CRYSTAL

GROWTH
Calcium Carbonate Crystallization in the Presence of Biopolymers & DESIGN
Michael F. Butler,* Nicole Glaser, Antony C. Weaver, Mark Kirkland, and 2006
Mary Heppenstall-Butler VOL. 6, NO. 3
UnileVer R&D Colworth, Sharnbrook, Bedfordshire, MK44 1LQ, U.K. 781-794
ReceiVed August 23, 2005; ReVised Manuscript ReceiVed NoVember 7, 2005

ABSTRACT: The influence on calcium carbonate crystallization of a series of biopolymers that contain carboxylic acid or sulfate
functional groups was studied using pH and turbidity measurements, optical microscopy, and scanning electron microscopy. Without
biopolymer, single calcite (104) rhombohedra were formed. In the presence of nongelling biopolymers (xanthan and gellan) in the
conditions used, (104) rhombohedra formed aggregates that were “stack-like”, but in the presence of gelling biopolymers (pectin,
κ-carrageenan, and sodium alginate) the aggregates were “rosette-like”. The “rosettes” were proposed to form by the nucleation of
calcite on a gelled microparticle template to form a hollow shell. Low methoxy (LM) pectin was particularly effective at directing
the growth of calcite rosettes and led to aggregates of radially aligned crystals. The influence of biopolymer concentration on calcite
crystallization was studied for LM pectin and κ-carrageenan. In the former case, an increasingly favorable influence of the pectin
molecules on the surface energy of calcite nuclei was proposed to result in an enhanced propensity for nucleation, until the pectin
concentration was so high that all of the calcium was sequestered. In the latter case, an increase in calcium binding with increasing
κ-carrageenan concentration decreased the solution supersaturation and hence decreased the propensity for calcite formation.

Introduction by those present in biomineral forming systems, have also been

shown to strongly influence calcium carbonate growth. In some
The ability to control crystallization is a critical requirement cases, calcite is formed, but with morphologies other than
in the synthesis of many industrially important materials. rhombohedra, whereas in others different forms of calcite, such
Selected examples of these are pharmaceutical molecules, as hollow vaterite spheres, have been observed.10
triglycerides, fatty acids, and inorganic materials such as calcium
In the more complex case, polymers have been used to control
carbonate. Scientifically, calcium carbonate is of enormous
the nucleation and growth of calcium carbonate crystals. In many
interest in the context of crystal control because it is produced
cases, the field of biomineralization has motivated these studies.
by a wide variety of biological organisms that exhibit exquisite
Polymeric systems containing aspartic acid and glutamic acid
control over the polymorph, location of nucleation, crystal size
have received much attention because they are components
and shape, crystallographic orientation, composition, stability,
present in several natural shell-forming systems. Different
and hierarchical assembly of the crystals.
polymorphs of calcium carbonate were obtained depending on
Calcium carbonate can exist in several different polymorphs
the amino acid sequence and polypeptide conformation.11
that are, in order of increasing solubility: calcite, aragonite,
Atomic force microscope (AFM) studies, combined with mo-
vaterite, calcium carbonate monohydrate, and calcium carbonate
lecular modeling, suggest that specific binding of natural
hexahydrate. Although aragonite and calcite are the most
polypeptides, extracted from shells, to particular calcite crystal
commonly found polymorphs in natural systems, vaterite is also
faces, is responsible for the modification in morphology of
present in some systems, and all of the polymorphs may be
calcite crystals in these cases.12 In at least one case,13 a peptide
obtained in the laboratory. In the presence of water, all of the
was designed to interact with a specific crystal face, and the
polymorphs eventually transform into the most stable form,
resulting crystals had the predicted morphology, demonstrating
calcite. They may, however, be stabilized in the presence of
the feasibility of this explanation for the natural polypeptides.
additives. Amorphous calcium carbonate also exists but has only
Both lysozyme14 and collagen15 have been shown to systemati-
been observed in a stabilized form as a biomineral.1
cally influence the calcium carbonate morphology with increas-
Many studies have investigated the influence of additives on ing polypeptide concentration. In the case of lysozyme, the
the crystallization of calcium carbonate.2,3 In the simplest case, calcite crystal habit was modified as the growth of different
the addition of magnesium is well known to lead to the crystal faces was sequentially inhibited. In the case of collagen,
formation of spicular aragonite crystals rather than rhombohedral a transition was observed with increasing collagen concentration,
calcite. Simple organic acids such as citric and maleic acid have from rhombohedral calcite crystals to spherulitic crystal ag-
been shown to strongly influence crystal morphology, especially gregates, via multiple layer crystals, that was explained by
in the presence of magnesium ions,4 and studies have shown preferential adsorption of collagen. In the presence of soluble
that acidic peptides, in particular, aspartic acid, can alter the proteins extracted from coralline algae, amorphous calcium
chiral morphology of calcite crystals by binding to specific step carbonate has been stabilized.16
edges at the crystal surface.5 In general, aspartic acid has a large
effect on calcium carbonate crystal growth, causing either calcite Synthetic polymers have also been used to alter calcium
or vaterite to form depending on the acid concentration.6-8 Other carbonate growth. In some cases, simple anionic polymers, such
amino acids, such as alanine, lysine, and glycine, have similar as poly(acrylic acid), poly(acrylonitrile), and poly(styrene
effects.9 More complicated sequences of amino acids, inspired sulfonate), have been used to inhibit crystallization, for the
purpose of preventing limescale formation in boilers or water
pipes17-19 and, in the presence of chitin and cellulose films, to
* To whom correspondence should be addressed. E-mail:
[Link]@[Link]. Phone: +44 (0)1234 222958. Fax: +44 confine crystallization to the film surface by preventing bulk
(0)1234 248010. crystallization.20 Molecular modeling has shown the importance
10.1021/cg050436w CCC: $33.50 © 2006 American Chemical Society
Published on Web 02/14/2006
782 Crystal Growth & Design, Vol. 6, No. 3, 2006 Butler et al.

of the adsorbed polymer fine structure in controlling the proteoglycans in some biomineralizing systems. Although
crystallization inhibition via the number and strength of the κ-carrageenan had a relatively small influence on calcium
polymer-crystal interactions.21 In other cases, specific poly- carbonate growth,47 heparin was shown to form either aragonite
morphs, including calcite, vaterite, and calcium carbonate crystals48 or helicoidally arranged calcite aggregates49 depending
hexahydrate, have been obtained in the presence of synthetic on the experimental conditions. Hyaluronic acid caused the
polymers containing chemical groups, such as carboxylate, formation of columnar stacks of rhombohedral calcite crystals,
amide, and sulfate, that are known to interact with inorganic whereas individual cuboctahedral calcite crystals formed in the
salts.17,22-26 presence of keratan sulfate.49 No systematic attempt was made
More complex polymer architectures have also been studied. to relate the molecular structure of the sulfated polysaccharide
For instance, anionic carboxylate terminated poly(amidoamine) to the crystal habit in these studies.
dendrimers27-29 have been found to stabilize spherical vaterite More detailed studies have been performed on the formation
particles. Double hydrophilic block copolymers containing a of thin films of calcium carbonate on biopolymer substrates, in
weakly hydrophilic block and a polyelectrolyte block that an attempt to mimic the formation of the nacreous layers of
interacts with calcium carbonate,30 which are a simplified model calcium carbonate in mollusc shells that grow on biopolymer
system for active polypeptides in biomineralization with acidic templates. Several studies have demonstrated the formation of
domains, have also been found to stabilize vaterite as well as calcite films on insoluble polysaccharide (chitosan) substrates
amorphous calcium carbonate for short times.23,31-34 They lead in the presence of poly(acrylic) or poly(aspartic acid).20,50-54
to the formation of well-defined microparticles with a shape In these cases, the interaction of the polyanion with the chitosan
that predominantly depends on the chemical functionality of surface provided nucleation sites for the initiation of calcium
the polyelectrolyte block. As well as showing crystallization carbonate formation, while the presence of free polyanion in
inhibition effects,35 a wide variety of crystal morphologies have solution prevented bulk nucleation and growth and confined the
been obtained in the course of these studies, from the conven- crystal growth to the film on the insoluble substrate. Calcite
tional calcite rhombohedra, aragonite spicules, and vaterite films have been grown in a similar way on simple glass
spheres to hollow shells23 (sometimes formed from aggregates substrates, in the presence of poly(acrylic), poly(aspartic), and
of spheres36,37), ellipsoids,38 dumbbells,23,26,39,40 plates,26 stacks poly(glutamic) acid.55 In the presence of magnesium, aragonite
of rhombohedra,36 and disks.36 Helices41 and spheres42 have been films were grown on chitosan substrates.56
observed in the presence of synthetic block copolypeptides In the present study, the crystallization of calcium carbonate
containing serine and aspartate residues, respectively. Interest- in the presence of a range of biopolymers that can be considered
ingly, in the former case the chirality of the copolypeptide was to be natural double-hydrophilic block copolymers was inves-
used to direct the sense of the helix. tigated. They were chosen so that the effects of calcium binding,
Many polysaccharides are natural block copolymers and number, and type of side group (i.e., carboxylate and sulfate)
consist of two or more different glycosidic monomer units. on the resulting calcium carbonate morphology and unit cell
Alginates, which are extracted from seaweed, are composed of could be studied. As such, this report presents the first
mannuronic and guluronic acid residues, for example. Despite investigation of the influence of a range of readily available
their importance in the field of biomineralization, where anionic food biopolymers on calcium carbonate crystallization. The
polysaccharides provide the template for the intricate hierarchical biopolymers studied, shown in Figure 1, consisted of two types.
The first type, which were biopolymers that did not gel in the
assemblies of calcium carbonate crystals in coccoliths,43,44
presence of calcium, were gellan, xanthan, and κ-carrageenan.
relatively little work has been done on the systematic study of
Gellan and xanthan are acidic biopolymers that contain residues
the influence of simple polysaccharides on calcium carbonate
with carboxylic acid groups. κ-Carrageenan is a sulfated
crystallization.
polysaccharide. The second type of biopolymer studied were
Sodium alginate and carboxymethyl inulin have been found those that did chelate calcium to form gels. These were sodium
to inhibit the crystallization of calcium carbonate45,46 by alginate and pectin, both of which contain residues with
interacting directly with the surface of the growing crystals. carboxylic acid groups.
Although sodium alginate did not alter the standard rhombo-
hedral habit of calcite, carboxyethyl inulin caused spherical
Experimental Section
vaterite particles to grow in addition to rhombohedral calcite
crystals. For carboxymethyl inulin, the inhibition efficacy Materials and Sample Preparation, Control Samples. Calcium
increased with the carboxylate content of the polymer. In another carbonate was formed by mixing and continually stirring 10 mL of
study, different calcium carbonate polymorphs were obtained calcium chloride solution and 10 mL of sodium bicarbonate solution
(supplied by Sigma Chemical Company), with the concentrations shown
in the presence of different polysaccharides.47 Calcite and in Tables 1 and 2. In the first and second set of experiments, where
vaterite were formed in the presence of beta cyclodextrin and the ratio of calcium chloride/sodium bicarbonate concentration was 1:2
soluble starch, respectively. The presence of the different and 1:1, respectively, the mixture was raised to pH 10.5 by the dropwise
polymorphs was explained by the differences in geometrical addition of 5 M sodium hydroxide, potassium hydroxide, or ammonium
matching and stereochemical complementarity between the hydroxide solutions (supplied by Sigma Chemical Company). To test
calcium ions and the hydroxyl groups on the biopolymers. The for the influence of pH, a similar procedure was adopted for calcium
chloride/sodium bicarbonate concentrations of 0.01 M:0.02 M and 0.01
importance of stereochemistry was also demonstrated in a study M:0.01 M, using sodium hydroxide to achieve pH values between 8
of the influence of poly(alginic acid) and poly(galacturonic acid) and 12 inclusive, in increments of 0.5 pH units. Measurements of pH
on calcium carbonate growth. Despite containing similar were performed using a pH meter (HI8424, from Hanna Instruments).
numbers of carboxylate groups, poly(alginic acid) was found Materials and Sample Preparation, Samples in the Presence of
to be more effective than poly(galacturonic acid) at influencing Biopolymers. The biopolymers used in this study were gellan (gel F,
calcium carbonate growth morphologies at equal concentra- supplied by Kelco), xanthan (cold water soluble Keltrol RD, supplied
by Kelco), κ-carrageenan (Genugel X0909, supplied by Kelco), sodium
tions.44
alginate (Manugel DMB, supplied by Kelco), LM pectin (LM12, 35%
Some studies on sulfated polysaccharides have been re- de-esterified, supplied by Kelco), and high-methoxy (HM) pectin (65%
ported47-49 that were inspired by the presence of sulfated de-esterified, supplied by Kelco).
Calcium Carbonate Crystallization Crystal Growth & Design, Vol. 6, No. 3, 2006 783

Figure 1. Molecular structure of (a) xanthan, (b) gellan, (c) κ-carrageenan, (d) sodium alginate, and (e) pectin.

Table 1. Solution Concentrations Used in Experiment 1a to determine the critical entanglement concentration that separates the
mix 1 mix 2 mix 3 mix 4 mix 5 dilute concentration regime, where the polymers can be regarded as
separate entities, and the semidilute regime, where the molecules are
cCaCl2 (mol/ L) 0.01 0.025 0.05 0.075 0.10 entangled. The critical entanglement concentrations (%,w/v) were
cNaHCO3 (mol/ L) 0.02 0.050 0.10 0.150 0.20 sodium alginate, 1.75; κ-carrageenan, 1.75; LM pectin, 1.83; HM pectin,
a The pH was raised to 10.5. 0.75; gellan, 1.12; xanthan, 0.005.
Calcium carbonate was formed by mixing 10 mL of calcium chloride
Table 2. Solution Concentrations Used in Experiment 2a (0.01 M) solution with 10 mL of sodium bicarbonate (0.02 M) solution
mix 6 mix 7 mix 8 mix 9 mix 10 containing the biopolymer at the required concentration, made from
the stock solution. The pH of the mixture was increased to 10.5 using
cCaCl2 (mol/ L) 0.01 0.025 0.05 0.075 0.10 5 M sodium hydroxide solution to trigger the precipitation of calcium
cNaHCO3 (mol/ L) 0.01 0.025 0.05 0.075 0.10 carbonate crystals. A biopolymer concentration of 0.2% (by weight)
a The pH was raised to 10.5. was used for all of the samples because this was below the critical
entanglement concentration for all of the biopolymers except for
Stock solutions of 4% (by weight) were made from sodium alginate, xanthan.
κ-carrageenan, LM and HM pectin by stirring the dry biopolymer pH Measurements During Crystallization. A benchtop pH meter
powder into de-ionized water. The sodium alginate and κ-carrageenan (Hanna model 302) was used to continually measure the pH in mixtures
solutions were heated to 80 °C for 30 min to ensure complete containing different biopolymers. 30 mL of 0.01 M calcium chloride
dissolution, whereas the pectin solutions were stirred at room temper- solution was added to 30 mL of biopolymer solution containing 0.02
ature until completely dissolved. Stock solutions of 0.2% (by weight) M sodium bicarbonate in a glass beaker. The pH was increased to 10.5
of gellan and xanthan were made by adding the dry biopolymer powder by the dropwise addition of 5 M sodium hydroxide solution, and its
to de-ionized water and stirring at 80 °C for 30 min to ensure complete value was subsequently measured at intervals of 2 min. The mixture
dissolution. These stock solutions were diluted to the required final was continuously stirred during the experiment.
concentrations for the calcium carbonate growth experiments. Optical Microscopy. Transmission optical microscopy (Leitz Dia-
The zero shear rate viscosity of the different biopolymer solutions plan, set up for Kohler illumination) was used to determine the crystal
was measured for a variety of concentrations, using a dynamic stress habit of the calcium carbonate crystals, which had precipitated and aged
rheometer (Rheometrics DSR200) equipped with a Couette geometry, for 3 days, in the control samples and in the presence of biopolymers.
784 Crystal Growth & Design, Vol. 6, No. 3, 2006 Butler et al.

Figure 3. Typical calcium carbonate crystals observed in the control

samples: (a) single calcite rhombohedron and (b) aggregate of calcite
rhombohedra.
Figure 2. Variation in turbidity (measured by absorbance) with time
for a calcium carbonate crystallization control sample. also from a sample at pH 10.5. Scanning electron micrographs,
shown in Figure 4for samples at pH 10.5, revealed that the
The crystals were obtained by centrifuging the suspension in which aggregates were formed from several crystals that had inter-
they formed and washing with 0.1 M sodium hydroxide solution (to grown.
prevent dissolution during washing) three times. A drop of the final, pH Measurement. The pH variation after addition of alkali
cleaned suspension was placed on a microscope slide beneath a standard to the target pH was different for the samples with 1:2 and 1:1
glass cover-slip and observed under bright-field conditions with and
calcium chloride/sodium bicarbonate ratios. The 1:2 samples
without crossed polarizers.
Electron Microscopy. One drop of suspension of each sample,
remained roughly at the target pH value that was attained after
washed (with 0.1 M sodium hydroxide) and centrifuged three times, addition of alkali, shown in Figure 5a. The 1:1 samples
was placed onto a carbon-coated copper grid. The drop was allowed experienced a variation in pH with time, shown in Figure 5b,
to dry in air, and the residual crystals were sputter coated with a 10- which depended on the target pH value. Above approximately
nm layer of gold/palladium metal. A JEOL1200EX transmission pH 11, the pH remained constant. Below pH 11, a rapid decrease
electron microscope equipped with an ASID10 scanning attachment, was observed that reached a plateau around pH 7.5. Because
operating at an accelerating voltage of 20 kV, was used to obtain high-
the pH was found to be constant after addition of alkali for the
magnification images of the crystal morphologies. The images were
recorded using software (INCA) supplied by Oxford Instruments. 1:2 samples, subsequent experiments in the presence of biopoly-
Close-up images of the crystal surfaces were imaged using a JEOL mer were made with a 1:2 ratio of calcium chloride to sodium
6301 scanning electron microscope. The samples were coated prior to bicarbonate.
examination with a gold/palladium mixture in an Oxford CP2000 Crystallization in the Presence of Biopolymers. Turbidity.
preparation chamber. For all of the biopolymers, there was an induction time before
Turbidity. In-situ crystallization experiments were performed using the onset of crystallization was measured by an increase in
an ultra-violet/visible wavelength (UV/Vis) spectrophotometer (Perkin-
Elmer Lambda 40) to measure the absorption of light in a crystallizing
turbidity of the mixture. Figure 6 shows a typical plot of the
suspension of calcium carbonate at 500 nm wavelength, every 20 s. change in turbidity with time of a sample containing 0.2% (w/
The onset of crystallization was detected via an increase in turbidity, v) biopolymer (κ-carrageenan) at pH 10.5, measured using the
as measured by increasing absorbance of the light. UV/Vis spectrophotometer. The value of the induction time
Five milliliters of 0.01 M CaCl2 solution was added to 5 mL of varied within each sample, although the average induction time,
biopolymer solution containing 0.02 M NaHCO3 and 0.2% (w/v) shown in Table 3, depended on the type of biopolymer for
biopolymer while being stirred using a magnetic follower. The pH was samples with the same polymer concentration and final pH
increased to 10.5 using 1 M NaOH. Immediately afterward, the mixture
was transferred into the UV/Vis spectrophotometer. The time taken
value. It also depended on the biopolymer concentration, shown
between setting the pH to 10.5 and the turbidity measurements starting in Figure 7 for LM pectin and κ-carrageenan. The induction
was measured and accounted for during analysis of the results. time decreased with increasing pectin concentration, although
X-ray Diffraction. X-ray diffraction was performed at beamline 16.1 the entire sample formed a gel at concentrations above 1%, and
at the Synchrotron Radiation Source, Daresbury. One-dimensional no crystallization was observed at and above this concentration.
diffraction patterns were obtained from samples consisting of a In contrast, the reverse trend was observed for κ-carrageenan
sedimented suspension of calcium carbonate crystals with a thickness that showed a positive correlation between induction time and
of approximately 1 mm. The diffraction patterns were obtained on a
curved Inel multiwire detector. An exposure time of 30 s was used to biopolymer concentration.
collect the diffraction patterns. Optical and Electron Microscopy. Figure 8 shows optical
micrographs of samples, taken in bright-field mode and between
Results crossed polarizers, containing different biopolymers at 0.2% (w/
v) and at pH 10.5. The bright-field images showed that in all
Control Samples. Turbidity, Optical Microscopy, and cases the majority of the crystals were in the form of aggregates.
Electron Microscopy. For all of the control samples, crystal- Single rhombohedra were extremely rare. Crystals formed in
lization occurred immediately upon reaching the final pH and the presence of gellan and xanthan were of similar size and
was measured by a sudden increase in sample turbidity using were anisotropic, elongated aggregates. Close inspection of the
the UV/Vis spectrophotometer (see Figure 2 for a sample polarizing optical micrographs suggested that these aggregates
measured at pH 10.5). The turbidity rapidly reached a maximum were stacks of rhombohedra. In contrast, dumb-bell-shaped
value and then decreased again, owing to sedimentation of the aggregates were observed in the presence of LM and HM pectin,
crystals in the cuvette. and a mixture of spherical and ellipsoidal aggregates was
For all of the experimental conditions, the control samples observed in the presence of κ-carrageenan and sodium alginate.
yielded rhombohedral crystals. Sometimes single crystals were The crystalline aggregates formed in the presence of κ-carra-
obtained, shown in the optical micrograph in Figure 3a for a geenan appeared to be the smallest, whereas similarly sized
sample at pH 10.5, but in most cases aggregates of rhombohedra aggregates were formed in the presence of the different pectins
were obtained, shown in the optical micrograph in Figure 3b, and sodium alginate. These observations were not quantified,
Calcium Carbonate Crystallization Crystal Growth & Design, Vol. 6, No. 3, 2006 785

Figure 4. Scanning electron micrograph of calcium carbonate crystals grown in a control sample, showing a single crystal and crystal aggregates.

Figure 6. Variation in turbidity (measured by absorbance) with time

for calcium carbonate crystallization in the presence of 0.2% (w/v)
κ-carrageenan, as a typical example of a system displaying an induction
time before the onset of crystallization.

Figure 5. Variation of pH with time for samples made with (a) a 1:2
ratio of calcium chloride/sodium bicarbonate and (b) a 1:1 ratio of
calcium chloride/sodium bicarbonate

Table 3. Induction Times before the Onset of Calcium Carbonate

Crystallization in the Presence of Different Biopolymer Additives Figure 7. Variation of induction time before the onset of crystallization
(0.2% w/v) with biopolymer concentration for LM pectin and κ-carrageenan.

biopolymer induction time (minutes)

stacks of rhombohedra. The stacks were better defined for the
gellan 1.8 ( 0.5 crystals in the xanthan sample. For the pectin, alginate, and
xanthan 2.0 ( 0.7
κ-carrageenan 3.0 ( 1.1
κ-carrageenan samples the ellipsoidal and dumb-bell-shaped
sodium alginate 82.0 ( 46.6 aggregates appeared to be formed from interconnected and
HM pectin 5.5 ( 1.0 radially arranged crystals, which was most apparent for the LM
LM pectin 46.0 ( 6.9 pectin sample. The size of the crystals in the sodium alginate
however. Inspection of the polarizing optical micrographs sample was greater than for the pectin and κ-carrageenan
revealed that for the κ-carrageenan, pectin, and sodium alginate samples, which gave the aggregates formed in the presence of
samples Maltese cross patterns were formed, although their sodium alginate a slightly less well ordered appearance. In some
appearance did depend on the position of the crystalline cases, shown in Figure 10, for the samples containing pectin
aggregate relative to the focal plane of the microscope. and κ-carrageenan, hollow crystalline shells were observed. The
Scanning transmission electron micrographs for these crystals number of these was increased when powdered samples of these
are shown in Figure 9. The higher magnification attainable using crystals were crushed in a press.
electron microscopy confirmed the observations made using Figure 11 shows scanning electron micrographs of the surface
optical microscopy for crystals grown in the presence of gellan of the crystals at higher magnification, and therefore in greater
and xanthan. In these cases, the crystal aggregates were indeed detail, than the scanning transmission electron micrographs. The
786 Crystal Growth & Design, Vol. 6, No. 3, 2006 Butler et al.

concentrations of LM pectin and κ-carrageenan, at pH 10.5,

respectively. In all cases, similar types of crystalline aggregates
were formed that exhibited a Maltese cross pattern between
crossed polarizers in the case of LM pectin. For the LM pectin,
fewer crystals were observed at the higher biopolymer concen-
trations, although the crystal size was similar. For the κ-carra-
geenan samples, the crystals appeared to become more dumb-
bell shaped as the biopolymer concentration increased from
0.3% (w/v) to 1.8% (w/v), while crystal size and number were
similar (being small and high, respectively) at low and high
concentrations, with a reduced amount but larger size at
intermediate concentrations.
Scanning transmission electron micrographs of calcium
carbonate grown in the presence of different concentrations of
LM pectin and κ-carrageenan are shown in Figure 14. Similar
crystal morphologies were obtained for all of the concentrations
used. Figure 15 shows scanning transmission electron micro-
graphs of calcium carbonate grown in the presence of 0.2% (w/
v) LM pectin at different pH values. For all of the pH values
studied, similar crystal morphologies were obtained.
pH Measurement. Figure 16 shows the variation in pH with
time after addition of 1 M sodium hydroxide to three different
samples containing 0.2% (w/v) LM pectin, to attain pH 10.5.
The pH initially increased to about 10.6 and after approximately
20 min began to decrease again toward a plateau in the region
between 10.0 and 10.2. On average, the time before the pH
began to drop was less than the time taken before crystallization
was first detected from the increase in turbidity measured by
the UV/Vis spectrophotometer.
Crystalline Form. Figure 17 shows the wide-angle scattering
patterns from all of the samples, indexed according to calcite,
taken at least 2 days after crystallization occurred. For all of
the samples, calcite was the only crystalline form present. The
broad background scattering is caused by the biopolymer present
in the sample.

Discussion
Although calcite was formed in the control samples and in
the presence of biopolymers, the crystalline aggregates that were
formed in both cases were different shapes. Whereas the control
samples contained single and randomly clustered rhombohedra,
the rhombohedra that grew in the presence of biopolymers were
structured in particular arrangements. Furthermore, the crystals
that grew in the presence of the biopolymers appeared after an
induction time, signifying that there was an interaction between
the incipient calcium carbonate crystals and the biopolymer that
altered the nucleation kinetics.
It should be noted that, for the samples containing biopoly-
mers, the 1:2 ratio of calcium chloride/sodium bicarbonate was
Figure 8. Transmission optical micrographs (left-hand side) and used because, at pH values greater than 10.5, only carbonate
corresponding polarizing optical micrographs (right-hand side) of ions are formed, as described in the following reaction. In this
calcium carbonate crystals grown in the presence of 0.2% biopolymer: case, the hydrogen carbonate ions were in excess and acted as
(a) gellan, (b) xanthan, (c) κ-carrageenan, (d) sodium alginate, (e) LM a buffer in the presence of sodium hydroxide to maintain a
pectin, and (f) HM pectin. constant pH value.

|
surfaces of the crystals grown in the presence of xanthan H2CO3
appeared to be roughened, or composed of many nanocrystallites - +
CaCl2 + 2NaHCO3 w CaCO3 + 2NaCl + HCO3 + H
with sizes of approximately 100 nm, on all of the crystal faces.
Crystals grown in the presence of κ-carrageenan and pectin CO32- + 2H+
appeared to be roughened on some of the faces, but other faces
were smooth. All of the faces appeared to be smooth for the Below pH 10.5, hydrogen carbonate is a weak acid, and it
crystals grown in the presence of sodium alginate. can dissociate and liberate hydrogen ions, leading to a decrease
Figures 12 and 13 show bright-field and polarizing optical in pH. For the 1:1 ratio of calcium chloride/sodium bicarbonate,
micrographs of crystals formed in the presence of different there was no excess hydrogen carbonate, as it was all converted
Calcium Carbonate Crystallization Crystal Growth & Design, Vol. 6, No. 3, 2006 787

Figure 9. STEM micrographs of calcium carbonate crystals grown in the presence of 0.2% biopolymer: (a) gellan, (b) xanthan, (c) κ-carrageenan,
(d) sodium alginate, (e) LM pectin, and (f) HM pectin.

to carbonate ions. The system therefore had no buffering alginate,44,45 κ-carrageenan,48 and poly(galacturonic acid),44
capacity, and the excess hydrogen ions caused the decrease in which is similar to pectin.
pH that was measured at all of the pH values at which The current results, which showed that sodium alginate was
crystallization was initiated. Since it was desirable to conduct more effective at inhibiting calcium carbonate crystallization
experiments in which the formation of calcium carbonate was than either pectin or κ-carrageenan, are consistent with the
as regulated as possible, all of the experiments with biopolymers previous studies on alginate, poly(galacturonic acid) and κ-car-
were performed at pH 10.5, with the 1:2 ratio of calcium rageenan.44,48 In those studies, it was argued that the main factors
chloride/sodium bicarbonate. that contributed to the crystallization inhibition were charge
The induction time that was measured prior to the onset of strength, amount, and polymer conformation. The final factor
crystallization in the presence of the biopolymers was indicative was proposed because some polyanions, such as chondroitin
of the crystallization inhibition that has been measured in the sulfate,57 were investigated that, despite possessing many anionic
presence of many polymers, including biopolymers such as groups, had very little effect on calcium carbonate crystalliza-
788 Crystal Growth & Design, Vol. 6, No. 3, 2006 Butler et al.

Figure 10. STEM micrographs of calcium carbonate shells grown in the presence of (a) κ-carrageenan and (b) LM pectin.

Figure 11. SEM micrographs of the surfaces of calcium carbonate crystals grown in the presence of (a) xanthan, (b) κ-carrageenan, (c) sodium
alginate, and (d) HM pectin.

tion. It was concluded that additional factors, such as polymer is unlikely to be the reason for the difference in induction time
conformation that may lead to a reduced availability or between LM pectin and sodium alginate. Pectin has been
dissipation of charge, were important. reported to be generally more flexible than sodium alginate.58
The first factor, strength of charge, explains why κ-carrag- It might therefore be expected that the carboxylate groups in
eenan had less effect on calcium carbonate crystallization than pectin would be more available for binding to calcium, leading
did HM pectin, despite possessing similar numbers of anionic to more efficient calcium sequestration by pectin and hence a
groups. The lower partial charge on the oxygen atoms of the longer induction time. However, the reverse was actually
anionic sulfate groups in κ-carrageenan compared to the anionic observed. It should be noted that the conformation of the
carboxylate groups in pectin will result in a weaker interaction polymers used in the current study is not known. In conclusion,
with the calcium ions at the surface of calcium carbonate crystals the difference between the induction times of LM pectin and
and therefore a weaker influence on calcium carbonate crystal- sodium alginate remains unknown.
lization. The second factor, number of charges, explains the The observation that, in the systems containing LM pectin,
much smaller induction times for gellan and xanthan compared the pH began to drop prior to the formation of calcium carbonate
to pectin and alginate, HM pectin compared to LM pectin, and detected by turbidity, suggests that events began to occur that
LM pectin compared to κ-carrageenan. All of these polymers influenced crystallization prior to the nucleation event itself.
possess anionic carboxylate groups, but xanthan and gellan One explanation is that the drop in pH demonstrates the
contain far fewer carboxylate groups than pectin or alginate. interaction of LM pectin with incipient calcium carbonate nuclei,
Similarly, HM pectin contains fewer carboxylate groups than hindering their formation and delaying crystallization until later
LM pectin. The smaller number of charges therefore leads to times. A previous study of calcium carbonate crystallization in
fewer possible interactions between the polyanion and the the presence of sodium alginate45 has shown the large influence
surface of the calcium carbonate crystal and weaker interaction that this polyanion can have on calcium carbonate crystal growth
overall. The third factor, conformation and availability of charge, kinetics. It was suggested that the presence of the alginate altered
Calcium Carbonate Crystallization Crystal Growth & Design, Vol. 6, No. 3, 2006 789

Figure 12. Transmission optical micrographs (left-hand side) and corresponding polarizing optical micrographs (right-hand side) of calcium carbonate
crystals grown in the presence of different concentrations of LM pectin: (a) 0.2%, (b) 0.3%, and (c) 0.4%.

the crystal growth mechanism from the normal spiral growth induced the formation of a monocrystalline aggregate, or for
mechanism,59 which has been observed using AFM in the the PEG-b-PMAA CTAB mixture, for the terpolymer it was
presence of some small molecular weight additives,5,60,61 to one suggested that the molecule existed in an extended conformation
controlled by surface nucleation. owing to the large number of ionizable sulfate units on the chain.
The crystal morphologies were all related to the (104) calcite Because the terpolymer also possessed units (>CdO, >SdO,
rhombohedra obtained in the control samples, as expected from and >NdH) that are known to interact with calcium ions, it
the XRD data that showed that calcite was the only crystal was suggested that they nucleated the formation of calcite
polymorph present in all cases. The effect of the biopolymers
crystals along the chain, leading to the formation of stacks. A
was apparent, however, in the overall morphology of the
similar explanation may favor the formation of stacks for
crystals. Most obvious was the distinction between the different
types of biopolymer: those that did not form gels in the presence hyaluronic acid and PEG-b-PMAA CTAB, as well as for
of calcium ions and those that did. The former biopolymers, xanthan and gellan in the current study. All of these polymers
gellan and xanthan, formed aggregates of crystals that could contain ionizable carboxylate groups that can become charged
be described as roughly stack-like, whereas the latter biolymers, at high pH values as well as interacting with calcium ions. In
pectin (LM and HM) and alginate, formed aggregates that could addition, xanthan and gellan are rather stiff molecules by virtue
be described as rosette-like. Furthermore, the rosette-like of being polysaccharides. It is therefore plausible to suggest
aggregates appeared to be hollow shells. κ-Carrageenan, that that, in the high pH conditions used in the present study
forms a gel in the presence of sodium but not calcium ions, (>pH10.5), the gellan and xanthan molecules became ionized,
formed a morphology that was intermediate between stack- and extended, and also nucleated calcite crystallization via the
rosette-like. interaction between the carboxylate groups on the polymer and
Stack-like morphologies have been obtained in the presence the calcium ions in solution, thereby forming stacks of calcite
of hyaluronic acid,49 a carboxylated biopolymer, a mixed rhombohedra. That there is an interaction between xanthan and
solution of a block-copolymer (poly(ethylene glycol)-block-poly- calcium carbonate is suggested by the high-resolution SEM
(methacrylic acid), PEG-b-PMAA) and a cationic surfactant image of the calcite crystal surface grown in the presence of
(cetyltrimethylammonium bromide, CTAB),36 and a water-
xanthan, which shows a high degree of surface roughness
soluble terpolymer (poly(acrylamide-co-2-acrylamido-2-methyl-
consistent with the presence of bound impurities, i.e., xanthan
1-propane sodium sulfonate-co-n-vinyl-pyrrolidone)26 that con-
tained several functional groups known to interact with calcium molecules, at the interface.
ions. Although no explanation was given for the formation of Rosette-like aggregates of calcium carbonate crystals have
the stacks in the case of hyaluronic acid, other than the acid been observed previously in the presence of biopolymers, such
790 Crystal Growth & Design, Vol. 6, No. 3, 2006 Butler et al.

was believed that lattice matching of the functional groups in a

helical conformation on the polymer, in the case of heparin, or
aggregates of proteins, in the case of tobacco mosaic virus, acted
as a template for the helical nucleation of calcite crystals into
a rosette-like aggregate. In the presence of gelatin and magne-
sium it was shown that the rosette-like structures, that were very
similar in appearance to those formed in the current study in
the presence of LM pectin, were spherulitic structures formed
from a nucleus of aligned magnesium calcite prisms. Subsequent
crystallization led to an angular spread in the orientation of the
prisms that increased until the aggregate ended in a globular
shape. A similar spherulitic growth has been observed in similar
inorganic systems: octacalcium phosphate in the presence of
poly(aspartic acid) or poly(acrylate)66 and fluoroapatite in the
presence of gelatin.67
The biopolymer systems in which rosette-like aggregates
formed in the present case, namely, pectin (HM and LM) and
alginate, neither form helical conformations nor aggregate into
them. Therefore, the rosettes are not likely to have formed via
direct templating on the biopolymer chains themselves. The
spherulite growth mechanism proposed in the gelatin study is
also not a likely explanation for the rosettes formed in the
present study, since that relied upon the formation of aligned
magnesium calcite prisms, and no such crystals were observed.
A more likely explanation for the formation of the rosettes in
the presence of pectin and alginate is suggested by previous
studies of templated growth of calcium carbonate on certain
substrates, combined with the observations in the present case
that the rosettes formed only in the presence of biopolymers
that gelled in the presence of calcium ions and that, in some
cases at least, the calcite crystals were shown to form hollow
shells.
Growth of calcium carbonate in the presence of p-mercap-
tophenol colloidal gold seeds leads to the formation of rosette-
like crystals with a radial arrangement of calcite rhombohedra
parallel to the [001] crystallographic direction.68 In this case,
the calcium carbonate was believed to have been directly
nucleated by the modifed gold colloids, with the final morphol-
ogy being influenced by crystal-crystal interactions at later
stages of growth that led to mutual frustration of growth in the
direction tangential to the growing rosette. In another study,
calcite platelets, formed in lamellar surfactant structures used
to stabilize aqueous foams, have been shown to assemble into
shells with a rosette-like appearance.69
Rosettes have also been formed in the presence of PEG-b-
PMAA,23,37,70 PEG-b-PMAA-aspartate (Asp),23 poly(styrene-
block-acrylic acid) (PS-b-PAA),34 and PEO-b-PHEE30 double
hydrophilic block copolymers as well as a solution containing
poly(ethylene oxide-block-methacrylic acid (PEO-b-PMAA) and
surfactant, sodium dodecyl sulfate (SDS).36 The block copoly-
mer was present in the calcite in all cases and was believed to
strongly interact by virtue of lattice matching between the
functional groups on the polymer and the calcium ions in the
crystal.30,34 For the PEG-b-PMAA-containing systems at least,
hollow aggregates of calcite crystals arose from the initial
formation of a calcium carbonate particle composed either of
amorphous calcium carbonate23 or an aggregate of vaterite
Figure 13. Transmission optical micrographs (left-hand side) and spheres.37 The block copolymer was then proposed to template
corresponding polarizing optical micrographs (right-hand side) of calcite nucleation and growth on the surface of the core calcium
calcium carbonate crystals grown in the presence of different concentra- carbonate particle. The outer layer of facetted calcite crystals
tions of κ-carrageenan: (a) 0.2%, (b) 0.3%, (c) 1.0%, (d) 1.8%, (e)
then grew at the expense of the core, which dissolved to provide
2.0%, (f) 2.5%, and (g) 3.0%.
material for crystal growth. In a similar manner, surface
as heparin,47,49 tobacco mosaic virus,62 gelatin (with added functionalized gold nanoparticles adsorbed onto vaterite spheres
magnesium),63-65 and collagen.15 In the former two cases, it have been shown to result in templated aragonite overgrowth
Calcium Carbonate Crystallization Crystal Growth & Design, Vol. 6, No. 3, 2006 791

Figure 14. STEM images of calcium carbonate grown in the presence of (a) LM pectin, 0.2%, (b) LM pectin, 0.4%, (c) κ-carrageenan, 0.2%, and
(d) κ-carrageenan, 2.5%.

leading to rosette-shaped crystal aggregates.71 Surface func- gel because there will be less free calcium in the interior of the
tionalized dendrimer molecules, containing moieties that interact gel that is available to form calcite crystals. An insufficient
with calcium at the outer surface of the dendrimer, also lead to number of open shells were observed to test this hypothesis,
a spherical overgrowth of calcium carbonate crystals.27,28 however. Future studies will concentrate on developing an
Pectin and alginate are biopolymers that interact strongly with understanding of the hollow calcite shells that are proposed to
calcium ions, to the extent that, at the right concentration, a gel form in the presence of pectin or alginate.
is formed whereby the calcium ions form physical cross-links Interestingly, the rosette-like aggregates formed in the pres-
between different biopolymer chains. It is therefore likely that ence of LM pectin appeared more ordered than those formed
the addition of calcium chloride to the solution containing pectin in the presence of HM pectin and alginate, shown both by the
or alginate and sodium bicarbonate led to the formation of small appearance of the rosettes in the STEM images and by the
regions of gelled biopolymer that contained bound calcium ions presence of Maltese crosses in the rosettes observed between
that could act as nuclei for calcium carbonate crystal growth. crossed polarizers in the optical microscope. The Maltese cross
In the same manner as the double hydrophilic block copolymers patterns showed that, for LM pectin, radial growth of the calcite
or colloidal gold particles templated the growth of calcite or rhombohedra occurred with the crystals all possessing a uniform
aragonite on a particle core, leading to the formation of a orientation, i.e., the growth occurred in a directed manner. This
crystalline shell, it is proposed that in the current study the pectin marked degree of orientation implies that there was a relation
or alginate that formed the core also templated the nucleation between the underlying LM pectin template and the overgrowth
of the calcite overgrowth. The common feature in all of the of calcite crystals that formed the rosette-like aggregate. The
systems is that a species is present that contains chemical groups, possibility of such a direct interaction is suggested by studies
such as carboxylate groups, that interact strongly with calcium. of calcium carbonate growth on biopolymer films, such as
The difference between the systems is that in the previous chitosan in the presence of poly(acrylic acid)20,50-55 and
studies the crystalline shell grew on a calcium carbonate core cellulose.54,72 In the case of calcite grown on poly(acrylic acid)
that subsequently dissolved, whereas in the present study the bound to insoluble chitosan films, a direct match between the
interior of the rosette-like aggregate is proposed to contained a separation of the calcium ions in the calcite crystal on the (001),
biopolymer core. Depending on the amount of free calcium ions (110), and (104) faces and the spacing of the carboxylate groups
present within the biopolymer gel microparticle, it is therefore on the polymer was used to explain the crystal morphologies
possible to form calcite within the gel as well as at the surface, obtained.50,51,55 Although detailed crystallographic data are not
which explains why, in some cases, separate calcite aggregates available from the present study, future investigations on
were observed that fitted perfectly within the rosette-like shell. templated growth on biopolymer films will explore the nature
Presumably, crystallization within pectin or alginate would be of any potential crystallographic templating effect in the
less likely to occur than in the solution near the surface of the presence of LM pectin.
792 Crystal Growth & Design, Vol. 6, No. 3, 2006 Butler et al.

Figure 15. STEM images of calcium carbonate grown in the presence of 0.2% LM pectin at different pH values: (a) 10, (b) 10.5, (c) 11, and (d)
11.5.

to some extent for all of the other systems containing rosette-

like aggregates, including κ-carrageenan. In the case of fluo-
roapatite grown in the presence of gelatin,67 dumb-bell-shaped
spherulites were obtained that were explained, and successfully
reproduced, by a fractal model of crystal growth. Modification
of crystal growth by local electric fields resulting from crystal-
crystal and crystal-polymer interactions were proposed to lead
to a divergence of crystal orientation from a central, flat,
hexagonal seed, eventually leading to a dumb-bell-shaped
aggregate with the crystals arranged in the same way as the
electrical field lines around a permanent dipole. However,
although the fractal model did not fully explain the dumb-bell
morphologies that grew on spherical calcium carbonate ag-
Figure 16. Change in pH with time during crystallization in solutions
containing LM pectin. gregates in the presence of double hydrophilic block copoly-
mers,39 enough similarities were observed with the fluoroapatite
aggregates to suggest that a similar spherulitic growth mecha-
nism could explain the calcium carbonate dumb-bells in the later
stages of aggregate growth. It is likely, therefore, that the dumb-
bells formed in the present study were a result of a spherulitic
growth mechanism in which the gel particle at the core of the
aggregate acted as the precursor wheat sheaf particle. If
nucleation did not occur uniformly across the gel microparticle,
the initial crystals that formed at the particle surface would
produce the asymmetry present in the wheat-sheaf shape of
spherulite precursors that have been definitely shown to produce
dumb-bells. It would therefore be possible for a spherical
nucleus to form dumb-bells.
In the case of κ-carrageenan, an intermediate morphology was
Figure 17. X-ray diffraction patterns of the calcium carbonate crystals
obtained that lay between the stack-like morphology observed
grown in the presence of different biopolymers, indexed according to
calcite. in the presence of nongelling biopolymers and the rosette-like
morphology obtained in the presence of calcium-gelling biopoly-
It was also observed that, in the present study, the rosette- mers. However, the aggregates formed in the presence of
like aggregates often formed in the shape of dumb-bells. This κ-carrageenan also appeared to be hollow, in at least some cases.
effect was most noticeable for LM pectin but could also be seen These observations can be explained using the hypothesis
Calcium Carbonate Crystallization Crystal Growth & Design, Vol. 6, No. 3, 2006 793

proposed for the calcium-gelling biopolymers, since κ-carrag- 0.4%, however, a decrease in induction time was measured that
eenan is known to form gels in the presence of sodium ions. cannot be explained by the effect of the polymer on supersatu-
Sodium ions were present in the current study since the reaction ration. Since LM pectin formed the most ordered radially
to form calcium carbonate was performed using sodium oriented aggregates of crystals and was therefore proposed to
bicarbonate. Therefore, it is likely that a weak gel of κ-carra- act as a direct template for calcite growth, it is possible that the
geenan formed. Furthermore, it is known that the cross-links in decrease in induction time is indicative of a decrease in surface
κ-carrageenan gels are formed from two chains interacting in a energy that is increasingly favorable for nucleation despite the
double helix. The presence of a gel therefore provides the reduction in solution supersaturation. Such an effect could occur
template for the formation of a hollow shell of calcite crystals if, at the higher LM pectin concentrations, the separation
in the same rosette form observed in the presence of pectin and between the templating carboxylate groups became increasingly
alginate, whereas the presence of extended double-helical matched to the lattice spacing between calcium ions on the
regions provides the reason for the extended nucleation in the templated crystal plane.
stack form observed in xanthan and gellan. For κ-carrageenan, that does not gel in the presence of calcium
From the above discussion, it is possible to explain the effect ions but is still expected to interact via the sulfate groups, the
of pH and biopolymer concentration on the crystal morphology. overall increase in induction time is expected owing to an
First, the overall similarity in aggregate morphology for the increase in the amount of bound calcium with increasing
crystals grown in the presence of LM pectin at different pH polymer concentration. The effect of supersaturation was
values is to be expected, since in all cases crystal growth was therefore dominant overall. Interestingly, however, a peak in
controlled only by the presence of the nucleating LM pectin the induction time was measured, superimposed on the overall
gel template and the presence of calcium and carbonate ions in increase, at a κ-carrageenan concentration around 2%, that
solution, which will be similar at all of the pH values studied. coincided with the observation of fewer nuclei and larger crystals
The less-ordered aggregates observed at higher pH values reflect using optical microscopy. The subsequent reduction in induction
the faster growth rates caused by the higher supersaturation of time at the κ-carrageenan concentration of 2.5% indicated that
carbonate ions in solution at higher pH. Second, the overall a compensating effect was present at this concentration that
similarity of the morphologies observed at different biopolymer served to increase the likelihood of crystallization. As for the
concentrations can also be explained for similar reasons. That LM pectin case, it is possible that at this concentration the
the biopolymers studied interact in solution to a lesser extent average spacing of the sulfate groups provided a better match
than the double hydrophilic block copolymers reported in the between the polymer and the incipient calcite crystals, thereby
literature is shown by the presence of the rhombohedral (104) reducing the surface energy of the calcite nuclei and promoting
calcite form in all of the biopolymer systems in the present nucleation. Further experiments will be performed on two-
study, over the range of concentrations used, in the present study. dimensional biopolymer surfaces to investigate these phenom-
For the block copolymer systems, differences in crystal habit ena.
were also observed over particular polymer concentration ranges,
indicating a strong interaction of those polymers with particular Conclusions
crystallographic planes that markedly altered their growth
rates.23,36 The influence on calcium carbonate crystallization of a series
The relationship between crystallization kinetics and the of food biopolymers, which contain carboxylic acid or sulfate
concentration of LM pectin and κ-carrageenan is less straight- functional groups, was studied using a variety of techniques,
forward to explain, however. The nucleating ability of polymers including pH and turbidity measurements, optical microscopy,
is correlated with the ability of the polymer to bind metal ions and scanning electron microscopy. The biopolymers chosen
that intiates the formation of subcritical nuclei that grow to the were gellan, xanthan, LM pectin, HM pectin, and sodium
critical size required for crystal growth.72 From standard alginate, which contain carboxylate groups, and κ-carrageenan,
nucleation theory, the induction time, τ, is related to the solution which contains sulfate groups.
supersaturation, Ω, by the following equation: In control samples containing no biopolymer, single calcite

( )
(104) rhombohedra were formed. In the presence of biopoly-
âυ2γs2 mers, (104) rhombohedra were formed as aggregates that were
1
log τ ∝ either “stack-like” or “rosette-like”. Stacks were formed in the
(2.303kBT)3 (log Ω)2 presence of nongelling biopolymers, which were nucleated by
the carboxylate groups on extended xanthan or gellan chains.
where â is a shape factor for the calcite nuclei () 16π/3 for Rosettes were formed in the presence of calcium-gelling
spherical shapes), ν is the molar volume of calcite () 1.89 × biopolymers and were proposed to form by the nucleation of
10-5 m-3), γs is the surface energy of the calcite nuclei, kB is calcite on a gelled microparticle template. LM pectin was
Boltzmann’s constant, and T is the temperature. Therefore, as particularly effective at directing the growth of calcite rosettes
the polymer concentration changes the two factors that may be and led to aggregates of radially aligned crystals. Evidence was
affected are the solution supersaturation, as the polymer binds found that the rosettes were hollow. The influence of biopolymer
more calcium from solution, and the nucleus surface energy, concentration on calcite crystallization was studied for LM
as the polymer may become associated with the growing crystal. pectin and κ-carrageenan. In the former case, an increasingly
In the case of LM pectin, it is known that the polymer binds favorable influence of the pectin molecules on the surface energy
calcium ions, as this is how the pectin gel is formed. At of calcite nuclei was believed to result in an enhanced propensity
concentrations above 1%, no calcite crystals formed, and the for nucleation, until the pectin concentration was so high that
entire solution became a single lump of gel. In this case, the all of the calcium was sequestered. In the latter case, an increase
LM pectin had bound all of the available calcium ions leaving in calcium binding with increasing κ-carrageenan concentration
none free for crystallization, which is represented in the above generally decreased the solution supersaturation and hence
equation for the induction time as the limit when supersaturation decreased the propensity for calcite formation. At certain higher
tended to zero. At the concentrations studied, between 0.2 and concentrations, however, it was possible that the κ-carrageenan
794 Crystal Growth & Design, Vol. 6, No. 3, 2006 Butler et al.

conformation became important and favored calcite nucleation (34) Linhai, Y.; Dalai, J. Chin. Sci. Bull. 2004, 49, 235.
via a surface energy decrease in the same way as for LM pectin, (35) Sedlak, M.; Antonietti, M.; Cölfen, H. Macromol. Chem. Phys. 1998,
199, 247.
as an enhancement in nucleation was observed. (36) Qi, L.; Li, J.; Ma, J. AdV. Mater. 2002, 14, 300.
(37) Yu, S.-H.; Cölfen, H.; Antonietti, M. J. Phys. Chem. B 2003, 107,
Acknowledgment. The authors thank Unilever for permis- 7396.
sion to publish this paper. (38) Marentette, J. M.; Norwig, J.; Stöckelmann, E.; Meyer, W. H.;
Wegner, G. AdV. Mater. 1997, 9, 647.
(39) Cölfen, H.; Qi, L. Chem. Eur. J. 2001, 7, 106.
References (40) Endo, H.; Schwahn, D.; Cölfen, H. J. Phys. Chem. 2003, 107, 7396.
(1) Addadi, L.; Raz, S.; Weiner, S. AdV. Mater. 2003, 15, 959. (41) Sugawara, T.; Suwa, Y.; Ohkawa, K.; Yamamoto, H. Macromol.
(2) Cölfen, H. Curr. Opin. Coll. Int. Sci. 2003, 8, 23. Rapid Commun. 2003, 24, 847.
(3) Mann, S.; Perry, C. C. AdV. Inorg. Chem. 1991, 36, 137. (42) Euliss, L. E.; Trnka, T. M.; Deming, T. J.; Stucky, G. D. Chem.
(4) Meldrum, F. C.; Hyde, S. T. J. Cryst. Growth 2001, 231, 544. Commun. 2004, 15, 1736.
(5) Orme, C. A.; Noy, A.; Wierzbicki, A.; McBride, M. T.; Grantham, (43) Marsh, M. E. in Biomineralization: From Biology to Biotechnology
M.; Teng, H. H.; Dove, P. M.; DeYoreo, J. J. Nature 2001, 411, and Medical Applications; Baeuerline, E., Ed.; Wiley-VCH: Wein-
775. heim, 2000; pp 251-268.
(6) Volkmer, D.; Fricke, M.; Huber, T.; Sewald, N. Chem. Commun. (44) Didymus, J. M.; Oliver, P.; Mann, S.; DeVries, A. L.; Hauschka, P.
2004, 16, 1872. V.; Westbroek, P. J. Chem. Soc., Faraday Trans. 1993, 89, 2891.
(7) Malkaj, P.; Dalas, E. Cryst. Growth Des. 2004, 4, 721. (45) Manoli, F.; Dalas, E. J. Mater. Sci.: Mater. Med. 2002, 13, 155.
(8) Tong, H.; Ma, W.; Wang, L.; Peng, W.; Hu, J.; Cao, L. Biomaterials (46) Verraest, D. L.; Peters, J. A.; van Bekkum, H.; van Rosmalen, G.
2004, 25, 3923. M. J. Am. Oil Chem. Soc. 1996, 73, 55.
(9) Manoli, F.; Kanakis, J.; Malkaj, P.; Dalas, E. J. Cryst. Growth 2002, (47) Falini, G.; Gazzano, M.; Ripamonti, A. J. Cryst. Growth 1994, 137,
236, 363. 577.
(10) Li, C. M.; Botsaris, G. D.; Kaplan, D. L. Cryst. Growth Des. 2002, (48) Yang, L.; Zhang, X.; Liao, Z.; Guo, Y.; Hu, Z.; Cao, Y. J. Inorg.
2, 387. Biochem. 2003, 97, 377.
(11) Levi, Y.; Albeck, S.; Brack, A.; Weiner, S.; Addadi, L. Chem. Eur. (49) Arias, J. L.; Neira-Carrillo, A.; Arias, J. I.; Escobar, C.; Bodero, M.;
J. 1998, 4, 389. David, M.; Fernández, M. S. J. Mater. Chem. 2004, 14, 2154.
(12) Wierzbicki, A.; Sikes, C. S.; Madura, J. D.; Drake, B. Calcif. Tissue (50) Zhang, S.; Gonsalves, K. E. J. Appl. Polym. Sci. 1995, 56, 687.
Int. 1994, 54, 133. (51) Zhang, S.; Gonsalves, K. E. Mater. Sci. Eng. C 1995, 3, 117.
(13) DeOliveira, D. B.; Laursen, R. A. J. Am. Chem. Soc. 1997, 119, (52) Kato, T.; Suzuki, T.; Amamiya, T.; Irie, T.; Komiyama, M.; Yui, H.
10627. Supramol. Sci. 1998, 5, 411.
(14) Jimenez-Lopez, C.; Rodriguez-Navarro, A.; Dominguez-Vera, J. M.; (53) Kato, T.; Suzuki, T.; Irie, T. Chem. Lett. 2000, 2, 186.
Garcia-Ruiz, J. M. Geochim. Cosmochim. Acta 2003, 67, 1667. (54) Hosoda, N.; Kato, T. Chem. Mater. 2001, 13, 688.
(15) Shen, F. H.; Feng, Q. L.; Wang, C. M. J. Cryst. Growth 2002, 242, (55) Kotachi, A.; Miura, T.; Imai, H. Chem. Mater. 2004, 16, 3191.
239. (56) Sugawara, A.; Kato, T. Chem. Commun. 2000, 6, 487.
(16) Raz, S.; Weiner, S.; Addadi, L. AdV. Mater. 2000, 12, 38. (57) Meyer, H. J. J. Cryst. Growth 1984, 66, 639.
(17) Kawaguchi, H.; Hirai, H.; Sakai, K.; Sera, S.; Nakajima, T.; Ebisawa, (58) Harding, S. E. Prog. Biophys. Mol. Biol. 1997, 2/3, 207.
Y.; Koyama, K. Colloid Polym. Sci. 1992, 270, 1176. (59) Liu, X. Y.; Boek, E. S.; Briels, W. J.; Bennema, P. Nature 1995,
(18) Rieger, J.; Thieme, J.; Schmidt, C. Langmuir 2000, 16, 8300. 374, 342.
(19) Shakkthivel, P.; Sathiyamoorthi, R.; Vasudevan, T. Desalination (60) Reyhani, M. M.; Oliveira, A.; Parkinson, G. M.; Jones, F.; Rohl, A.
2004, 164, 111. L.; Ogden, M. I. Int. J. Mod. Phys. B 2002, 16, 25.
(20) Iwatsubo, T.; Sumaru, K.; Kanamori, T.; Yamaguchi, T.; Sinbo, T. (61) de Yoreo, J. J.; Dove, P. M. Science 2004, 306, 1301.
J. Appl. Polym. Sci. 2004, 91, 3627. (62) Sinha, A.; Chakraborty, J.; Das, S. K.; Ramachandrarao, P. Curr.
(21) Hädicke, E.; Rieger, J.; Rau, I. U.; Boeckh, D. Phys. Chem. Chem. Sci. 2003, 84, 1437.
Phys. 1999, 1, 3891. (63) Falini, G.; Fermani, S.; Gazzano, M.; Ripamonti, A. J. Chem. Soc.,
(22) Verdoes, D.; van Landschoot, R. C.; van Rosmalen, G. H. J. Cryst. Dalton Trans. 2000, 10, 3983.
Growth 1990, 99, 1124. (64) Falini, G.; Fermani, S.; Gazzano, M.; Ripamonti, A. Chem. Eur. J.
(23) Cölfen, H.; Antonietti, M. Langmuir 1998, 14, 582. 1997, 3, 1807.
(24) Dalas, E.; Klepetsanis, P.; Koutsoukos, P. G. Langmuir 1999, 15, (65) Falini, G. Int. J. Inorg. Mater. 2000, 2, 455.
8322. (66) Bigi, A.; Boanini, E.; Walsh, D.; Mann, S. Angew. Chem., Int. Ed.
(25) Malkaj, P.; Chrissanthopoulos, A.; Dalas, E. J. Cryst. Growth 2002, 2002, 41, 2163.
242, 232. (67) Busch, S.; Schwarz, U.; Kniep, R. Chem. Mater. 2001, 13, 3260.
(26) Pai, R. K.; Hild, S.; Ziegler, A.; Marti, O. Langmuir 2004, 20, 3123. (68) Künther. J.; Seshadri, R.; Nelles, G.; Assenmacher, W.; Butt, H.-J.;
(27) Naka, K.; Tanaka, Y.; Chujo, Y. Langmuir 2002, 18, 3655. Mader, W.; Tremel, W. Chem. Mater. 1999, 11, 1317.
(28) Naka, K.; Chujo, Y. C. R. Chimie 2003, 6, 1193. (69) Rautaray, D.; Sinha, K.; Shankar, S. S.; Adyanthaya, S. D.; Sastry,
(29) Naka, K. Top. Curr. Chem. 2003, 228, 141. M. Chem. Mater. 2004, 16, 1356.
(30) Dimova, R.; Lipowsky, R.; Mastai, Y.; Antonietti, M. Langmuir 2003, (70) Yu, S. H.; Cölfen, H.; Hartmann, J.; Antonietti, M. AdV. Funct. Mater.
19, 6097. 2002, 12, 541.
(31) Sedlak, M.; Cölfen, H. Macromol. Chem. Phys. 2001, 202, 587. (71) Keum, D.-K.; Naka, K.; Chujo, Y. Chem. Lett. 2004, 33, 310.
(32) Kaluzynscki, K.; Pretula, J.; Lapienis, G.; Basko, M.; Bartzcak, Z.; (72) Dalas, E.; Klepetsanis; Koutsoukos, P. G. J. Coll. Int. Sci. 2000, 224,
Dworak, A. J. Polym. Sci. A, Polym. Chem. 2001, 117, 200. 56a.
(33) Bolze, J.; Pontoni, D.; Ballauff, M.; Narayanan, T.; Cölfen, H. J.
Coll. Int. Sci. 2004, 277, 84. CG050436W