization, the pH is raised even further. While most
paper is prepared at about a pH of 5.0, there are reported advantages to using
higher and even alkaline pH values. If the proportionality between swelling and
papermaking properties is accepted, our results suggest that no further change
in paper properties should be expected for a wide range of pH following neu-
tralization of the acid groups. This is by no means clear from studies of pH in
the literature where no attention is given to the nature of the cations present in
the pulp or in the aqueous medium. Thus, a pulp in the commonly occurring
calcium form may lose strength if it is made acid by the addition of alum and gain
strength if it is made alkaline by the use of sodium hydroxide. However, where
the pH is not greatly removed from neutrality, these are the effects of counter-ion
exchange rather than the effect of pH. In our study, we have deliberately in-
volved only one cation to isolate the effects of pH from those of ion exchange.
A trend in papermaking is towards increased recycling of drainage waters,
hence there is growing interest in knowing what effects the higher electrolyte
concentrations have upon paper properties. For a salt of a monovalent cation,
our studies show the order of concentration which will bring about a reduction
of swelling. Again, however, in extrapolating the effects we have found to mill
situations, the nature as well as the concentration of recycled cations needs to
be considered.
If the concepts put forward are to be accepted, it is apparent that the con-
centration of ions in a pulp suspension is not necessarily that inside the fiber wall
in the absence of appreciable ionic strength. This is not generally taken into
account in studies of the kinetics of cellulose reactions. However, if h is high,
any cationic reactant will be more concentrated at the location of the reaction
inside the cell wall than in the external solution. On the other hand, anionic
reactants will be largely confined to the external solution. These phenomena
are of increased importance in laboratory studies where, unlike most industrial
processes, there is a tendency to use pure reagents and deionized water.
 EFFECTS OF pH/NEUTRAL SALTS ON CELLULOSE GELS 2843

References
1. A. M. Scallan and J. Grignon, Suensk. Papperstidn., 82 (2), 40 (1979).
2. T. Lindstrom and G. Carlsson, Proc. Eucepa Conf., Warsaw, Poland (1978).
3. P. Lepoutre, S. H. Hui, and A. A. Robertson, J . Appl. Polyrn. Sci., 17,3143 (1973).
4. W. L. Dean and G. N. Ferguson, U.S. Pat. 3,589,364 (1971).
5. H. R. Proctor, J . Chem. Soc., 105,313 (1914).
6. F. G. Donnan and A. B. Harris, J . Chem. Soc., 99,1554 (1911).
7. F. G. Donnan,Z. Electrochem., 17,572 (1911).
8. H. R. Proctor and J. A. Wilson, J . Chem. Soc., 109,307 (1916).
9. (a) S. M. Neale, J . Textile Inst., 20,373 (1929); (b) 21,225 (1930).
10. J. Farrar and S. M. Neale, J . Colloid Sci., 7, 186 (1952).
11. A. J. Pennings and W. Prins, J . Polym. Sci., 58,229 (1962).
12. D. I. Hitchcock, J . Gen. Physiol., 9,97 (1926).
13. A. M. Scallan and J. E. Carles, Suensk. Papperstidn., 75,699 (1972).
14. J. E. Stone and A. M. Scallan, Cellul. Chem. Technol., 3,343 (1968).
15. P. H. Gregor, J . Am. Chem. Soc., 73,642 (1951).
16. P. H. Gregor and M. Frederick, Ann. N . Y . Acad. Sci., 57,87 (1953).
17. J. E. Stone and A. M. Scallan, Pulp Paper Mag. Can., 69,288 (1968).
18. A. M. Scallan, Wood Sci., 6 (3), 266 (1974).
19. P,Lepoutre and A. A. Robertson, TAPPI, 57 (10). 87 (1974).
20. R. Kunin, Ion Exchange Resins, Wiley, New York, 1958, p. 34.
21. F. H. Chowdhury and S. M. Neale, J . Pc~lym.Sci. Part A , 1,2881 (1963).
22. I. Michaeli and A. Katchalsky, J . Polyrn. Sci., 23,683 (1957).
23. J. Hasa, M. Ilavsky, and K. Dusek, J . Polyrn. Sci., Polyrn. Phys. Ed., 13,253 (1975).
24. J. Hasa and M. Ilavsky, J . Polyrn. Sci., Polyrn. Phys. Ed., 13,263 (1975).
25. P. Lepoutre, in Solvent Spun Rayon, Modified Cellulose Fibres and Deriuatiues, A. F. Turbak,
Ed., ACS Symp. Ser. No. 58, Washington, D.C., p. 217 (1977).

Received June 28,1979

Accepted January 16,1980