Environ. Sci. Technol. 2003, 37, 261-267
Metal Selectivity of Sargassum
spp. and Their Alginates in Relation
to Their r-L-Guluronic Acid Content
and Conformation
T H O M A S A . D A V I S , †,‡
FRANCISCO LLANES,§
BOHUMIL VOLESKY,† AND
A L F O N S O M U C C I * ,‡
Department of Chemical Engineering, McGill University,
3610 University Street, Montreal, Quebec, Canada H3A 2B2,
Department of Earth and Planetary Sciences, McGill
University, 3450 University Street, Montreal, Quebec,
Canada H3A 2A7, and Biomaterials Center, University of
Havana, Havana, 10400 Cuba
The discovery of a consistent and unusual enrichment in
homopolymeric R-L-guluronic acid G-blocks in alginates
extracted from a suite of Sargassum brown algae is described
in this study. 1H NMR spectroscopy was used to
characterize these alginates which display homopolymeric
guluronic acid block (G-block) frequency values (FGG)
between 0.37 and 0.81. The presence of these G-blocks
results in an enhanced selectivity for cadmium or calcium
relative to monovalent ions such as sodium and the
proton as well as smaller divalent ions such as magnesium.
Results of competitive exchange experiments for the CdCa-alginate system yield selectivity coefficient, K*CdCa,
values between 0.43 ( 0.10 and 1.32 ( 0.02 for a range in
FGG of 0.23 to 0.81. In contrast to the Cd-Ca-alginate
system, the Mg-Ca-alginate and Mg-Cd-alginate systems
yielded maximum values of K*MgCa (18.0 ( 1.4) and
K*MgCd (16.0 ( 0.9) for the alginates extracted from
Sargassum fluitans (FGG ) 0.81; Cuba) and Sargassum
thunbergii (FGG ) 0.75; Korea), respectively. Selectivity
studies with mixed-metal pair alginate systems highlight
the importance of the specific macromolecular conformation
of the alginate polymer in determining metal binding
behavior in multiple-metal systems. Furthermore, they
demonstrate the importance of the conformation of the
alginate as it occurs within the tissue of Sargassum in
determining the metal binding behavior of this algal biosorbent.
The unique composition of the alginates present in
species of Sargassum may represent a distinct advantage
over other brown algal species when considering their
implementation for the strategic removal of toxic heavy
metals from contaminated and industrial wastewaters.
Introduction
Significant efforts have been invested to develop the use of
raw biomass for removal of toxic or strategic elements from
* Corresponding author phone: (514)398-4892; fax: (514)398-4680;
e-mail: alm@eps.mcgill.ca.
† Department of Chemical Engineering, McGill University.
‡ Department of Earth and Planetary Sciences, McGill University.
§ University of Havana.
10.1021/es025781d CCC: $25.00
Published on Web 11/27/2002
 2003 American Chemical Society
contaminated and industrial wastewaters (1). The ability of
biomass such as fungi, algae, bacteria, and actinomycetes to
sequester heavy metals has already been reported (2, 3). The
term biosorption has been applied to this process, and it
refers specifically to the passive, or rather, nonmetabolically
mediated binding of ions to the functional groups of the
biomass.
The brown algae Sargassum possesses the required
mechanical properties (4), chemical affinity, and sorption
capacity (4, 5) necessary to bind metals such as Cu, Pb, Hg,
and Cd in an effective, reversible, and potentially costeffective manner. Previous studies of this brown algae have
also focused on the implementation of a fully operational
remediation system using a fix-bed column design (6-8),
modeling of electrostatic effects (9-11), and identification
of the primary binding sites (12). Alginate (Figure 1) was
identified as the most important component (12) leading to
metal binding by the brown algal tissue of Sargassum.
However, very little information exists about the detailed
composition of this highly variable poly-uronide as it naturally
occurs within Sargassum and more importantly about the
architecture of this biopolymer as it occurs within the cell
wall. The alginate composition varies widely depending on
the season and location of natural harvesting (13-15), and
with limited knowledge about the consistency of the structure
of the biosorbent, it becomes increasingly challenging to
model and predict its metal binding performance in remediation scenarios.
In preparation for the implementation of a remediation
program for solutions of mixed metals, the elucidation of the
binding mechanism, taking into consideration the variable
nature of the biopolymers contained within the biosorbent
as well as their macromolecular structure, remains a critical
step. This study represents an extension of previous work
(12) that investigated the role of alginate in the binding of
single heavy metals by brown seaweed. That work established
that cadmium binding by bulk Sargassum fluitans and its
isolated alginate arises from bridging or bidentate complex
formation with the carboxylate groups of the alginate. Another
investigation (4) attempted to relate heavy metal uptake to
the macromolecular conformation of the alginate, but that
study was limited to single metal systems (i.e., Cd, Zn, Cu,
Ca, and Pb) at a fixed pH and no correlation could be
established. A correlation, however, may emerge from the
study of binary systems (i.e., in the presence of two metals)
where competitive adsorption will reveal metal selectivity.
Alginic acid or alginate, the salt of alginic acid, is the
common term applied to a family of linear polysaccharides
containing 1,4-linked β-D-mannuronic (M) and R-Lguluronic (G) acid residues arranged in a block-wise, nonregular order along the chain (Figure 1). The proportion of
M and G residues and their macromolecular conformation
determine the physical properties and the affinity of the
alginate for divalent metals (16). Polyguluronic acid contains
two diaxially (1a,4a) linked R-L-guluronic acid residues in
the chair form which produce a rodlike conformation with
a molecular repeat of 8.7 Å ((17) Figure 1b). In contrast,
polymannuronic acid forms a flat ribbonlike chain, its
molecular repeat is 10.35 Å, and it contains two diequatorially
(1e,4e) linked β-D-mannuronic acid residues in the chair form
(18). It is this difference in conformation between the two
homopolymeric blocks which is believed to be chiefly
responsible for their strong but variable affinity for divalent
heavy metals.
Selectivity coefficients derived from ion-exchange reactions between sodium and divalent metals for two alginates
VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
261
goals we obtained five different species of Sargassum seaweed
from various localities and whose detailed compositions have
not previously been described. The alginates were isolated
using a newly modified technique which yields low-viscosity
Na-alginate solutions suitable for characterization by 1Hnuclear magnetic resonance spectroscopy (NMR).
Materials and Methods
FIGURE 1. Alginate molecular structure: (a) alginate monomers
(uronic acids: M vs G); (b) macromolecular conformation of the
alginate polymer; (c) chain sequences; block copolymer structure
(after ref 44); (d) calcium induced gelation of alginate: schematic
representation in accordance with the “egg-box” structure (after
ref 45).
(19) confirmed the higher affinity of G-blocks for divalent
heavy metals. The higher specificity for divalent metals is
explained by the “zig-zag” structure of poly-G which can
accommodate the Ca2+ (and other divalent metals) ion more
easily (20-23). The alginates are believed to adopt an ordered
solution conformation, through dimerization of the poly-G
sequences in the presence of calcium or other divalent cations
of similar size. It is the rigid and buckled shape of the polyG-sections which results in an alignment of two chain sections
yielding an array of coordination sites with cavities favorable
to divalent cations because they are lined with carboxylate
and other electronegative oxygen atoms. This description is
known as the “egg-box” model ((21-23) Figure 1d). The
regions of dimerization are terminated by chain sequences
of poly-M. As a result, several different chains may become
interconnected and this promotes gel network formation.
We report here on a detailed study of the compositional
variability of alginates extracted from several species of
Sargassum seaweed. Its composition, which is unusually rich
in homopolymeric guluronic acid blocks (FGG), results in an
enhanced selectivity for cadmium or calcium relative to
smaller divalent ions such as magnesium. In contrast to
previous single metal (e.g., Cd) studies, mixed-metal pair
systems highlight the importance of the specific macromolecular conformation of the alginate biopolymer in
determining metal binding behavior in multiple-metal
systems. The goals of this study were to (i) extend the limited
knowledge base of Sargassum alginate compositions, (ii)
demonstrate that the alginates of Sargassum, relatively unique
among their class of algal polysaccharides, display selectivity
among metal cations, and (iii) show that calcium competes
with cadmium in binding to alginate-based materials irrespective of their uronic acid composition. To achieve these
262
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003
Samples. Sargassum fluitans originates from the Sargasso
Sea of the northwest Atlantic Ocean and is one of the few
known pelagic species. It is carried by winds and tides to the
shores of Cuba and other Caribbean Islands where it
accumulates in copious quantities along the beaches. The
biomass used in this work was collected fresh at Guanabo
Beach, 30 kilometers east of Havana. Sargassum siliquosum
and Sargassum oligocystum were collected from the reef flat,
1 to 4 m below tide level, on the fringing reef at Goold Island
(18°10.9′ S, 146°10.2′ E) on the inshore, central Great Barrier
Reef, Australia. These benthic species are found as part of
mixed species assemblages that dominate the reef flat of
many inshore reefs on the Great Barrier Reef (24-26).
Sargassum thunbergii was collected at Songjong Beach, Pusan
Bay, Korea, Sargassum muticum originated from the south
of England, and Macrocystis pyrifera was sampled in Nova
Scotia, Canada. The latter three species are also benthic
species. A sample of purified alginate extract from Macrocystis
pyrifera was obtained from Sigma (A-7128, Lot 58H0472). A
highly purified sample of chitosan was provided courtesy of
the Norwegian Biopolymer Laboratory, for use as a control
for the carbon:nitrogen analyses.
Alginate Extraction. Alginate was extracted from the bulk
samples in a 2% solution of Na2CO3 according to a slight
modification of the method of Percival and McDowell (27).
The extraction was carried out at 80 °C instead of room
temperature in order to reduce the viscosity (for 1H NMR)
and ensure complete extraction of the alginate (28). This
procedure eliminates the need for a separate prehydrolysis
step (28), which is normally applied to high viscosity alginate
extracts prior to high resolution 1H NMR characterization.
Integration and preliminary interpretation of the relative peak
heights revealed that the alginates extracted by the hightemperature method applied in this study displayed an
average polymer length of 20 to 25 carbohydrate residues.
This falls within the range (i.e., 20-30) where corrections for
the presence of end-groups are not required for the interpretation of 1H NMR polymer spectra (15). In the presence
of excess Na2CO3, the alginic acid is converted to a Naalginate and is solubilized. The resulting Na-alginate solution
was separated from the solid phase by filtration (Whatman
filter paper #4). This step was followed by the precipitation
of the alginic acid upon addition of dilute hydrochloric acid
and conversion of the sodium salt to the insoluble acid (pH
< 1.0). The alginic acid precipitate was pelletized by
centrifugation and washed with a 95% aqueous ethanol
solution prior to conversion back to the sodium salt upon
addition of a concentrated sodium carbonate solution.
Alginate Content of Sargassum. The alginate content of
Sargassum fluitans has previously been reported (12) to
account for 45% of the dry weight of the biomass once it has
been stripped of its sea salts and converted to the protonated
form by washing with a 0.1 N HCl solution and extensive
rinsing with distilled water. Figueira et al. (29) have shown
that, for a variety of brown algae, between approximately 40
and 50% of the original biomass’ dry weight is lost during
this acid treatment. The alginate yields of S. fluitans and S.
oligocystum extracted for this work were on the order of 45
and 37%, respectively.
1H NMR Spectroscopy, Carbon and Nitrogen Analyses.
The freeze-dried Na-alginates were dissolved in deuterium
oxide (D2O, 99.9%) and dried several times prior to NMR
spectrum acquisition. The 1H spectra were recorded with a
Varian Unity 500 spectrometer at a temperature of 70 °C, a
sweep width of 5999.7 Hz, an 80° pulse, and an acquisition
time of 2.048 s. Typically, 128 or 256 repetitive scans were
acquired, and the data were processed with a line broadening
of 0.6 Hz. Sodium 3-trimethylsilylpropionate-2,2,3,3-d4 (TSP)
(Aldrich) was used as an internal reference. The solvent peak
(HDO) was partly eliminated using a decoupler with a 5.0 s
delay period. We also recorded several spectra at 90 °C without
the decoupler since the HDO peak was shifted further upfield
away from the peaks of interest. Results of the integration
of the peaks of interest were not affected by the change in
temperature nor the use of the decoupler. Consequently,
results in this work stem from spectral acquisitions at both
temperatures.
The purity of the extracted alginates and, more specifically,
the ratio of carbon to nitrogen in the freeze-dried samples
was determined with a Carlo Erba carbon, nitrogen, and sulfur
elemental analyzer (model NA 1500). Highly purified chitosan
(C/N ) 6) was used as a standard (error < 3.0%), and duplicate
analyses were performed for all samples.
Alginate Maximal Cd Uptake. To determine the relative
uptake of cadmium by alginates of varying guluronic acid
composition, at a uniform (final) equilibrium cadmium
concentration, the so-called ‘tea-bag’ method of Fourest and
Volesky (12) was employed. A large beaker was filled with 3.0
L of a 4.51 mM CdCl2 solution (pH ) 4.5) and kept constantly
stirred. In it were placed a total of 15 dialysis bags (Spectra/
Pore Membrane MWCO: 6-8000), three bags for each of the
alginates extracted from S. siliquosum, S. thunbergii, and M.
pyrifera, three samples of native-unextracted S. fluitans, and
the three empty control dialysis bags. The alginate and
seaweed samples were left to equilibrate with the solution
for one week, subsequently removed and transferred into
individual plastic containers for rinsing with deionized
distilled water. A series of final acid rinses (0.1 HCl; see
detailed description in next section) were used to leach each
sample, determine the amount of bound metal, and estimate
the total metal uptake for each sample.
Determination of Exchange Equilibria. Solutions of
sodium alginate (1%, 2 mL) extracted from the six brown
algae chosen for this study were dialyzed (Spectra/Pore
Membrane MWCO: 6-8000) against a solution containing
the salts of the two cations to be investigated at a total
concentration of 0.20 ( 0.01 M, according to a modification
of the method of Smidsrød and Haug (30). Different
combinations of MgCl2, CaCl2, and CdCl2 were used (pH )
5.0). In addition, the ratio of Mg to Ca or Cd was varied in
order to shift the equilibrium and hence the mole fraction
of Ca or Cd bound to the alginates. The dialysis bags were
placed in a small plastic container (120 mL) and filled with
50 mL of the mixed-metal solution. The solutions were
changed a minimum of 4 times, every 12 h, while being
continuously stirred on an orbital shaker. After the alginates
were equilibrated with the fixed metal concentration solutions, they were rinsed in deionized, distilled water 4 times
in periods of 12 h. A final acid rinse (0.1 HCl) was used to
leach the bound metals from the alginates, with several rinses
(12 h, 20 mL, 4-5 times) being pooled in order to quantify
the total metal content. The original metal solutions and the
acid rinse were analyzed by inductively coupled plasma
atomic emission spectrometry (0.1 ppm detection limit,
reproducibility to > 95%).
Results and Discussion
Composition of Sargassum Derived Alginates. The approach
of Grasdalen et al. (15, 31) was used in order to determine
the M:G ratio and block structure of the alginates described
in this study. The abundance or frequency of individual
TABLE 1. Compositional Data of Na-Alginates Extracted from
Species of Sargassum
composition,
fractions
doublet frequencies
source
FM
FG
FMM
FMG
FGM
FGG
M/G
S. filipendula
S. muticum
S. oligocystum (no. 1)
S. oligocystum (no. 2)
S. polycystum
S. thunbergii (no. 1)
S. thunbergii (no. 2)
0.16
0.24
0.43
0.44
0.18
0.34
0.20
0.84
0.76
0.57
0.56
0.82
0.66
0.80
0.07
0.07
0.24
0.35
0.12
0.17
0.16
0.08
0.17
0.20
0.09
0.05
0.17
0.05
0.08
0.17
0.20
0.09
0.05
0.17
0.05
0.76
0.59
0.37
0.47
0.77
0.48
0.75
0.19
0.31
0.77
0.77
0.21
0.53
0.25
monomer guluronic (FG) and mannuronic (FM) acid residues
as well as the frequencies of the doublet uronic acid pairs
(i.e., FGG, FMM, and FMG) were determined by integration of
the appropriate 1H NMR peaks of the Na-alginates according
to the protocol proposed by Grasdalen et al. (15). A detailed
description of the protocol can be found in Davis et al. (28).
The quantity FGG is typically used to characterize the degree
of guluronic acid block structure for a given alginate sample.
The 1H NMR results are summarized in Table 1. Seven
individual Na-alginate samples represent five different
species of Sargassum brown algae. Two different extracts of
Na-alginate were prepared from different samplings of both
S. oligocystum and S. thunbergii. The M:G ratio is defined
explicitly as
M 1 - FG
)
G
FG
(1)
following substitution of FG (determined directly from the
appropriate integrations of the spectra). The M:G ratio of the
Na-alginates varies from 0.19 to 0.77, with the alginate
derived from S. filipendula having the highest guluronic acid
content (FG ) 0.84). Guluronic acid is the major component
of all samples, with the lowest FG being 0.57. Homopolymeric
guluronic acid blocks (FGG) also dominate in all samples over
homopolymeric mannuronic acid blocks (FMM) and alternating block sequences (FMG). The composition of the two
alginate samples extracted from S. oligocystum compare
favorably. They yield the same M:G ratio and only minor
differences in homopolymeric GG diad frequencies (0.37 and
0.47). On the other hand, the two alginate samples extracted
from S. thunbergii differ in both M:G ratio (0.25 and 0.53)
and FGG (0.48 and 0.75). These results reflect the potential
compositional variability of alginate extracted from different
batches of Sargassum from the same species (31-33). Despite
this difference, the alginates are still considered to be highly
enriched in guluronic acid when compared to the vast
majority of alginate compositions (15, 31-33). To the best
of the authors’ knowledge, the compositional analysis of this
suite of samples represents the most extensive characterization of the block structure of Sargassum derived alginates.
Llanes et al. (33) used solid state 13C NMR spectroscopy
to characterize dried alginate samples from a mixed Sargassum sample. The authors concluded that the alginate
sample (M:G ≈ 0.60) contained guluronic acid which mainly
occurred as homopolymeric blocks. The corroboration of
our data derived from aqueous Na-alginate solutions with
information obtained on the block structure in the solidstate further strengthens our conclusion that Sargassum
derived alginates display an unusual composition of uronic
acids along the block copolymer. Similar compositions,
among other brown seaweeds, are only known to occur in
the stipes of Laminaria hyperborea which typically yield a
FGG of ≈ 0.42.
VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
263
TABLE 2. Alginate Uronic Acid Content and Maximal Cd Uptake
at pH ) 4.5
alginate (source)
M/G
FGG
uptake
(mmol/g)
Macrocystis pyrifera
1.70 0.23 2.01 ( 0.19
Sargassum siliquosum
0.72 0.57 1.82 ( 0.05
Sargassum thunbergii
0.25 0.75 1.73 ( 0.10
native S. fluitans biomassa (control)
0.95 ( 0.04
dialysis membrane (control)
0.06 ( 0.01
a The alginate composition of this sample is similar to that listed in
Table 3.
Alginate-Maximal Cadmium Uptake. Cadmiumalginate uptake studies were performed on alginates of
varying guluronic acid content (M:G ) 0.25 to 1.70) in order
to test the hypothesis that the macromolecular conformation
may be responsible for differential metal uptake due to the
preferential binding by the guluronic block sections. In
contrast to the whole algal tissue, which may be placed
directly into the metal bearing solution, Na-alginate is
soluble and therefore must be contained within a dialysis
membrane. For lack of abundance of the alginate extracts,
only one initial metal concentration was used. This concentration was sufficiently high (4.51 mM) to ensure that
maximal uptake for the mass of biopolymer was achieved in
the experiment. Previous studies (5) have demonstrated that
sorption can be described by a single metal Langmuir sorption
isotherm whereby maximal uptake occurs upon saturation
of the binding sites. The minimal, final metal concentration
required to reach the isotherm plateau or maximal metal
uptake is approximately 0.53 mM (or 60 ppm) for cadmium
bound to Sargassum spp. (5). Furthermore, to make a fair
comparison of the various samples, one large batch experiment was performed according to the method of Fourest
and Volesky (12) so that all samples were in equilibrium with
the same final cadmium concentration. The initial and final
concentrations of the bulk cadmium chloride solution were
determined to be the same (4.51 mM), indicating that the
loss of metal to the alginate phase was insignificant compared
to the bulk metal solution concentration. In addition, and
for comparison purposes, a native sample of Sargassum
fluitans was also placed into dialysis membranes and
included in the metal uptake experiments. An empty dialysis
membrane served as a control. The experiments were carried
out and maintained at pH ) 4.5 by incremental additions of
LiOH throughout the equilibration.
The results are summarized in Table 2. The sample of
native S. fluitans biomass yielded a maximal cadmium uptake
(0.95 ( 0.04 mmol Cd/g) in keeping with the level expected
from previous experiments. Davis et al. (5) reported a range
in cadmium uptake for various Sargassum species from 0.66
to 0.90 mmol/g at pH ) 4.5. Some cadmium binding could
also be ascribed to the dialysis membrane since it took up
an estimated 0.06 mmol of Cd/g. Results of the cadmium
uptake experiments do not reveal any significant difference
between alginate samples of varying guluronic acid content
(FGG). At first glance, it appears as though more cadmium is
bound by the alginate extracts with the lower guluronic acid
content, i.e., Macrocystis pyrifera (2.01 mmol/g). However,
considering the maximum cumulative protocol and analytical
errors (estimated at 9.5%) and the uncertainty associated
with cadmium binding by the dialysis membrane, no
statistically significant relationship (to a 95.4% level of
confidence) could be established. Thus, it appears that the
selectivity coefficients of the divalent cation relative to the
monovalent proton, for the various alginates tested, are so
large that practically all sites are satisfied by the divalent
cation (in this case, cadmium). Hence, differential selectivity
264
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003
FIGURE 2. Selectivity coefficients (K*CdCa) for the cadmium-calcium
ion-exchange with alginates extracted from M. pyrifera (O), S.
oligocystum (4), and S. fluitans (0). XCa is the mole fraction of
calcium bound by the alginate in the gel phase. Error bars show
errors with a confidence level of 95.4%.
of the alginates in single metal systems may not be
ascertained.
Alginate Metal Exchange Equilibria. The selectivity
coefficient for the ion-exchange of two cations, A and B, is
defined by the following reaction and equilibrium relationship
≡S-A(g) + B(aq) a ≡S-B(g) + A(aq)
(2)
K*AB ) XB • CA/XA • CB
(3)
where (g) ) gel or alginate phase; (aq) ) aqueous phase; ≡S
) binding site; XA and XB are the equivalent fractions of the
counterions in the polymer phase, whereby XA + XB ) 1, and
CA and CB are the concentrations of the same ions in solution.
Three metal-pair-alginate systems (Cd-Ca; Mg-Ca; MgCd) were studied across a range of equivalent fractions of
the counterions in the polymer phase. The selectivity
coefficient, K*, for each metal-pair-alginate system is
operationally defined for the bulk electrolyte solutions, at a
total molar metal concentration of 0.20 M, and is therefore
a stoichiometric constant. Nevertheless, the selectivity coefficients derived in this study still reflect the relative affinity
of the alginates at environmentally relevant metal concentration levels. For example, the results of a study (34) of the
competitive uptake of Cu and Co by a Na-alginate gel at
intermediate (100 ppm or ∼1.6 mM) as well as low Cu (18
ppm or 0.28 mM) and high Co (300 ppm or 5.1 mM)
concentrations revealed a high selectivity for Cu over Co
(20.9 based on batch absorption and an extended Langmuir
model) in both cases such that more than 90% of the copper
was sequestered from the solution. The alginate (obtained
from CP Kelco) used in that study contained only 31%
guluronic acid, less than half the value than for four of the
alginates extracted from Sargassum in this work. The authors
acknowledged that an even greater selectivity for Cu over Co
could be expected for alginates containing higher amounts
of guluronate.
Figure 2 shows the results of the Cd-Ca exchange
experiments. The detailed uronic acid composition of each
of the alginates used in the experiments can be found in
Table 3, and Table 4 summarizes the maximal and
minimal values of the selectivity coefficients for 0.1 < XCa
< 0.72. The majority of the Cd-Ca experiments yield
reproducible results, but the large relative errors in several
of the exchange experiments (XCa < 0.3) indicate that this
experimental technique may not be the most suitable for
TABLE 3. Compositional Data for Alginates Used in Metal
Selectivity Experiments
composition,
fractions
source
FM
FG
doublet frequencies
FMM
FMG
FGM
FGG
M/G
Cadmium-Calcium (Figure 2)
Macrocystis pyrifera 0.63 0.37 0.49 0.14 0.14 0.23 1.70
S. oligocystum
0.43 0.57 0.24 0.20 0.20 0.37 0.77
S. fluitans
0.16 0.84 0.13 0.03 0.03 0.81 0.19
Magnesium-Calcium (Figure 3)
Macrocystis pyrifera 0.63 0.37 0.49 0.14 0.14 0.23 1.70
S. fluitans
0.16 0.84 0.13 0.03 0.03 0.81 0.19
Magnesium-Cadmium (Figure 4)
Macrocystis pyrifera 0.59 0.41 0.35 0.24 0.24 0.17 1.44
(Sigma)
S. siliquosum
0.42 0.58 0.41 0.01 0.01 0.57 0.72
S. thunbergii
0.20 0.80 0.16 0.05 0.05 0.75 0.25
systems that yield very low selectivity coefficients. The greater
relative errors at the lower XCa values likely result from a
more loosely formed Ca-Cd-alginate gel which is apparently
related to the smaller amount of network forming calcium
ions bound to the G-block sequences of the alginate.
Nevertheless, the selectivity coefficients for the Cd-Ca system
(K*CdCa) are similar for all the alginate extracts tested and
independent of the M:G ratio or FGG, over a broad range of
XCa. These data are, in fact, tightly clustered about unity when
compared to selectivity coefficients for other mixed metal
systems. For example, Smidsrød and Haug (30) reported
selectivity coefficients for various metal-pair-alginate systems that vary from 5 to 50 depending on the guluronic acid
content of the alginate. The higher coefficient values corresponded to extracts that were deliberately enriched in
guluronic acid by fractionation.
Figure 3 shows the results of the Mg-Ca exchange
experiments. For comparison, the results of the Cd-Ca
exchange experiments are included in the same figure and
are depicted by open symbols. In contrast to the cadmiumcalcium system, the selectivity coefficients for the magnesium-calcium system are highly variable and range from
2.10 to 18.0 (Table 4). The alginate extracted from Sargassum
fluitans (FGG ) 0.81) clearly displays an enhanced selectivity
for calcium over magnesium when compared to the alginate
extracted from Macrocystis pyrifera (FGG ) 0.23). The Mg-Ca
selectivity coefficients, K*MgCa, for the two alginate extracts
are maximal at relatively low mole fraction of bound calcium
(i.e., XCa < 0.5) and decrease with increasing XCa.
Figure 4 summarizes the results of the Mg-Cd exchange
experiments with three different alginate extracts (Tables 3
and 4). Like the Mg-Ca system, the selectivity coefficients,
K*MgCd, increase with increasing FGG but maximum values
occur at high mole fraction of bound cadmium (i.e., XCd >
0.5). The results of the Cd-Ca exchange experiments are
also included in Figure 4 except that, for comparison
purposes, they are plotted as a function of XCd.
The presence of nitrogen in the alginate extracts would
indicate that proteins may have been retained during the
extraction process. Proteins may display a high selectivity
for divalent ions and thus, the purity of the alginate extracts
may be critical in the interpretation of the metal exchange
results. Table 5 summarizes the results of duplicate elemental
analyses of the alginate extracts and chitosan standard. The
carbon and nitrogen contents are reported both in absolute
concentrations (i.e., µmoles/mg of sample and wt %) and
molar ratios. The commercial alginate (Sigma) as well as the
alginates extracted from S. fluitans, S. siliquosum, and M.
pyrifera contain no detectable nitrogen (detection limit <
0.5 µg of N/mg of sample). The alginate extracted from S.
thunbergii contains 1.1% nitrogen or the equivalent (6.25 ×
%N (35)) of 6.9% protein by weight. Nevertheless, this did
not appear to affect its metal selectivity. This conclusion is
based on the similarity of selectivity coefficient values
obtained for the other alginate-metal systems. The near unity
value of K*CaCd, for the range of XCa investigated, leads us to
expect that the selectivity coefficients for both the Mg-Ca
and Mg-Cd systems should yield comparable values for
alginates with equivalent frequencies of G-blocks (FGG).
Accordingly, the maximal selectivity for S. thunbergii in the
Mg-Cd system (FGG ) 0.75, K*MgCd ) 16.0) was only slightly
lower than that for S. fluitans in the Mg-Ca system (FGG )
0.81, K*MgCa ) 18.0) as would be expected on the basis of the
slightly lower frequency of G-blocks.
The presence of a maximum in selectivity curves, as
observed in this study for the Mg-Ca/Cd metal pair systems,
was previously described (36) and explained in terms of a
theoretical model for the gelation of alginates and the nearestneighbor cooperative effects in the binding of calcium to
G-blocks. In other words, binding of a calcium ion to one
site in the chain favors the binding of another calcium ion
in the neighboring position and the formation of sequences
of bound calcium ions along the chain. According to the
model proposed by Andresen et al. (36), only a fraction, R
(i.e., 0 < R < 1), of the binding sites in G-blocks participate
in this cooperativity. Low R values (i.e., R < 0.3) result in
maximal selectivity at values of XCa lower than 0.5. Conversely,
for high values of R (i.e., 1 > R > 0.5), maximal selectivity
occurs at values of XCa greater than 0.5. Previous studies (37)
have shown that alginate fragments rich in mannuronic acid
residues (M-blocks) and those with alternating structures
(MG-blocks) are characterized by low selectivities and
minimal cooperativity and, hence, display selectivity curves
with little or no maxima. The appearance of maxima in the
selectivity curves presented in Figures 3 and 4 is therefore
expected given the high proportion of homopolymeric
G-blocks of these Sargassum derived alginates. The presence
of a maximal selectivity coefficient at XCa values lower than
0.5 for the Mg-Ca system is compatible with the observations
of Andresen et al. (36) and, in terms of the model, indicates
that a small fraction, R, of G-block sites participate in
cooperative binding of calcium. The opposite was true for
the Mg-Cd system and suggests that cadmium binds
differently than calcium to the alginate, since a maximum
occurs at high XCd and implies that a larger fraction of G-block
sites participate in cooperative Cd binding.
Application to the Raw Biomass. The results of selectivity
studies performed on alginates extracted from Sargassum
bear directly on the selectivity of the raw biomass, the material
most likely to be used in biosorption remediation systems.
A previous study (38) has demonstrated that selectivity
coefficients for the strontium-calcium and strontiummagnesium metal pair systems of extracted alginates are
closely correlated to coefficients obtained with the raw brown
algal tissue. In both cases, enhanced selectivity for strontium
over magnesium or calcium was observed as the guluronic
acid content of the alginate in the algal tissue increased.
Similar results were reported for the calcium-magnesium
system (39). In that study, the ion-exchange reaction was
used as a “probe” to test whether the network structure was
different in the gel and in the plant. Laminaria hyperborea
stipes were used to prepare alginates rich in G-blocks, and
the samples were subsequently dialyzed extensively against
different calcium-magnesium chloride solutions in order
to obtain the selectivity coefficients. The same protocol was
applied to raw L. hyperborea tissue, and the results revealed
an even higher selectivity for the stipes than the extracted
alginates at low XCa values. These results weigh strongly in
favor of the hypothesis that the selectivity imparted by the
VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
265
TABLE 4. Maximal and Minimal Values of Selectivity Coefficients Obtained for Alginates with Varying G-Block Compositions
source
FGG
Max. K*CdCa
Min. K*CdCa
Max. K*MgCa
Min. K*MgCa
Macrocystis pyrifera
Sargassum oligocystum
Sargassum fluitans
Macrocystis pyrifera (Sigma)
Sargassum siliquosum
Sargassum thunbergii
0.23
0.37
0.81
0.17
0.57
0.75
1.40 ( 0.01
1.01 ( 0.01
1.32 ( 0.02
0.43 ( 0.10
0.36 ( 0.03
0.80 ( < 0.01
12.0 ( < 0.1
4.83 ( 0.76
18.0 ( 1.4
6.67 ( 1.41
Max. K*MgCd
Min. K*MgCd
10.1 ( 1.1
11.7 ( 0.3
16.0 ( 0.9
2.10 ( 0.05
2.35 ( 0.20
2.64 ( 0.08
TABLE 5. C:N Analyses of Alginates Used in Metal Selectivity
Experiments
sample
µmol
of C/mg
µmol
of N/mg
%N
(wt/wt)
C:N
ratio
M. pyrifera (Sigma)
chitosana
S. fluitans
S. siliquosum
S. thunbergii
M. pyrifera
25.7 ( 0.3
26.1 ( 0.3
24.3 ( 0.3
24.6 ( 0.3
25.8 ( 0.1
24.7 ( 1.0
NDb
4.18 ( 0.10
NDb
NDb
0.76 ( 0.02
NDb
NDb
5.9
NDb
NDb
1.1
NDb
NDb
6.2
NDb
NDb
34
NDb
a Sample standard obtained courtesy of the Norwegian Biopolymer
Laboratory. b ND ) not detected.
FIGURE 3. Selectivity coefficients (K*MgCa) for the magnesiumcalcium ion-exchange (left axis) with alginates extracted from S.
fluitans (b) and M. pyrifera (9). Results are compared to selectivity
coefficients (right axis) from Figure 2 (K*CdCa) shown with open
symbols. XCa and error bars as in Figure 2.
FIGURE 4. Selectivity coefficients (K*MgCd) for the magnesiumcadmium ion-exchange (left axis) with alginates extracted from S.
thunbergii (b), S. siliquosum (2), and M. pyrifera (9). Results are
compared to selectivity coefficients (right axis) from Figure 2 (K*CdCa)
shown with open symbols. XCd and error bars as in Figure 2.
G-block rich alginates of Sargassum result in at least a
concomitant level of selectivity in the raw algal tissue.
Whereas experiments in this study were carried out at
high metal concentrations and ionic strength, previous
studies performed on Sargassum spp. in our laboratory have
demonstrated the selective nature of this brown algae at
concentration levels relevant to remediation applications.
These were both in the form of competitive equilibrium batch
experiments for the Cd-Fe system (40) as well as in studies
of flow-through columns packed with Sargassum algal
biosorbent in the Ca- and K-form (8, 41-42). The following
sequence of decreasing metal affinity for the Sargassum
biosorbent was determined on the basis of batch equilibrium
data (41)
Cu > Ca > Cd > Zn > Fe
266
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 2, 2003
and is well reflected by the differential breakthrough curves
which develop following passage of simulated waste-feed
streams through Sargassum packed columns. The higher
affinity of Ca relative to Cd for Sargassum in this affinity
sequence contrasts with the near unity of the Cd-Ca
selectivity coefficients determined in this study. Our results
may be explained by the observations of Andresen et al. ((36),
above), whereby the raw brown algal biomass can display an
even higher selectivity for calcium than its extracted alginate.
Practical Issues for Implementation of the Biosorption
Process. For the Cd-Ca exchange, virtually no selectivity is
observed for a range of FGG that spans the entire spectrum
of naturally occurring (13-15, 28, 31-33) brown algal derived
alginates. It is clear then, that the potential interference of
calcium in the binding of targeted toxic heavy metals such
as cadmium by alginates is great due to this lack of selectivity.
Calcium and cadmium have nearly the same ionic radius
(i.e. 1.00 and 0.95 Å, for a 6-fold coordination, respectively),
and their similar affinity can be attributed to the importance
of steric placement in the alginate gel network. The cavity
formed by the alignment of two chain sections of R-Lguluronic acid (Figure 1) results in a multidentate environment that consists of the carboxylate groups, the ringoxygens, and the hydroxyl groups of the G-residues, all of
which contribute to the binding of the cations. We believe
that the size of the cation is a key variable in metal binding,
both due to the rigid nature of the GG-linkages, as well as
to the steric arrangement of the electronegative ions surrounding the divalent cation. The lower selectivity of
magnesium for the GG-sequences is likely due to its smaller
ionic radius (0.72 Å, for a 6-fold coordination) which prevents
the formation of a tightly bound coordination environment
such as with calcium or cadmium.
In the case of monovalent cations such as Na+, the single
charge renders them incapable of forming network binding
junctions. Consequently, it is not surprising that in the NaCa mixed cation system, a high selectivity of Ca for G-block
rich alginates is observed (19, 43). Clearly network junctions
are formed by the calcium in the G-block sections but Na+
can, nevertheless, compete with Ca for binding sites in
alginates displaying lower FGG values.
The influence of Na+ on the sorption of cadmium by
Sargassum was studied by Schiewer and Volesky (11). The
addition of Na+ (at [Na] ) 0.6 mM for [Cd]f ) 0.01 mM or
at [Na] ) 2.5 mM for [Cd]f ) 0.1 mM) to a Na-free system
resulted in an estimated 10% reduction in Cd binding. They
concluded that for most biosorption applications (i.e. for
use as a ‘polishing’ step in the treatment of industrial
wastewaters) the influence of ionic strength should be
included in the modeling of such systems. This would
correspond to scenarios where [Metal] , [Na] < 1000 mM.
These Na concentrations are comparable to those employed
in Haug’s (19, 43) study of the Na-Ca alginate ion-exchange
system. The work described herein indicates that the
reduction in cadmium sorption observed by Scheiwer and
Volesky (11) likely represents a minimum because Na would
compete more effectively for cadmium if the brown algal
biomass contained alginate with significantly lower frequencies of G-blocks (or conversely, higher frequencies of
M-blocks or MG-blocks, as is the case of un-fractionated
Laminaria or bulk Macrocystis).
It is now clear that the composition of alginates extracted
from Sargassum species is relatively unique among the brown
algae, with the only other known source of alginates rich in
G-blocks being the stipes of L. hyperborea. Isolation of the
latter, G-block enriched alginate requires a mechanical
separation of the various parts of the algal tissue and would
be undesirable for the simple and economic implementation
of a remediation scheme, irrespective of the fact that this
Laminaria tissue is already a valuable commodity for other
markets. Furthermore, it has previously been demonstrated
(4) that certain species of Laminaria leach more of their
polysaccharide matrix than does Sargassum, under the acidic
conditions normally employed for biomass regeneration. This
important implementation parameter, which ultimately
dictates the lifetime of an ion-exchange column, is an
unavoidable downfall of certain brown algae (e.g., Ascophyllum nodosum, Fucus vesiculosus, Laminaria) despite
their very high metal uptake capacities. The combination of
mechanical stability (4), high metal uptake capacity (5), and
high guluronic acid content all contribute to make Sargassum
algal tissue a suitable material for toxic, heavy metal
remediation by the biosorption process.
Acknowledgments
This work was made possible by a Natural Sciences and
Engineering Research Council of Canada (NSERC) Seed Grant
to A.M. Additional financial support was provided by
individual NSERC Research grants to A.M. and B.V. as well
as funding from NSERC and the Fonds pour la Formation
des Chercheurs et l’Aide à la Recherche du Québec (FCAR)
to T.D. in the form of post-graduate scholarships. Courtesy
samples of Sargassum siliquosum and Sargassum oligocystum
were obtained from G. Diaz-Pulido and Dr. L. McCook
(Australian Institute of Marine Science and CRC Reef
Research), Sargassum thunbergii from Dr. Y. S. Yun (Postech
University), and Sargassum muticum from Dr. B. Farnham
(University of Portsmouth). The skillful technical assistance
of Joanna Hobbins and Ashley Meek is gratefully acknowledged. The authors also wish to acknowledge the assistance
of Drs. F. Morin and Z. Xia (Department of Chemistry, McGill
University) in acquisition of the NMR spectra. The authors
extend their gratitude to Professors Emeritus Arthur Perlin
and Bjørn Larsen for many stimulating discussions.
Note Added after ASAP
This paper was released ASAP on 11/27/2002 with Sargassum
thunbergii misspelled (including Figure 4). The correct
version was reposted on 12/04/2002.
Literature Cited
(1) Brierly, C. L. Geomicrobiol. J. 1990, 8, 201-223.
(2) Gadd, G. M. In Biotechnology; Rehm, H.-J., Ed.; VCH Verlagsgesellschaft: Weinheim, Germany, 1988; Vol. 6b: Special
Microbial Processes, pp 401-433.
(3) Volesky, B., Ed. Biosorption of Heavy Metals; CRC Press: Boca
Raton, Fl, 1990.
(4) Fourest, E.; Volesky, B. Appl. Biochem. Biotechnol. 1997, 67,
33-44.
(5) Davis, T. A.; Volesky, B.; Vieira, R. H. S. F. Water Res. 2000, 34,
4270-4278.
(6) Kratochvil, D.; Fourest, E.; Volesky, B. Biotechnol. Lett. 1995, 17,
777-782.
(7) Kratochvil, D.; Volesky, B.; Demopoulos, G. Water Res. 1997, 31,
2327-2339.
(8) Kratochvil, D.; Volesky, B. Water Res. 2000, 34, 3186-3196.
(9) Schiewer, S.; Volesky, B. Environ. Sci. Technol. 1995, 29, 30493058.
(10) Schiewer, S.; Volesky, B. Environ. Sci. Technol. 1997, 31, 18631871.
(11) Schiewer, S.; Volesky, B. Environ. Sci. Technol. 1997, 31, 24782485.
(12) Fourest, E.; Volesky, B. Environ. Sci. Technol. 1996, 30, 277282.
(13) McKee, J. W. A.; Kavalieris, L.; Brasch, D. J.; Brown, M. T.; Melton,
L. D. J. Appl. Phycol. 1992, 4, 357.
(14) Penman, A.; Sanderson, G. R. Carbohydr. Res. 1972, 25, 273282.
(15) Grasdalen, H.; Larsen, B.; Smidsrød, O. Carbohydr. Res. 1979,
89, 23-31.
(16) Haug, A.; Myklestad, S.; Larsen, B.; Smidsrød, O. Acta Chem.
Scand. 1967, 21, 768-778.
(17) Atkins, E. D. T.; Nieduszynski, I. A.; Mackie, W.; Parker, K. D.;
Smolko, E. E. Biopolymers 1973, 12, 1879-1887.
(18) Atkins, E. D. T.; Nieduszynski, I. A.; Mackie, W.; Parker, K. D.;
Smolko, E. E. Biopolymers 1973, 12, 1865-1878.
(19) Smidsrød, O.; Haug, A. Acta Chem. Scand. 1965, 19, 329-340.
(20) Kohn, R. Pure Appl. Chem. 1975, 42(3), 371-397.
(21) Morris, E. R.; Rees, D. A.; Thom, D. Carbohydr. Res. 1978, 66,
145-154.
(22) Morris, E. R.; Rees, D. A.; Thom, D. Carbohydr. Res. 1980, 81,
305-314.
(23) Rees, D. A. Pure Appl. Chem. 1981, 53, 1-14.
(24) McCook, L. J. Mar. Ecol. Prog. Ser. 1996, 139, 179-192.
(25) McCook, L. J. Mar. Biol. 1997, 129, 713-722.
(26) McCook, L. J.; Price, I. R.; Klumpp, D. W. Proc 8th Int. Coral Reef
Symp. 1997, 2, 1851-1856.
(27) Chemistry and enzymology of marine algal polysaccharides;
Percival, E., McDowell, R. H., Eds.; Academic Press: London,
U.K., 1967; pp 137-143.
(28) Davis, T. A.; Llanes, F.; Volesky, B.; Diaz-Pulido, G.; McCook, L.;
Mucci, A. Appl. Biochem. Biotechnol. (in press).
(29) Figueira, M. M.; Volesky, B.; Ciminelli, V. S. T.; Roddick, F. A.
Water Res. 2000, 34, 196-204.
(30) Smidsrød, O.; Haug, A. Acta Chem. Scand. 1968, 22, 1989-1997.
(31) Grasdalen, H.; Larsen, B.; Smidsrød, O. Carbohydr. Res. 1983,
118, 255-260.
(32) Haug, A.; Larsen, B.; Smidsrød, O. Carbohydr. Res. 1974, 32,
217-225.
(33) Llanes, F.; Sauriol, F.; Morin, F. G.; Perlin, A. S. Can. J. Chem.
1997, 75, 585-590.
(34) Jang, L. K.; Nguyen, D.; Geesey, G. G. Water Res. 1995, 29, 307313.
(35) Helrich, K. Ed. Official Methods of Analysis of the Association of
Official Analytical Chemists, 15th ed.; Association of Official
Analytical Chemists, Inc.: Arlington, Virginia, U.S.A., 1990; pp
59-60.
(36) Andresen, I.; Skipnes, O.; Smidsrød, O.; Østgaard, K.; Hemmer,
P. C. ACS Symp. Ser. 1977, No. 48, 361-381.
(37) Smidsrød, O. J. Chem. Soc., Faraday Trans. 1974, 57, 263-274.
(38) Haug, A.; Smidsrød, O. Nature 1967, 215, 757.
(39) Haug, A.; Smidsrød, O. Nature 1967, 215, 1167-1168.
(40) Figueira, M. M.; Volesky, B.; Ciminelli, V. S. T. Biotechnol. Bioeng.
1997, 54, 344-350.
(41) Kratochvil, D.; Volesky, B. Water Res. 1998, 32, 2760-2768.
(42) Figueira, M. M.; Volesky, B.; Azarian, K.; Ciminelli, V. S. T.
Environ. Sci. Technol. 2000, 43(20), 4320-4326.
(43) Haug, A. Acta Chem. Scand. 1959, 13, 1250-1251.
(44) Smidsrød, O.; Draget, K. Carbohydr. Europe 1996, 14, 6-13.
(45) Christensen, B. E.; Indergaared, M.; Smidsrød, O. Carbohydr.
Poly. 1990, 13, 239-255.
Received for review May 11, 2002. Revised manuscript received October 16, 2002. Accepted October 22, 2002.
ES025781D
VOL. 37, NO. 2, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
267