International Journal of Biological Macromolecules 34 (2004) 4347

Non-equilibrium processes of interchain association induced

by Cs+ ions in -carrageenan aqueous solutions
F. Cuppo a , H. Reynaers a, , S. Paoletti b , S.P.B. Kremer c , J.A. Martens c
a

Department of Chemistry, Laboratory of Macromolecular Structural Chemistry, Catholic University of Leuven,

Celestijnenlaan 200F, B-3001 Heverlee, Belgium
b Department of Biochemistry, Biophysics and Macromolecular Chemistry, University of Trieste,
Via Giorgieri 1, I-34136 Trieste, Italy
c Centre for Surface Chemistry and Catalysis, Catholic University of Leuven,
Kasteelpark Arenberg 23, B-3001 Heverlee, Belgium

Abstract
We present preliminary results of the investigation of interchain association processes induced by Cs+ ions in -carrageenan aqueous
solution. The solutions contained variable amounts of NaI and CsI, under the condition that the total concentration of 1:1 electrolyte
was 0.1 M. The associative processes were observed by static light scattering under isothermal conditions (at T = 25 C), after cooling
molecularly dispersed solutions obtained at high temperature (80 C). It was found that, under all the investigated conditions of polymer
concentration (from 0.2 to 2 g l1 ) and ionic composition, the onset of time-dependent association fails to lead to an equilibrium, but
proceeds up to physical gelation of the associating system. Depending on the experimental variables, however, the gelation threshold may
take up to several days to be achieved.
2004 Elsevier B.V. All rights reserved.
Keywords: -Carrageenan; Light scattering; Intermolecular association; Gelation

1. Introduction
-Carrageenan is an industrially important anionic
polysaccharide, due to its property to form ionotropic, thermoreversible gels in aqueous environment. A vast literature
has been devoted to the characterisation of the properties
and structure of -carrageenan gels, especially by rheological and microscopic techniques. Among the most interesting results, it was established that -carrageenan gels
possess a fibrillar morphology, analogously to the majority
of physical gels of biological and synthetic nature [1,2].
On the other hand, in spite of notable experimental efforts,
the very basic molecular processes leading to interchain
association and gel formation by -carrageenan have been
a matter of lively debate in the past decades (see refs. [3,4]
and references therein), and several important aspects remain still unclear. Much of such controversy arose from
the discrepancy between experimental findings presented
by different laboratories involved in the research. In our

Corresponding author. Tel.: +32-16-327355; fax: +32-16-327990.

E-mail address: harry.reynaers@chem.kuleuven.ac.be (H. Reynaers).

0141-8130/$ see front matter 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijbiomac.2004.02.001

previous work [3,5], we stressed the fundamental role of

the preparative procedures in order to obtain reproducible
and reliable results when dealing with -carrageenan, due
to the extreme sensitivity of the biopolymer to the physicochemical parameters applied and, in particular, because
of the very marked tendency towards macromolecular aggregation. Recently [6], we demonstrated, by use of static
light scattering, the possibility to accurately control the
process of intermolecular association in -carrageenan solutions, ranging from conditions of very limited cluster
formation up to conditions of extensive, time-dependent association. To this purpose, we applied the mixed-counterion
method firstly employed by Borgstrm et al. [7]. It consists in preparing -carrageenan solutions in the presence
of mixed NaI and CsI, at variable Cs+ /Na+ molar ratio
and at fixed 0.1 M iodide concentration. In fact, it is well
established, from optical rotation measurements [8,9], that
-carrageenan takes on a helical conformation in 0.1 M NaI
at 25 C. In these conditions, and at low polymer concentration, no association between helices takes place, hence
the system is molecularly dispersed and in thermodynamic
equilibrium. However, the gradual replacement of Na+
ions with Cs+ ions progressively induces association of

44

F. Cuppo et al. / International Journal of Biological Macromolecules 34 (2004) 4347

the single-stranded helices of -carrageenan, phenomenon

that we could suitably observe by means of light scattering
experiments.
A remarkable aspect that turned out from our previous investigation was the very long time-dependence of the associative phenomena. It should be observed that, surprisingly
enough, the kinetic effects involved in the associative process have been very often neglected, or at least seriously
underestimated, in the previous research on -carrageenan.
To our knowledge, only the relatively recent work by Meunier et al. [10] was devoted to a better characterisation of
such aspects. On the other hand, several papers dealing with
the associative behaviour of -carrageenan appeared in the
last years in the literature, where the importance of kinetic
effects might not have been properly recognised. For example, Ueda et al. [11] studied by SEC-LALLS the molar mass
distribution of -carrageenan in KCl, in conditions where
the polymer aggregates. The measurements were performed
immediately after cooling molecularly dispersed solutions
obtained at high temperature (60 C). As already noted by
Meunier et al. [10], it is clear that in those conditions only
the rather small aggregates formed in the very early stages
of the aggregation processes were analysed. It is worth mentioning, also, the paper by Wittgren et al. [12], devoted to the
investigation of the associative behaviour of -carrageenan
by flow field-flow fractionation coupled to multiangle light
scattering. In the latter case, the authors used an equilibration time of 24 h prior to injection of the samples into the
experimental apparatus, therefore, assuming that the aggregation process would reach an equilibrium, corresponding
to a distribution of polymer clusters of limited size, within
that period. It is important to note that in the latter study the
experiments were performed under physicochemical conditions (temperature, ionic composition and molar mass of the
polymer sample) thoroughly analogous to the ones applied
in our present investigation.
In this communication we present new results of the
light scattering investigation of association and gelation
processes of -carrageenan in mixed NaI/CsI solutions.
More particularly, we wish to address here the issue of
whether an equilibrium state may eventually be achieved in
the course of such processes. We will show that, under all
the presently investigated conditions (including a constant
value of weight-average molar mass in the order of 105 ),
the onset of interchain association does not lead to any
equilibrium, but unavoidably proceeds up to physical gelation of the macromolecular systems. However, depending
on the experimental parameters, the gelation threshold may
take up to several days to be achieved.

2. Materials and methods

-Carrageenan (Sigma, type III, lot no. 127H1222)
of weight-average molar mass Mw = 1.79(0.02)
105 g mol1 (as determined by wide-angle light scattering)

was purified as previously described [6]. Aqueous solutions in mixed NaI/CsI salt (with constant 0.1 M iodide
concentration and varying cesium mole fraction, the latter being defined as XCs = [CsI]/([CsI] + [NaI]) were
prepared according to a standard procedure [6]. A peristaltic pump in a closed circuit was used to filter the
solutions (Millipore membrane, pore size 0.22 m) directly in the light scattering cell. Static light scattering
measurements were performed on a back-scattering set-up
(ALV-NIBS/HPPS, ALV, Langen, Germany), operating at
angle = 173 , at wavelength = 632.8 nm (He/Ne laser,
with output power of 3 mW). The temperature control system of the light scattering apparatus was used to perform a
pre-heating stage before each measurement. The solutions
were heated from 25 to 80 C, kept at 80 C for 1 h and finally cooled to 25 C. The heating and cooling steps lasted
13(1) min and 16(1) min, respectively. The scattered intensity was measured under isothermal conditions (25 C),
every 3 or 10 min depending on the total duration of the
experiment.

3. Results and discussion

A typical plot of scattered intensity versus time, obtained
for an associating system of -carrageenan (at polymer concentration Cp = 0.6 g l1 and cesium fraction XCs = 0.4)
is shown in Fig. 1. It can be noted that a continuous increase of scattered intensity takes place for about 14 h from
the beginning of the experiment. Such an increase is clearly
due to the process of intermolecular association, with formation of progressively larger polymer clusters in the system. However, a visible change occurs at t 14 h, corresponding to the onset of large fluctuations of the experimental signal. Such intensity fluctuations have been reported
for several gelling systems of various chemical nature [13].
They indicate a drastic slowing down of diffusion dynamics, which gives rise to non-ergodic scattering (speckle
pattern). The phenomenon should be interpreted as the occurrence of a connectivity transition in the associating system, therefore, denoting the achievement of the gelation
threshold.
We obtained similar curves at several values of polymer
concentration (at constant cesium fraction, XCs = 0.4), in
the range 0.22 g l1 . They are collectively shown, in logarithmic scale, in Fig. 2. The initial smooth increase of
scattered intensity indicates that association takes place in
all cases, although the kinetics of the process is strongly
dependent on the polymer concentration. Most remarkably,
in no case the association process comes to an equilibrium,
which would be revealed by a levelling off of the intensity
curves towards a plateau. Analogous observations on associating -carrageenan (where the association process was
induced by the presence of K+ ions) were reported by Meunier et al. [10]. In that case, however, the investigation was
limited to the initial conditions of ergodic scattering, and

F. Cuppo et al. / International Journal of Biological Macromolecules 34 (2004) 4347

45

Fig. 1. Time evolution of the scattered intensity from a sample of -carrageenan, at polymer concentration Cp = 0.6 g l1 and cesium fraction XCs = 0.4.
The onset of large fluctuations of the experimental signal, at t 14 h, denotes the achievement of the gelation threshold.

the gelation threshold was only achieved in a single case, at

the highest concentration investigated by the authors. Therefore, it remained unclear whether an equilibrium would be
achieved at some time in the systems at lower concentration,
where the experiments were interrupted (at about 24 h from
the onset of the association processes) still in the presence
of a steady increase of scattered intensity. We should note

that our results have been obtained in different conditions of

temperature and ionic composition with respect to the investigations by Meunier et al. [10]. The main difference, from
the molecular point of view, consists in the helical content
of the associating -carrageenan molecules. In the experiments by Meunier et al. [10] the helical content was comparatively low, and however, dependent on the temperature

Fig. 2. Time evolution of the scattered intensity from -carrageenan samples, at different values of polymer concentration ranging from 0.2 to 2 g l1
and at constant cesium fraction XCs = 0.4. The double-log scale has been used for a better collective representation of the several datasets. The inset
shows the data at the lowest polymer concentration (Cp = 0.2 g l1 ) in linear scale.

46

F. Cuppo et al. / International Journal of Biological Macromolecules 34 (2004) 4347

at which each experiment was performed. On the contrary,

we decided to work in conditions under which association
takes place between fully ordered chains (namely, with a helical content higher than 90% [8,9]). By doing so, we aimed
at keeping the conformational and the associative properties of -carrageenan well separate. However, at least at a
qualitative level, our results well compare to those by Meunier et al. [10]. The data shown in Fig. 2 indicate that no equilibrium is attained even by decreasing the polymer concentration down to 0.2 g l1 at XCs = 0.4. In all cases the intensity curves come to the onset of visible fluctuations, that is,
the gelation threshold is achieved. It should be remarked that
such a process may require a very long time, up to about 80 h
at the lowest investigated value of polymer concentration
(see inset in Fig. 2). We found it useful to report the gelation
time, defined as the onset of large fluctuations of the scattered intensity signal, as a function of the polymer concentration. The resulting plot is given in Fig. 3, where the strong
concentration dependence of the gelation time is highlighted.
It is interesting to note that the double-log representation of
the experimental data (see inset in Fig. 3) suggests the presence of scaling behaviour of the gelation time as a function of
polymer concentration. This latter issue, which has been previously addressed by several authors on the basis of both theoretical and experimental evidence (see, for example, refs.
[14,15]), will be treated in deeper detail in a forthcoming
paper.
Due to the limits of sensitivity of the experimental apparatus and because of the very long duration of such experiments, we decided not to extend our investigation to even
lower values of polymer concentration. Rather, we found it
interesting to perform similar experiments by keeping the

polymer concentration constant, at 1 g l1 , and by decreasing the cesium fraction, XCs , in the range 0.4 to 0.0. In fact,
as noted in the Introduction, a decrease of XCs involves a
reduction of the thermodynamic driving force leading to association of -carrageenan. The results are given in Fig. 4.
They show that a reduction of XCs from 0.4 to 0.3 strongly
affects the kinetics of the association process, although the
gelation threshold is achieved in both cases. However, a further reduction to XCs = 0.2 prevents the occurrence of any
visible time-dependent association on a time scale of about
40 h. Therefore, the three datasets in the range of XCs from
0.0 to 0.2 appear as a flat baseline, denoting the substantial
stability of the systems in these conditions. No intermediate
behaviour appears between the systems at XCs 0.3, where
physical gelation is achieved at the end of the slow association processes, and the systems at XCs 0.2, where no
time-dependent association appears to be detectable at all.
We should note that in a previous study [6] we found evidence for the occurrence of very limited association even at
XCs values of 0.1 and 0.2. The two findings, however, do not
conflict, given the much higher sensitivity to the presence of
very little association offered by the experimental method
used in the previous investigation. With respect to this, it
should be emphasised that the remarkable advantage of the
method presently employed consists in the possibility to
observe continuously, and without introducing any physical
disturbance, the slow evolution of the macromolecular systems in the course of association and physical gelation. From
this viewpoint such an approach, although providing mostly
qualitative information, offers unique advantages compared
to any other commonly used experimental technique (e.g.
rheology, ultracentrifugation, fractionation techniques, etc.).

Fig. 3. Gelation time vs. polymer concentration, at constant XCs = 0.4. The inset shows the data in double-log scale.

F. Cuppo et al. / International Journal of Biological Macromolecules 34 (2004) 4347

47

Fig. 4. Cesium fraction dependence of the associative behaviour of -carrageenan, at constant polymer concentration Cp = 1 g l1 .

The observation of the data in Fig. 4 leads to two important considerations: (i) the ionic composition plays a critical
role in the associative behaviour of -carrageenan. The increase of XCs produces a sharp transition from conditions
of no (or very little) association to conditions of extensive
association, with long time dependence eventually leading
to physical gelation. Such a transition takes place within a
narrow range of ionic composition (from XCs = 0.2 to 0.3
in the present conditions of Mw , polymer concentration,
ionic strength and temperature). This result is in agreement
with our previous findings [6] and partially confirms the observations of other investigators [12,16]. (ii) Also in the case
of the cesium fraction dependence, we failed to find suitable
conditions for the associative processes to attain an equilibrium without involving physical gelation of the systems.
This is in analogy to the case of the polymer concentration
dependence (Fig. 2). When significant time dependence
appears, in fact, intermolecular association unavoidably
proceeds up to the achievement of the gelation threshold.
As a final comment, it should be noted that even the gel
phase produced at the gelation threshold certainly does not
constitute an equilibrium state, since the achievement of
the connectivity transition merely represents an intermediate state in the evolution of the associating systems. The
very presence of slow fluctuations of the speckle pattern
denotes the continuous internal evolution and structural
rearrangement of the gel network. Through very slow dynamics, indeed, the system attempts to approach a truly
thermodynamically stable state, which remains, however,
unattainable on an experimentally relevant time scale.

Acknowledgements
H.R. and F.C. are indebted to FWO-Vlaanderen, INTAS
(00-243) and the Catholic University of Leuven for supporting this research and for a research grant (F.C.).

References
[1] Hermansson AM, Eriksson E, Jordansson E. Carbohydr Polym
1991;16:297.
[2] Guenet JM. J Rheol 2000;44:947.
[3] Bongaerts K, Reynaers H, Zanetti F, Paoletti S. Macromolecules
1999;32:675.
[4] Viebke C, Borgstrm J, Piculell L. Carbohydr Polym 1995;27:145.
[5] Bongaerts K, Reynaers H, Zanetti F, Paoletti S. Macromolecules
1999;32:683.
[6] Cuppo F, Reynaers H, Paoletti S. Macromolecules 2002;35:539.
[7] Borgstrm J, Piculell L, Viebke C, Talmon Y. Int J Biol Macromol
1996;18:223.
[8] Slootmaekers D, De Jonghe C, Reynaers H, Varkevisser FA, Bloys
van Treslong CJ. Int J Biol Macromol 1988;10:160.
[9] Ciancia M, Milas M, Rinaudo M. Int J Biol Macromol 1997;20:35.
[10] Meunier V, Nicolai T, Durand D. Macromolecules 2000;33:2497.
[11] Ueda K, Itoh M, Matsuzaki Y, Ochiai H, Imamura A. Macromolecules 1998;31:675.
[12] Wittgren B, Borgstrm J, Piculell L, Wahlund KG. Biopolymers
1998;45:85.
[13] Shibayama M, Norisuye T. Bull Chem Soc Jpn 2002;75:641.
[14] Gimel JC, Durand D, Nicolai T. Phys Rev B 1995;51:11348.
[15] Clark AH, Kavanagh GM, Ross-Murphy SB. Food Hydrocol
2001;15:383.
[16] Viebke C, Borgstrm J, Carlsson I, Piculell L, Williams P. Macromolecules 1998;31:1833.