Studies on chitosan as an admixture for cement-based materials: assessment of its

viscosity enhancing effect and complexing ability for heavy metals

M. Lasheras-Zubiate, I. Navarro-Blasco, J.M. Fernndez, J.I. Alvarez*

Department of Chemistry and Soil Sciences, School of Sciences, University of Navarra,

c/ Irunlarrea, 1, 31080, Pamplona, Spain

*Dr. Jos Ignacio Alvarez Galindo

Dpto. de Qumica
Facultad de Ciencias
Universidad de Navarra
C/ Irunlarrea, 1
31.080 Pamplona (Navarra)
Spain
Phone: 34 948 425600
Fax: 34 948 425740
E-mail: [email protected]

Studies on chitosan as an admixture for cement-based materials: assessment of its

viscosity enhancing effect and complexing ability for heavy metals

Abstract
Chitosans of different molecular weights were assessed as cement admixtures in order
to modify the rheological properties of a fresh cement paste. Their sequestering
capability with respect to several heavy and toxic metals was checked and the
complexation processes were investigated by different approaches based on
electrochemical measurements. The aim was to conduct research on a potentially useful
polymer to be added to cementitious materials acting as a viscosity modifier as well as a
heavy metal immobilising agent. Pb, Cu, Zn, Cd, Mo and Cr were used as target metals.
The highest complexation affinity was found towards Zn, with stoichiometries up to
1:40 having been calculated. Chitosan was proved to act as a thickener in cement
mixtures. The highest molecular weight chitosan showed the greatest effectiveness
taking into account its combined role as heavy metal retainer and thickener. In addition,
a clear interaction between this high molecular weight chitosan and cement particles
was also demonstrated by means of zeta potential measurements.

Keywords
metal-polymer complexes, viscosity, additives, composites, electrochemistry

INTRODUCTION
The use of viscosity-enhancing admixtures (VEAs) has been shown to be efficient in
improving the stability and cohesion of cement-based materials. VEAs usually increase
the viscosity of the fresh cement paste, thus reducing the risk of the separation of the
heterogeneous constituents of a cement-based mortar or concrete.1 Especially when
used for rendering, VEAs also prevent cement mortars from sagging. They have also
been reported as anti-washout admixtures for underwater concrete placement. At the
same time that they show a viscosity-enhancing behaviour, these additives might act as
set-retarders, water-reducers, and water-retainers.2 Regarding the chemical composition
of these admixtures, polysaccharides can be quoted as some of the most widely used
products.3-5 Sugars, starches, and cellulose ethers have been studied and their effect
when added to cement-based materials reported.6
Cement materials have also shown the ability to act as effective agents towards the
solidification/stabilisation of hazardous waste substances, especially heavy metals
(studies on Cr(VI) and divalent Pb, Cd, Ni and Hg have been published).7 Cement
functions as a non or less hazardous deposit for unwanted by-products and rejects.
Besides low cost, availability and versatility are additional advantages for cement-based
systems.8
Certain polymers, present in cement as typical additives (such as CMC and HEC), show
a good adsorption of heavy metals and may act as effective removing agents from
polluted waters and untreated sludges.9,10 Understanding of metal retention in polymercement composites requires a previous knowledge of the mechanisms involved in
adsorption and fixation of those metals by the components of the material. Apart from
some interaction with certain components of cement, complexation processes may be
expected to occur between metallic species and the added polymer.11 Voltammetric
studies have shown usefulness to elucidate complexation behaviour in aqueous
solutions, providing relatively simple approaches to the estimation of polymer-metal
stability constant, stoichiometry of formed complex and complexing capacity of the
ligand (i.e., the amount of metal that the polymer may uptake and stabilise in the
matrix).12 This set of data is very useful when determining what amount of metal can be
added to the cement matrix and, at the same time, will serve in the modelling of the
leaching patterns that may follow these metals when solidified/stabilised on the
polymer-cement.13

Chitosan, the deacetylated derivative of chitin, is the second most abundant natural
polysaccharide. This polymer and its derivatives have been used as biomaterials because
of their biocompatibility, biodegradability, and biological activities.14,15 Purified
qualities of chitosans have been used as additive for many biocements, dental
biocements, and cements for reparation of bones. 16 In these composites, chitosan has
been reported i) to enhance injectability of these composites; ii) to increase setting time
in a similar way to cellulose derivatives in cement-polymer systems,17 and iii) to act as a
thickener - i.e. a cohesion-improving agent - in calcium phosphate cement for the repair
of bones.18
Furthermore, chitosan has shown its successful role as heavy metal retainer. Some
recent studies have pointed to its excellent capacity to act as an effective complexing
and removing agent for heavy metals.19-21 The binding mechanism of metal ions onto
chitosan is not yet fully understood; mechanisms such as adsorption, ion exchange and
ion chelation have so far been proposed.22,23 The hydroxyl groups and amine sites in the
polymeric chain are the more active binding sites for metal ions, establishing dative
bonds with transition metals.24
Literature is rather lacking on the behaviour of chitosan and its derivatives as cement
admixtures for building purposes. The strong resemblance that chitosan bears to other
polysaccharides like cellulose endorses the interest of these studies. In addition, as
mentioned above, chitosan and its derivatives have shown a clear capacity for chelating
heavy metals. The rationale for this work is that chitosans -or derivatives- would be
potential admixtures (rheological modifiers) in order to obtain polymer-modified
cements with heavy metals retaining and/or immobilizing properties.
As a preliminary study devoted to assessing the aforementioned potential use of
chitosan, the present paper intends i) to test whether most toxic metals -Cu, Zn, Cd, Pb,
Cr and Mo- undergo complexation with chitosan, and ii) to evaluate the behaviour in
fresh state of cement mortars modified by chitosan addition.

EXPERIMENTAL
Voltammetric studies on chitosan complexing ability of heavy metals
Chitosans with different molecular weights [low molecular weight, 50-190 kDa
(LMW), Ref. 448869-50G; medium molecular weight, 190-310 kDa (MMW), Ref.
448877-50G; and high molecular weight, 310-375 kDa (HMW), Ref. 419419-50G] and
different deacetylation degrees (92%, 80% y 76%, respectively) used in this study are

commercial products from Sigma-Aldrich (Steinheim, Germany). The MW and DDA

data were supplied by the manufacturer, and their validity was subsequently checked
according to previously published procedures.25,26 Acetic acid and sodium acetate were
purchased from Merck (Darmstadt, Germany). The Certipure standard solutions
containing 1000 ppm of the heavy metals, Copper (Cu(NO3)2 in HNO3 2-3%), Zinc
(Zn(NO3)2 in HNO3 2-3%), Cadmium (Cd(NO3)2 in HNO3 2-3%), Lead (Pb(NO3)2 in
HNO3 2-3%), Chromium (K2CrO4 in H2O) and Molybdenum (NH4)6Mo7O24 in H2O)
were supplied by Merck. Chitosan and metal solutions were dissolved in 0.44 M acetic0.04 M acetate buffer solution (pH = 4.0). The optimum concentration of ligand for the
determinations was found to be 0.5 M. Lower concentrations than 0.1 M showed a
very weak signal and values higher than 1 M caused the saturation of the electrode.
Voltammetric measurements, both differential pulse anodic stripping voltammetry
(DPASV) and cathodic differential pulse voltammetry (DPV), were performed with a
Metrohm 746 VA Trace Analyzer coupled with a 747 VA Stand equipped with a static
mercury drop electrode (SMDE). A conventional three-electrode arrangement
consisting of a glassy carbon counter electrode, an Ag/AgCl/KCl 3M reference
electrode and Metrohm multimode mercury electrode were used. Before each titration,
the procedure involved the transfer of the solution to the cell and purging with nitrogen
for 10 min. The conditions used for each metal are specified in Table I. All the
complexes were proved to be formed quasi-instantaneously, so that a minimum
conditioning time (4 min.) was allowed after every spike of metal.
When the assayed metals amalgamated onto the mercury electrode, a known amount of
ligand was placed in the electrochemical cell and then increasing amounts of Cu, Zn, Cd
and Pb, were spiked on the solution. On the other hand, when metals did not
amalgamate onto the electrode, as in the case of Cr and Mo, increasing amounts of
ligand were added to a solution containing a fixed metal concentration.

Assessment of the fresh state behaviour of cement mortars modified by chitosan

addition
Materials
An ordinary Portland cement (OPC) (CEM II 32,5 N, supplied by Portland S.A.
Olazaguta, Spain) and a standardised siliceous aggregate (mineralogical composition as
well as grain size distribution can be found in Izaguirre et al.27) were used to prepare the
mortars. The selected binder: aggregate ratio (B:Ag) was 1:3, by weight. Instead of

volume proportions, weight proportions were selected in order to avoid any imprecise
measurement. The three different commercial chitosans previously reported were used
as admixtures. One only additive was incorporated in each one of the remaining
mortars, in order to check clearly its possible efficiency, avoiding other variables. For
each chitosan, a dosage of 0.1% of the total dried mortars weight was incorporated.

Mortars preparation
The mixer was a Proeti ETI 26.0072. 450 g of cement, 1350 g of aggregate and additive
(when necessary) were blended for 90 seconds at low speed. The water was added in
accordance with a 0.5 water/cement ratio, which makes it possible to achieve suitable
workability, avoiding any excess of water.
Water was then added and mixed for 90 more seconds at low speed. Mortars settled for
10 minutes before being tested, to let the additive take effect. Straight afterwards, fresh
state properties were determined as described below.

Fresh state properties

The consistency and water-retention capacity of the fresh state of cement mortars were
determined.28,29 The first one was carried out by means of the flow table test. This
method is based on the measurement of the spread of slurry placed into a cone-shaped
mould. When the mould, filled with fresh mortar, is vertically lifted, 15 knocks of the
table take place for 15s, and the slump is determined.
Water-retention capacity was determined by weighing absorbent materials placed on the
fresh sample before and after 5min of contact under pressure, as reported elsewhere.27

Viscosity measurements
In order to assess any rheological change, viscosities of cement pastes modified by
chitosan addition were determined. To this end, a HAAKE Viscotester VT 550, with a
programme of rotation speeds varying from 1.5 to 510 s-1 during 5.5 minutes was used.

Particle size distribution

As reported by Chiou et al.30, Khayat31, and Saric-Coric et al.32, fixation of water,
interaction between cement particles and polymer molecules, and intertwining process
between polymer chains are expected to produce agglomerates, hence particle size
distribution of cement-polymer mixtures should change.

20 g of cement were added to 100 g of solutions with different amounts of chitosan in

water, the resulting chitosan/cement ratios ranging from 0% to 1%. The particle size
distribution of these samples was determined using a Malvern Mastersizer (Malvern
Instruments, Ltd., U.K.).

Conductivity
Set retarding ability of different polysaccharides used in cement pastes has been
reported to be determined by means of conductivity measurement.2 A Thermo Orion 4Star benchtop pH/conductivity meter equipped with a Dura Probe 4-electrode
conductivity cell (Orion 013605MD) with a nominal cell constant of 0.55 cm-1 intended
for a conductivity measurement range from 10 S cm-1 to 200 mS cm-1, was used for
conductivity assays.
Different cement pastes were prepared by mixing 5 g of cement and 0.05 g of each one
of the tested chitosans with 100 mL of ultrapure water previously thermostated at 25 C.
The experiments were carried out in a reaction vessel magnetically stirred inside a water
bath thermostatically adjusted at 25 C. The conductivity measurements were recorded
each 10 minutes for 24 hours. These data were used in order to draw the conductivity
curve.

Zeta potential measurements

Specific mixtures of cement, water and additive were prepared to assess the zeta
potential. The determination of zeta potential of solid-liquid interface permitted us to
study the adsorption of the admixture on the solid surface of binder, since it modifies
this property, as stated by Papo and Piani,33 and Zhang et al.34,35
Following the method proposed by Zhang et al.34,35 and assayed previously 36 solutions
with different concentrations of chitosans were prepared (percentages w/w of 0.025,
0.05, 0.1, 0.15, 0.2 and 0.25), and cement was added with a solute/liquid ratio of
0.0025. After 5 minutes of stirring, the mixtures were separated by centrifuge, and the
supernatants were analysed using a Zeta Potential Analyzer ZETA PLUS (Brookhaven
Instruments Corporation, New York, U.S.A.). The average of ten measurements was
regarded as the zeta potential of the cement particles.

RESULTS AND DISCUSSION

Voltammetric studies on chitosan complexing ability of heavy metals
Electrochemical study of Pb, Cd and Zn
The electrochemical approach consisted of the titration of a given amount of chitosan
(acting as complexing ligand) with increasing amounts of added standard metals. All
three metals are able to amalgamate -upon reduction at suitable electrode potential- on
the mercury working electrode. Subsequent scanning of the potential in the positive
going sense forces the reoxidation of the metal, giving rise to a current that is monitored
in the Differential Pulse mode. Depending on the previous situation of the metal in
solution (either free or chitosan-bound), the amount of metal to be deposited on the
mercury electrode and subsequently reoxidized will differ. Thus, if chitosan is in excess
of the stoichiometry of the possible complex with a given metal, most of the metal
added to the chitosan-containing solution will be stabilized by the ligand (chitosan) and
it will amalgamate on the mercury electrode to a much lesser extent with respect to the
amount of deposit that would be achieved when free. This results in diverse patterns
when current vs. metal concentration are plotted. In general, two linear portions are
obtained: a first one with a lower slope, corresponding to the anodic stripping current
measured for the respective metal when complexed by the chitosan in solution, and a
second steeper branch indicative of free metal in solution, that is to say, when in excess
of the stoichiometric ratio of the chitosan-metal complex. The break in this type of
graph allows the stoichiometry to be estimated and the population of chitosan acting as
ligand to be calculated.
The background electrolyte has to be carefully selected in order to ensure that it does
not play any competitive complexing role with respect to the main ligand under
observation. An acetic acid/acetate buffer solution of pH 4 was used, guaranteeing, on
one hand, the prevalence of ionic state of metals

37

and, on the other hand, the absence

of any significant complexing activity towards the metals, as previously reported.38

Besides, at this pH protonation of the amino groups (and the consequent reduction of
the complexation capacity) is prevented, which could not be ensured with lower pH
values. Finally, in this buffered solution the precipitation of both chitosan and metal
hydroxides is precluded.
Stripping voltammograms were recorded for Pb, Cd and Zn when accumulated on the
mercury electrode at the optimal potentials. Titration data were obtained for all of them
by plotting their respective stripping peak currents vs. increasing concentrations of the

added metals as shown in Fig. 1. In all instances, the voltammograms showed a constant
potential peak at -379, -560 and -840 mV for Pb, Cd and Zn, respectively, the height of
which increased with increasing metal concentrations. A surprising feature observed
was the fact that the Zn stripping peak in presence of chitosan showed, in the first place,
a marked decrease in its intensity compared to that of the free metal and, in the second
place, a displacement of its peak potential towards less negative values (from -960 mV
to -840 mV in presence of chitosan).
Stoichiometries were calculated from titration data collected for the three assayed
molecular weights of chitosan and three different concentrations (0.1, 0.5 and 1 M) of
a fixed molecular weight chitosan (MMW), as presented in Table II.
As can be seen, the higher the molecular weight of chitosan, the larger the binding
capacity observed for Pb and Zn. On the other hand, the highest complexation capacities
were found for both Zn and Pb when more diluted chitosan solutions were assayed,
irrespective of the MW of the chitosan studied. It is also noteworthy that whatever the
MW of chitosan, the coordination of Zn is always larger than that of Pb.
These results can be explained bearing in mind the structure of chitosan. A higher
molecular weight implies a larger number of available amino groups to bind the metals,
which explains the first assertion. As for the second, dilution of chitosan implies a
corresponding diminishing of the aggregation of the polymer and the steric hindrances,
so that metals can be easily placed between sheets of the polymer. Furthermore, the
breaking of stabilizing hydrogen bonds within the structure of the diluted chitosan
contributes to this trend. Finally, ionic radii of both Zn and Pb differ significantly (0.74
and 1.20 , respectively) so that inclusion of Zn into the polymer is sterically more
favourable. Moreover, the complex of the polymer with this metal seems to have a
different geometry to the binding with the other two metals.39
Linearization algorithms of titration data are usually employed to estimate affinity
constants of complexes, such as those developed by Buffle
1:1 stoichiometries, both Scatchard

42

and Langmuir

43

40

and Ruzic.

41

For simple

algorithms are straightforward,

allowing the disclosure of differently behaving ligands. Experimental data of Pb and Zn

were processed according to Scatchard and Langmuir in spite of the fact that the
stoichiometry of the complexes is far from 1:1.
Equations [1] and [2] are obtained from these mathematical treatments:

ML K L K ML
T
M
M M 1
ML L T K L T

[1]
[2]

where [M] is the free metal concentration, [ML] is the complexed metal concentration
obtained by subtracting [M] from the total metal concentration [MT]; LT is the total
ligand concentration as calculated from these expressions and K is the conditional
stability constant derived from the slope and from the ordinate respective plots.
Equation [1] is commonly known as the Scatchard equation. The Langmuir linearization
(equation [2]), also called Ruzic/van den Berg linearization method, usually gives rise
to a linear relationship when the ratio [M]/ [ML] is plotted against [M]. The slope is
equal to 1/LT and the intercept yields 1/KLT.
Surprisingly, both metals adjust well to these theoretical patterns. Although calculations
of total available chitosan cannot properly resemble the real amount of ligand present in
the electrochemical cell, the data obtained corroborate the above described trends. The
calculated effective concentration of chitosan in every instance increases with its ability
to link the metal (stoichiometry) and a linear tendency (R2 = 0.9905 and 0.9175 for Pb
and Zn respectively) is observed. Even when using the Scatchard algorithm, a single
straight line was always achieved, indicating the existence of a unique type of complex
with these two metals.
As for Cd (Fig. 1b) its behaviour differs from that of Pb and Zn; chitosan ability to
complex Cd is lower (see Table II) and either chitosan concentration or MW influences
the stoichiometry variations to a much lesser extent.

Electrochemical study of Cu
Copper is, amongst the heavy metals, the most troublesome to determine
voltammetrically due to its low solubility in mercury, the coincidence of its stripping
peak potential with the potential interval in which many organic species are adsorbed on
the mercury electrode and the presence of a pre-wave.44
The assayed concentrations of both ligand and metal were lower than those used for the
previous metals in order to avoid the saturation of the electrode. Experiments were
carried out in the presence of the fixed 0.05 M chitosan and stripping voltammograms
were recorded for increasing Cu concentrations.

10

The peaks obtained in this case were difficult to evaluate due to their broadening,
showing a pre-wave that is not observed in free metal solutions. For this reason, the
correction of the base line is critical to obtain reliable results. It is also worth
mentioning that the peak potential of the metal in presence of chitosan suffers a
displacement towards more positive potentials (ca. 38 mV) in comparison with the peak
potential of free Cu, confirming that the complexation occurs.
An interesting aspect inferred from the titration curve (Fig. 1d) is that this binding
seems to be independent of the molecular weight of chitosan, as the concentration of
metal that saturates the ligand is almost the same no matter what the chitosan
characteristics are. It has been previously reported that Cu (as well as Zn) interacts
differently with chitosan from other metals39 and a square-planar or tetrahedral
geometry has been previously proposed for the complex.45 According to our results, an
inter-layer binding mechanism is proposed as the main retention factor of Cu by
chitosan, so that cavity between the sheets of the polymer would be more relevant than
length of the chains.

Electrochemical study of Mo and Cr

Since both Mo and Cr do not amalgamate on Hg electrodes, their complexation by
chitosan was followed-up in a more traditional way, consisting of the monitoring of
their reduction peak upon increasing concentrations of ligand (chitosan). Each metal
showed a diverse pattern so that attention has been focused on either peak intensity or
peak potential for Mo and Cr, respectively. It must be taken into account that, unlike the
previous metals, Mo and Cr were added to the solutions as anionic forms, as reported in
the experimental section.
Mo peak intensity decreased regularly as chitosan concentration in cell increased up to a
point at which the peak intensity remained constant upon addition of chitosan (Fig. 2a).
Mo peak potential did not suffer any displacement in the presence of chitosan. As has
been reported in the literature, different Mo chemical species co-exist in solution at the
selected pH.46 Besides, we have, in this case, detected that complexation of Mo by the
background electrolyte takes place to a certain extent, giving rise to non-linear
responses of Mo peak intensities vs. Mo concentration in acetate buffer. Nevertheless,
when chitosan is present in the solution, it turns out to be the most important
complexing agent, displacing the acetate, as proved by the fact that Mo peak intensity
signal decreases linearly in the presence of increasing concentrations of chitosan.

11

Furthermore, this effect is more pronounced as the molecular weight of chitosan

increases. The indent of Fig. 2a shows the linear trend of this latter effect.
On the other hand, Cr reduction peak potential was more negative when in the presence
of chitosan. This displacement was more acute once the added chitosan was enough to
fully bind the initial Cr in solution. The degree of displacement is related to the strength
of the retention of the metal.
Experiments carried out at different polymer MW showed that the higher the molecular
weight, the larger the retention of Cr as measured in terms of peak potential
displacement, as shown in Fig. 2b.
Since electrochemical approaches were different for these two metals with respect to the
one used before -for Zn, Cd, Pb and Cu- absolute numerical figures cannot be compared
for the two groups of metals; however, from breaks in the linear portions of the
respective titration curves obtained for both Mo and Cr (Fig. 2) it can be concluded that
chitosan is able to retain 7-fold more Cr than Mo.
As has been emphasized before, as the polymer chain increases, there is a greater
number of linking groups available for the retention of the metals within the threedimensional structure of the acting ligand. This fact is consistent with a predominant
intra-chain linkage of both metal atoms rather than the binding between the sheets
formed by the different chitosan chains.47
Different models have been previously proposed to elucidate the chitosan-metal
linkages. One of them is the bridge model in which the metallic ion is coordinated
with several nitrogen atoms from the same or different chains of chitosan.48 The other
model proposed is called the pendant model, which considers that the metal is bound
to an amino group of the chain as a pendant (Fig. 3).45 For this reason, depending on the
characteristics of the metal, some metal ions could bind only a helical loop of polymer,
while others might correlate with several helical loops.39 In our case, all the metals
under study seem to bind with several polymer chains as the stoichiometry is always
superior to 1:1.

Assessment of the fresh state behaviour of cement mortars modified by chitosan

addition
Table III shows the slump determined through the flow table test. As can be seen, the
chitosan addition increases the viscosity of the fresh cement mortars, giving rise to a
drop in fluidity. Taking into account the slump value of the control mortar (OPC

12

without admixture), LMW chitosan led to a reduction of 9.7% in terms of fluidity;

medium and high molecular weight chitosans caused a decrease of 17.4% and 20.3%,
respectively. It might be expected that the addition of a higher molecular weight
polymer resulted in a greater thickening effect, owing to the higher possibility of
entanglement between polymer chains.49 The obtained results confirmed the
expectations.
Apparent viscosity measurements performed on cement pastes (1:1 water/cement ratio)
modified by the addition of the chitosan polymer fit in with the flow table test results
(Fig. 4 shows the apparent viscosity values of cement pastes with chitosan additions as a
function of an increasing shear rate). Cement pastes with HMW chitosan showed the
highest apparent viscosity at low shear rates. All the tested polymers conferred a shearthinning behaviour on the cement pastes, which has been reported to be very useful for
many practical functions (such as rendering, grouts, and shotcrete). A relatively large
viscosity was determined at low shear rates but it showed a reduction when great stress
was applied. Actually, materials of pseudoplastic nature are supposed to resist
sedimentation and sagging at rest but can be easily mixed, pumped and sprayed (during
a wet-mix or dry-mix process).50 The fall in viscosity when the stress is applied can
be explained because linkages between polymer and cement particles may break, and
linear chains of the polymer could align with the flow direction. When the stress slows
down, a recoagulation process could take place.51
In order to understand the fluidity decrease, the insolubility of the chitosan in alkaline
pH (a pH of 12.12 was obtained from a solution of 5% w/w of the binding material), as
well as in calcium-rich systems, must also be taken into account.52 This high sensitivity
towards alkaline pH and free Ca2+ ions of polysaccharide molecules has been previously
reported by Goycoolea et al.53 and Knaus and Bauer-Heim54, respectively. To a certain
extent, chitosans might precipitate in such alkaline and calcium ion rich matrices, giving
rise to an entanglement between different polymer chains, thus increasing the viscosity.
A cross-linking phenomenon owing to the interaction with calcium ions could also be
responsible for a viscosity rise.18 In addition, OH and NH2 groups would undergo
ionization under alkaline conditions, becoming the negatively charged O- and NHgroups. Given that LMW chitosan showed the largest degree of deacetylation, this
polymer presented the highest number of negatively charged groups. As a consequence,
its charge density should be larger and responsible for the appearance of electrostatic

13

repulsions between chains of the polymer, which inhibited intermolecular association

and hence reduced viscosity. 34
To sum up, both the lower molecular weight of the LMW chitosan and the appearance
of electrostatic repulsions between chains in this polymer could be responsible for its
lower fluidity reduction. Based on these considerations, a similar trend in terms of
fluidity can be proved with respect to DDA for MMW and HMW chitosans. The lower
the DDA and the higher the MW, the larger the viscosity of the cement solutions was.
Water retention ability of these samples (Table III) was also assessed. All the modified
samples retained larger percentages of mixed water than the control mortar. These data
can be related to the DDA. The higher the deacetylation degree in chitosan, the larger
was the water retention value. A polymer molecule with a high DDA shows a greater
number of hydrophilic substituents (-NH2 groups), being capable of binding water
molecules by hydrogen bonds.55 Therefore, owing to its high DDA, samples with LMW
showed a slightly high ability of water retention.
Particle size distribution measurements carried out in cement-polymer mixtures
confirmed that chitosan addition shifted the distribution of the largest particles towards
larger sizes (Fig. 5 depicts, as an example, the particle size distribution of cement pastes
with different MMW chitosan dosages). The bimodal distribution of the plain cement
mixture contrasted with the trimodal and sometimes even tetramodal distribution of the
cement-polymer mixtures. All the samples showed a maximum around 0.27 m, which
did not present further modifications. The plain cement mixture showed other maximum
at around 12 m. Samples with chitosan presented other relevant population at around
26 m (sometimes depicted as a shoulder). The ratio of particles with sizes bigger than
30 m was larger for chitosan-added samples. Finally, a population above 100 m can
be also observed (with an average value ca. 200 m). Polymer entanglement could be
responsible for these displacements towards larger sizes and for the appearance of the
largest populations, thus suggesting that a certain kind of interaction with cement
particles takes place. In general, a dosage-dependence could be proved: the larger the
dosage, the greater the displacement of the particle population towards larger sizes, as it
can be seen in Fig. 5 that collects data from MMW chitosan samples.
Set-retarding ability has been reported to be one of the roles played by some of the
polysaccharides used as cement admixtures.2 The setting process evolution has been
assessed by means of conductometric curves in cement suspensions, according to the

14

works by Peschard et al.2 and Pourchez et al.3 The conductometric curve of a cement
suspension presents different steps, the portlandite precipitation time, shown by an
electrical conductivity drop, being the most useful marker in order to assess the
hydration delay. In Fig. 6 the conductometric curves of cement suspensions can be seen
(conductivity of the cement suspensions as a function of the time): one as control
sample, without chitosan addition, and the others with the addition of the three different
chitosans tested. There is no change either in the dissolution rate of the cement particles
or in the nucleation rate of CSH (belonging to the first part of the curves). However, the
conductivity drop (portlandite precipitation) shows outstanding differences, involving
hydration delays as different chitosans were added to the cement suspensions. Results
indicate that chitosans exert a set-retarding action in cement suspensions. As far as the
interaction mechanisms are concerned, the reported insolubility of chitosans in alkaline
pH suggests some kind of adsorption between precipitants and anhydrous cement
particles, hydrates or crystal lattices rather than water unavailability, because the water
retention differences do not fully justify such a long delay in the hydration time.2 The
particle size distribution further reinforces this assumption on the interaction
mechanism (Fig. 5).
With the aim of assessing the interaction between chitosan molecules and cement
particles, zeta potential measurements were obtained for increasing quantities of
admixture (Fig. 7). The net-surface charge for cement suspensions without chitosan was
negative. This matches previous results, which stated that the zeta potential of early
hydration of ordinary Portland cement suspensions is dominated by the negative zeta
potential of the ettringite.56
LMW chitosan does not modify the zeta potential of the cement particles at most
dosages. This fact suggests that there is no adsorption of the polymer (owing to the
alkaline pH, a negatively charged polyelectrolyte) on the cement particles. Only the
highest dosage of polymer shifts the zeta potential towards more negative values,
showing a certain degree of interaction. These experimental results agree with the
poorer set-retarding capability of LMW chitosan (Fig. 6). They also point to the
adsorption on cement particles as the mechanism for the set-retarding effect.
The surface charge of cement particles with MMW chitosan showed more negative
values, proving that adsorption of this chitosan on cement particles took place. This
interaction could be useful to explain the delay in the setting process when MMW
chitosan was added. Further studies will be necessary to elucidate where the adsorption

15

takes place: stern or diffuse double layer, and to decide the kind of phases that interact:
silicates (C3S, CSH), CH (calcium hydroxide), or aluminate phases (C3A, ettringite,
monosulfate).56
Finally, HMW chitosan led to a charge inversion of the zeta potential (high positive
values), suggesting a very diverse interaction between cement particles and this
chitosan. The strong flocculation ability of this chitosan could give rise to the formation
of agglomerates of cement particles and polymer molecules (as proved by particle size
distribution measurements). As a hypothesis, these agglomerates should be negatively
charged, taking into account both the net-surface charge of the cement particles and the
ionized negatively charged groups of the chitosan chains, the agglomerates being
sheltered by a layer of positive counterions (calcium ions). As a result, the zeta potential
turned out to be positive and the hydration of cement particles inside such agglomerates
was hindered. This approach confirms the most effective thickening behaviour observed
for HMW chitosan, together with both its set-retarding ability and its lower water
retention.

CONCLUSIONS
In the present study, a natural biodegradable polymer, chitosan, was tested as a heavy
metal retainer and a viscosity-enhancing admixture with the aim of investigating its
usefulness as an addition to cement-based materials.
With regard to chitosans ability to retain heavy metals, voltammetric techniques have
been shown to be suitable for this study. This polymer proves to be an effective ligand
with all the six assayed metals. For all of them - Pb, Cd, Mo, Cu, Zn and Cr-, it can be
said that not only adsorption but also complexation actually takes place in acetate buffer
(pH=4) with chitosans of different MW.
The length of the chains of the polymer, together with the assayed concentration, have
been seen to play a key role in the chitosan binding activity with Zn and Pb as well as
Cr and Mo.
A predominant intra-chain linkage is proposed in all instances except for Cd and Cu,
which response seem more independent of the chitosan characteristics (molecular
weight and concentration) for which an inter-layer binding mechanism fits better.
Regarding the viscosity modifying role of the chitosan, it has been proved to increase
viscosity in fresh mortars while retaining more water than plain cement. Owing to the
greater probability of entanglement between chains of the polymer as well as a cross

16

linking phenomenon in a calcium-rich system and a low degree of deacetylation, the

HMW polymer was responsible for the greatest thickening effect. The degree of
deacetylation of the polymer could also be connected to the slight differences in waterretention ability.
The rise of these properties is enough to prove the capability of the chitosan as a
modifier of the rheological behaviour of cement-based materials. Furthermore, the
hydration delay assessed by means of conductometric curves suggests a set-retarding
role of the chitosan in cement mortars, the HMW chitosan being the polymer with the
greatest retarding ability. The mechanism for this set-retarding action has been related
to an interaction of the chitosan with cement particles according to the zeta potential
measurements.
Among the assayed chitosans, both voltammetric studies and rheological tests have
shown that the HMW chitosan has the best effectiveness as a complexing agent for Zn,
Pb and Cr, and also the best thickening behaviour in cement-based materials. These
combined methodologies allow us to conclude that high molecular chitosan derivatives
will be most appropriate for the final aim of this research.

ACKNOWLEDGEMENTS
This work was fully funded by the Ministry of Education and Science of Spain
(MAT2007-65478) and FUNA (Fundacin Universitaria de Navarra). M. Lasheras
would like to thank the Friends of the University of Navarra, Inc. for funding support.
REFERENCES
1. Phan, T.H.; Chaouche, M.; Moranville, M. Cem Concr Res 2006, 36, 1807.
2. Peschard, A.; Govin, A.; Pourchez, J.; Fredon, E.; Bertrand, L.; Maximilien, S.;
Guilhot, B. J Eur Ceram Soc 2006, 26, 1439.
3. Pourchez, J.; Govin, A.; Grosseau, P.; Guyonnet, R.; Guilhot, B.; Ruot, B. Cem
Concr Res 2006, 36, 1252.
4. Pourchez, J.; Peschard, A.; Grosseau, P.; Guyonnet, R.; Guilhot, B.; Vallee, F. Cem
Concr Res 2006, 36, 288.
5. Vieira, M.C.; Klemm, D.; Einfeldt, L.; Albrecht, G. Cem Concr Res 2005, 35, 883.
6. Peschard, A.; Govin, A.; Grosseau, P.; Guilhot, B.; Guyonnet, R. Cem Concr Res
2004, 34, 2153.
7. Peysson, S.; Pra, J .; Chabannet, M. Cem Concr Res 2005, 35, 2261.

17

8. Qiao, X.C.; Poon, C.S .; Cheeseman, C.R. J Hazard Mater 2006, 129, 290.
9. Arica, M.Y .; Bayramolu, G. Colloids Surf A 2005, 253, 203.
10. avu, S.; Grdag, G.; Yaar, M.; Gl, K.; Grkaynak, M.A. Polym Bull 2006,
57, 445.
11. Halim, C.E.; Short, S.A.; Scott, J.A.; Amal, R.; Low, G. J Hazard Mater 2005, 125,
45.
12. Ferri, T.; Sangiorgio, P. Anal Chim Acta 1999, 385, 337.
13. Marion, A.M.; De Lanve, M.; De Grauw, A. Cem Concr Res 2005, 35, 951.
14. Park, J.H.; Cho, Y.W.; Chung, H.; Kwon, I.Ch .; Jeong, S.Y. Biomacromolecules
2003, 4, 1087.
15. Ogawa, K.; Yui, T .; Okuyama, K. Int J Biol Macromol 2004, 34, 1.
16. Pan, Z.H.; Cai, H.P.; Jiang, P.P.; Fan, Q.Y. J Polym Res 2006, 13, 323.
17. Muzzarelli, C .; Muzzarelli, R.A.A. J Inorg Biochem 2002, 92, 89.
18. Alkhraisat, M.H.; Rueda, C.; Blanco, L.; Mario, F.T.; Torres, J.; Gbureck, U.;
Lopez, E. Acta Biomater 2010, 6, 257.
19. Trimukhe, K.D.; Varma, A.J. Carbohydr Polym 2008, 71, 66.
20. Trimukhe, K.D.; Varma, A.J. Carbohydr Polym 2008, 71, 698.
21. Trimukhe, K.D.; Varma, A.J. Carbohydr Polym 2009, 75, 63.
22. Mcafee, B.J.; Gould, W.D.; Nadeau, C.J.; da Costa, A.C.A. Sep Sci Technol 2001,
36, 3207.
23. Vold, I.M.N.; Vrum, K.M.; Guibal, E.; Smidsrd, O. Carbohydr Polym 2003, 54,
471.
24. Schmul, R.; Krieg, H.M.; Keizer, K.; Water SA 2001, 27, 1.
25. Muzzarelli, R.A.A. Carbohydr Polym 1985, 5, 461.
26. Roberts, G.A. F.; Domszy, J.G. Int J Biol Macromol 1982, 4, 374.
27. Izaguirre, A.; Lanas, J.; Alvarez, J.I. Cem Concr Res 2009, 39, 1095.
28. UNE 1015-3.; Methods of test for mortar for masonry. Part 3, Determination of
consistence of fresh mortar (by flow table): 2000.
29. UNE 83-816-93.; Test methods. Mortars. Fresh mortars. Determination of water
retentivity: 1993.
30. Chiou, H.; Fellows, C.M.; Gilbert, R.G.; Fitzgerald, M.A. Carbohydr Polym 2005,
61, 61.
31. Khayat, K.H. Cem Concr Compos 1998, 20, 171.
32. Saric-Coric, M.; Khayat, K.H.; Tagnit-Hamou, A. Cem Concr Res 2003, 33, 1999.

18

33. Papo, A.; Piani, L. Cem Concr Res 2004, 34, 2097.
34. Zhang, D.F.; Ju, B.Z.; Zhang, S.F.; He, L.; Yang, J.Z. Carbohydr Polym 2007, 70,
363.
35. Zhang, D.F.; Ju, B.Z.; Zhang, S.F.; Yang, J.Z. Carbohydr Polym 2008, 71, 80.
36.

Izaguirre,

A.;

Lanas,

J.;

Alvarez,

J.I.

Carbohydr

Polym

in

press

doi,10.1016/j.carbpol.2009.11.010
37. Rangel-Mendez, J.R.; Monroy-Zepeda, R.; Leyva-Ramos, E.; Daz-Flores, P.E .;
Shiari, K. J Hazard Mater 2009, 162, 503.
38. Salinas, I.; Esparza, I.; Gmez, S.; Santamara, C.; Fernndez, J.M. Electroanalysis
2005, 17, 469.
39. Webster, A.; Halling, M.D.; Grant, D.N. Carbohydr Res 2007, 342, 1189.
40. Buffle, J. J Electroanal Chem 1981, 125, 273.
41. Ruzic, I. Anal Chim Acta 1982, 140, 99.
42. Scatchard, G. Ann NY Acad Sci 1949, 51, 660.
43. Langmuir, D. Techniques of estimating thermodynamic properties for some aqueous
complexes of geochemical interest, In Chemical Modeling in Aqueous Systems, Ed
Jenne EA. American Chemical Society: Washington, 1979.
44. Batley, G.E. Anal Chim Acta 1986, 189, 371.
45. Rhazi, M.; Desbrires, J.; Tolaimate, A.; Rinaudo, M.; Vottero, P.; Alalgui, A.
Polymer 2002, 43, 1267.
46. Guibal, E.; Milot, C.; Eterradossi, O.; Gauffier, C.; Domard, A. Int J Biol Macromol
1999, 24, 49.
47. Ravi Kumar, M.N.V. React Funct Polym 2000, 46, 1.
48. Guibal, E. Sep Purif Technol 2004, 38, 43.
49. Sonebi, M. Cem Concr Res 2006, 36, 1609.
50. Yahia, A.; Khayat, K.H. Cem Concr Res 2001, 31, 731.
51. Wallevik, J.E. Cem Concr Res 2009, 39, 14.
52. Gomes, P.; Gomes, C.A.R.; Batista, M.K.S.; Pinto, L.F.; Silva, P.A.P. Carbohydr
Polymer 2008, 71, 54.
53. Goycoolea, F.M.; Morris, E.R.; Gidley, M.J. Carbohydr Polymer 1995, 27, 69.
54. Knaus, S.; Bauer-Heim, B. Carbohydr Polymer 2003, 53, 383.
55. Chen, L.; Du, Y.; Zeng, X. Carbohydr Res 2003, 338, 333.

19

56. Zingg, A.; Winnefeld, F.; Holzer, L.; Pakusch, J.; Becker, S.; Gauckler, L. J
Colloids Interf Sci 2008, 323, 301.
Table I. Electrochemical conditions used for the different metals.

Parameters

Pb

Cd

Mo

DPASV DPASV
Type
9
9
Drop size
300
300
Pruge time (s)
120
120
Conditioning time (s)
2000
2000
Rotation speed (rpm)
-600
-900
Accumulation potential (E, mV)
60
60
Accumulation time (s)
20
20
Equilibration time (s)
-600
-900
Initial potential (mV)
0
-200
Final potential (mV)
100
100
Pulse amplitude (mV)
20
12
Scan speed (mV s-1)

Cu

DPV
9
300
120
2000
150
-1200
25
40

Zn

DPASV DPASV
9
9
300
300
120
120
2000
2000
-300
-1200
60
60
20
20
-300
-1200
150
-700
100
100
12
12

Cr
DPV
9
300
120
2000
-200
-1200
25
40

Table II. Influence of chitosan molecular weight and chitosan dilution on the number of
metal nuclei complexed by each chitosan molecule.

Element

LMW
0.5 M

0.1 M

MMW
0.5 M

1.0 M

HMW
0.5 M

Zn

10

40

20

10

40

Pb

24

12

16

Cd

10

Table III. Slump and water retention of cement pastes with chitosan additions.

Cement pastes

Slump (mm)

Water retention (%)

Control group

190.0

81.1

CEM + LMW

171.5

85.2

CEM + MMW

157.0

84.6

CEM + HMW

151.5

83.5

20


Figure 1. Titration curves for increasing concentrations of Pb (a), Cd (b), Zn (c) in 0.5 M
chitosan,andCu(d)in0.05Mchitosansolution.
Figure2.EffectofincreasingconcentrationsofchitosanonthepeakintensityofMo104M(a)
andonpeakpotentialdisplacementforCr3.4104M(b).
Figure3.Pendant(a)andbridge(b)modelsaccordingtoGuibal48andRhazietal.45
Figure4.Apparentviscosityofcementpasteswithchitosanadditionsvs.increasingshearrate.
Figure 5.Particlesizedistributionofcementpasteswithdifferent medium molecularweight
chitosandosages(w/wpercentageschitosan/cement).
Figure6.Effectofthedifferentchitosansoncementhydration.
Figure 7. Zeta potential of the cement particles for different molecular weight chitosans vs.
polymerconcentration.

21

Fig.1

22

Fig.2

23

Fig.3

24

Fig.4

Fig.5

25


Fig.6

Fig.7

26