Calcium Carbonate Crystallization in the Presence of Biopolymers

Calcium Carbonate Crystallization in the Presence of BiopolymersMichael F. Butler,* Nicole Glaser, Antony C. Weaver, Mark Kirkland, and Mary Heppenstall-ButlerUnileVer R&D Colworth, Sharnbrook, Bedfordshire, MK44 1LQ, U.K. ReceiVed August 23, 2005; ReVised Manuscript ReceiVed NoVember 7, 2005

CRYSTAL GROWTH & DESIGN2006 VOL. 6, NO. 3 781-794

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, -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 -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 -carrageenan concentration decreased the solution supersaturation and hence decreased the propensity for calcite formation. Introduction The ability to control crystallization is a critical requirement in the synthesis of many industrially important materials. Selected examples of these are pharmaceutical molecules, triglycerides, fatty acids, and inorganic materials such as calcium carbonate. Scientifically, calcium carbonate is of enormous interest in the context of crystal control because it is produced by a wide variety of biological organisms that exhibit exquisite control over the polymorph, location of nucleation, crystal size and shape, crystallographic orientation, composition, stability, and hierarchical assembly of the crystals. Calcium carbonate can exist in several different polymorphs that are, in order of increasing solubility: calcite, aragonite, vaterite, calcium carbonate monohydrate, and calcium carbonate hexahydrate. Although aragonite and calcite are the most commonly found polymorphs in natural systems, vaterite is also present in some systems, and all of the polymorphs may be obtained in the laboratory. In the presence of water, all of the polymorphs eventually transform into the most stable form, calcite. They may, however, be stabilized in the presence of additives. Amorphous calcium carbonate also exists but has only been observed in a stabilized form as a biomineral.1 Many studies have investigated the influence of additives on the crystallization of calcium carbonate.2,3 In the simplest case, the addition of magnesium is well known to lead to the formation of spicular aragonite crystals rather than rhombohedral calcite. Simple organic acids such as citric and maleic acid have been shown to strongly influence crystal morphology, especially in the presence of magnesium ions,4 and studies have shown that acidic peptides, in particular, aspartic acid, can alter the chiral morphology of calcite crystals by binding to specific step edges at the crystal surface.5 In general, aspartic acid has a large effect on calcium carbonate crystal growth, causing either calcite or vaterite to form depending on the acid concentration.6-8 Other amino acids, such as alanine, lysine, and glycine, have similar effects.9 More complicated sequences of amino acids, inspired* To whom correspondence should be addressed. E-mail: [email protected]. Phone: +44 (0)1234 222958. Fax: +44 (0)1234 248010.

by those present in biomineral forming systems, have also been shown to strongly influence calcium carbonate growth. In some cases, calcite is formed, but with morphologies other than rhombohedra, whereas in others different forms of calcite, such as hollow vaterite spheres, have been observed.10 In the more complex case, polymers have been used to control the nucleation and growth of calcium carbonate crystals. In many cases, the field of biomineralization has motivated these studies. Polymeric systems containing aspartic acid and glutamic acid have received much attention because they are components present in several natural shell-forming systems. Different polymorphs of calcium carbonate were obtained depending on the amino acid sequence and polypeptide conformation.11 Atomic force microscope (AFM) studies, combined with molecular modeling, suggest that specific binding of natural polypeptides, extracted from shells, to particular calcite crystal faces, is responsible for the modification in morphology of calcite crystals in these cases.12 In at least one case,13 a peptide was designed to interact with a specific crystal face, and the resulting crystals had the predicted morphology, demonstrating the feasibility of this explanation for the natural polypeptides. Both lysozyme14 and collagen15 have been shown to systematically influence the calcium carbonate morphology with increasing polypeptide concentration. In the case of lysozyme, the calcite crystal habit was modified as the growth of different crystal faces was sequentially inhibited. In the case of collagen, a transition was observed with increasing collagen concentration, from rhombohedral calcite crystals to spherulitic crystal aggregates, via multiple layer crystals, that was explained by preferential adsorption of collagen. In the presence of soluble proteins extracted from coralline algae, amorphous calcium carbonate has been stabilized.16 Synthetic polymers have also been used to alter calcium carbonate growth. In some cases, simple anionic polymers, such as poly(acrylic acid), poly(acrylonitrile), and poly(styrene 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 confine crystallization to the film surface by preventing bulk 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 crystallization inhibition via the number and strength of the polymer-crystal interactions.21 In other cases, specific polymorphs, including calcite, vaterite, and calcium carbonate hexahydrate, have been obtained in the presence of synthetic polymers containing chemical groups, such as carboxylate, amide, and sulfate, that are known to interact with inorganic salts.17,22-26 More complex polymer architectures have also been studied. For instance, anionic carboxylate terminated poly(amidoamine) dendrimers27-29 have been found to stabilize spherical vaterite particles. Double hydrophilic block copolymers containing a weakly hydrophilic block and a polyelectrolyte block that interacts with calcium carbonate,30 which are a simplified model system for active polypeptides in biomineralization with acidic domains, have also been found to stabilize vaterite as well as amorphous calcium carbonate for short times.23,31-34 They lead to the formation of well-defined microparticles with a shape that predominantly depends on the chemical functionality of the polyelectrolyte block. As well as showing crystallization inhibition effects,35 a wide variety of crystal morphologies have been obtained in the course of these studies, from the conventional calcite rhombohedra, aragonite spicules, and vaterite spheres to hollow shells23 (sometimes formed from aggregates of spheres36,37), ellipsoids,38 dumbbells,23,26,39,40 plates,26 stacks of rhombohedra,36 and disks.36 Helices41 and spheres42 have been observed in the presence of synthetic block copolypeptides containing serine and aspartate residues, respectively. Interestingly, in the former case the chirality of the copolypeptide was used to direct the sense of the helix. Many polysaccharides are natural block copolymers and consist of two or more different glycosidic monomer units. Alginates, which are extracted from seaweed, are composed of mannuronic and guluronic acid residues, for example. Despite their importance in the field of biomineralization, where anionic polysaccharides provide the template for the intricate hierarchical assemblies of calcium carbonate crystals in coccoliths,43,44 relatively little work has been done on the systematic study of the influence of simple polysaccharides on calcium carbonate crystallization. Sodium alginate and carboxymethyl inulin have been found to inhibit the crystallization of calcium carbonate45,46 by interacting directly with the surface of the growing crystals. Although sodium alginate did not alter the standard rhombohedral habit of calcite, carboxyethyl inulin caused spherical vaterite particles to grow in addition to rhombohedral calcite crystals. For carboxymethyl inulin, the inhibition efficacy increased with the carboxylate content of the polymer. In another study, different calcium carbonate polymorphs were obtained in the presence of different polysaccharides.47 Calcite and vaterite were formed in the presence of beta cyclodextrin and soluble starch, respectively. The presence of the different polymorphs was explained by the differences in geometrical matching and stereochemical complementarity between the calcium ions and the hydroxyl groups on the biopolymers. The importance of stereochemistry was also demonstrated in a study of the influence of poly(alginic acid) and poly(galacturonic acid) on calcium carbonate growth. Despite containing similar numbers of carboxylate groups, poly(alginic acid) was found to be more effective than poly(galacturonic acid) at influencing calcium carbonate growth morphologies at equal concentrations.44 Some studies on sulfated polysaccharides have been reported47-49 that were inspired by the presence of sulfated

proteoglycans in some biomineralizing systems. Although -carrageenan had a relatively small influence on calcium carbonate growth,47 heparin was shown to form either aragonite crystals48 or helicoidally arranged calcite aggregates49 depending on the experimental conditions. Hyaluronic acid caused the formation of columnar stacks of rhombohedral calcite crystals, whereas individual cuboctahedral calcite crystals formed in the presence of keratan sulfate.49 No systematic attempt was made to relate the molecular structure of the sulfated polysaccharide to the crystal habit in these studies. More detailed studies have been performed on the formation of thin films of calcium carbonate on biopolymer substrates, in an attempt to mimic the formation of the nacreous layers of calcium carbonate in mollusc shells that grow on biopolymer templates. Several studies have demonstrated the formation of calcite films on insoluble polysaccharide (chitosan) substrates in the presence of poly(acrylic) or poly(aspartic acid).20,50-54 In these cases, the interaction of the polyanion with the chitosan surface provided nucleation sites for the initiation of calcium carbonate formation, while the presence of free polyanion in solution prevented bulk nucleation and growth and confined the crystal growth to the film on the insoluble substrate. Calcite films have been grown in a similar way on simple glass substrates, in the presence of poly(acrylic), poly(aspartic), and poly(glutamic) acid.55 In the presence of magnesium, aragonite films were grown on chitosan substrates.56 In the present study, the crystallization of calcium carbonate in the presence of a range of biopolymers that can be considered to be natural double-hydrophilic block copolymers was investigated. They were chosen so that the effects of calcium binding, number, and type of side group (i.e., carboxylate and sulfate) on the resulting calcium carbonate morphology and unit cell could be studied. As such, this report presents the first investigation of the influence of a range of readily available food biopolymers on calcium carbonate crystallization. The biopolymers studied, shown in Figure 1, consisted of two types. The first type, which were biopolymers that did not gel in the presence of calcium, were gellan, xanthan, and -carrageenan. Gellan and xanthan are acidic biopolymers that contain residues with carboxylic acid groups. -Carrageenan is a sulfated polysaccharide. The second type of biopolymer studied were those that did chelate calcium to form gels. These were sodium alginate and pectin, both of which contain residues with carboxylic acid groups. Experimental SectionMaterials and Sample Preparation, Control Samples. Calcium carbonate was formed by mixing and continually stirring 10 mL of calcium chloride solution and 10 mL of sodium bicarbonate solution (supplied by Sigma Chemical Company), with the concentrations shown in Tables 1 and 2. In the first and second set of experiments, where the ratio of calcium chloride/sodium bicarbonate concentration was 1:2 and 1:1, respectively, the mixture was raised to pH 10.5 by the dropwise addition of 5 M sodium hydroxide, potassium hydroxide, or ammonium hydroxide solutions (supplied by Sigma Chemical Company). To test 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 M:0.01 M, using sodium hydroxide to achieve pH values between 8 and 12 inclusive, in increments of 0.5 pH units. Measurements of pH were performed using a pH meter (HI8424, from Hanna Instruments). Materials and Sample Preparation, Samples in the Presence of Biopolymers. The biopolymers used in this study were gellan (gel F, supplied by Kelco), xanthan (cold water soluble Keltrol RD, supplied by Kelco), -carrageenan (Genugel X0909, supplied by Kelco), sodium alginate (Manugel DMB, supplied by Kelco), LM pectin (LM12, 35% de-esterified, supplied by Kelco), and high-methoxy (HM) pectin (65% 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) -carrageenan, (d) sodium alginate, and (e) pectin.Table 1. Solution Concentrations Used in Experiment 1a mix 1 cCaCl2 (mol/ L) cNaHCO3 (mol/ L)a

mix 2 0.025 0.050

mix 3 0.05 0.10

mix 4 0.075 0.150

mix 5 0.10 0.20

0.01 0.02

The pH was raised to 10.5. Table 2. Solution Concentrations Used in Experiment 2a mix 6 mix 7 0.025 0.025 mix 8 0.05 0.05 mix 9 0.075 0.075 mix 10 0.10 0.10

cCaCl2 (mol/ L) cNaHCO3 (mol/ L)a

0.01 0.01

The pH was raised to 10.5.

Stock solutions of 4% (by weight) were made from sodium alginate, -carrageenan, LM and HM pectin by stirring the dry biopolymer powder into de-ionized water. The sodium alginate and -carrageenan solutions were heated to 80 C for 30 min to ensure complete dissolution, whereas the pectin solutions were stirred at room temperature until completely dissolved. Stock solutions of 0.2% (by weight) of gellan and xanthan were made by adding the dry biopolymer powder to de-ionized water and stirring at 80 C for 30 min to ensure complete dissolution. These stock solutions were diluted to the required final concentrations for the calcium carbonate growth experiments. The zero shear rate viscosity of the different biopolymer solutions was measured for a variety of concentrations, using a dynamic stress rheometer (Rheometrics DSR200) equipped with a Couette geometry,

to determine the critical entanglement concentration that separates the dilute concentration regime, where the polymers can be regarded as separate entities, and the semidilute regime, where the molecules are entangled. The critical entanglement concentrations (%,w/v) were sodium alginate, 1.75; -carrageenan, 1.75; LM pectin, 1.83; HM pectin, 0.75; gellan, 1.12; xanthan, 0.005. Calcium carbonate was formed by mixing 10 mL of calcium chloride (0.01 M) solution with 10 mL of sodium bicarbonate (0.02 M) solution containing the biopolymer at the required concentration, made from the stock solution. The pH of the mixture was increased to 10.5 using 5 M sodium hydroxide solution to trigger the precipitation of calcium carbonate crystals. A biopolymer concentration of 0.2% (by weight) was used for all of the samples because this was below the critical entanglement concentration for all of the biopolymers except for xanthan. pH Measurements During Crystallization. A benchtop pH meter (Hanna model 302) was used to continually measure the pH in mixtures containing different biopolymers. 30 mL of 0.01 M calcium chloride solution was added to 30 mL of biopolymer solution containing 0.02 M sodium bicarbonate in a glass beaker. The pH was increased to 10.5 by the dropwise addition of 5 M sodium hydroxide solution, and its value was subsequently measured at intervals of 2 min. The mixture was continuously stirred during the experiment. Optical Microscopy. Transmission optical microscopy (Leitz Diaplan, set up for Kohler illumination) was used to determine the crystal habit of the calcium carbonate crystals, which had precipitated and aged 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. The crystals were obtained by centrifuging the suspension in which they formed and washing with 0.1 M sodium hydroxide solution (to prevent dissolution during washing) three times. A drop of the final, cleaned suspension was placed on a microscope slide beneath a standard glass cover-slip and observed under bright-field conditions with and without crossed polarizers. Electron Microscopy. One drop of suspension of each sample, washed (with 0.1 M sodium hydroxide) and centrifuged three times, was placed onto a carbon-coated copper grid. The drop was allowed to dry in air, and the residual crystals were sputter coated with a 10nm layer of gold/palladium metal. A JEOL1200EX transmission electron microscope equipped with an ASID10 scanning attachment, operating at an accelerating voltage of 20 kV, was used to obtain highmagnification images of the crystal morphologies. The images were recorded using software (INCA) supplied by Oxford Instruments. Close-up images of the crystal surfaces were imaged using a JEOL 6301 scanning electron microscope. The samples were coated prior to examination with a gold/palladium mixture in an Oxford CP2000 preparation chamber. Turbidity. In-situ crystallization experiments were performed using an ultra-violet/visible wavelength (UV/Vis) spectrophotometer (PerkinElmer Lambda 40) to measure the absorption of light in a crystallizing suspension of calcium carbonate at 500 nm wavelength, every 20 s. The onset of crystallization was detected via an increase in turbidity, as measured by increasing absorbance of the light. Five milliliters of 0.01 M CaCl2 solution was added to 5 mL of biopolymer solution containing 0.02 M NaHCO3 and 0.2% (w/v) biopolymer while being stirred using a magnetic follower. The pH was increased to 10.5 using 1 M NaOH. Immediately afterward, the mixture was transferred into the UV/Vis spectrophotometer. The time taken between setting the pH to 10.5 and the turbidity measurements starting was measured and accounted for during analysis of the results. X-ray Diffraction. X-ray diffraction was performed at beamline 16.1 at the Synchrotron Radiation Source, Daresbury. One-dimensional diffraction patterns were obtained from samples consisting of a sedimented suspension of calcium carbonate crystals with a thickness of approximately 1 mm. The diffraction patterns were obtained on a curved Inel multiwire detector. An exposure time of 30 s was used to collect the diffraction patterns.

Results Control Samples. Turbidity, Optical Microscopy, and Electron Microscopy. For all of the control samples, crystallization occurred immediately upon reaching the final pH and was measured by a sudden increase in sample turbidity using the UV/Vis spectrophotometer (see Figure 2 for a sample measured at pH 10.5). The turbidity rapidly reached a maximum value and then decreased again, owing to sedimentation of the crystals in the cuvette. For all of the experimental conditions, the control samples yielded rhombohedral crystals. Sometimes single crystals were obtained, shown in the optical micrograph in Figure 3a for a sample at pH 10.5, but in most cases aggregates of rhombohedra were obtained, shown in the optical micrograph in Figure 3b,

also from a sample at pH 10.5. Scanning electron micrographs, shown in Figure 4for samples at pH 10.5, revealed that the aggregates were formed from several crystals that had intergrown. pH Measurement. The pH variation after addition of alkali to the target pH was different for the samples with 1:2 and 1:1 calcium chloride/sodium bicarbonate ratios. The 1:2 samples remained roughly at the target pH value that was attained after addition of alkali, shown in Figure 5a. The 1:1 samples experienced a variation in pH with time, shown in Figure 5b, which depended on the target pH value. Above approximately pH 11, the pH remained constant. Below pH 11, a rapid decrease was observed that reached a plateau around pH 7.5. Because the pH was found to be constant after addition of alkali for the 1:2 samples, subsequent experiments in the presence of biopolymer were made with a 1:2 ratio of calcium chloride to sodium bicarbonate. Crystallization in the Presence of Biopolymers. Turbidity. For all of the biopolymers, there was an induction time before the onset of crystallization was measured by an increase in turbidity of the mixture. Figure 6 shows a typical plot of the change in turbidity with time of a sample containing 0.2% (w/ v) biopolymer (-carrageenan) at pH 10.5, measured using the UV/Vis spectrophotometer. The value of the induction time varied within each sample, although the average induction time, shown in Table 3, depended on the type of biopolymer for samples with the same polymer concentration and final pH value. It also depended on the biopolymer concentration, shown in Figure 7 for LM pectin and -carrageenan. The induction time decreased with increasing pectin concentration, although the entire sample formed a gel at concentrations above 1%, and no crystallization was observed at and above this concentration. In contrast, the reverse trend was observed for -carrageenan that showed a positive correlation between induction time and biopolymer concentration. Optical and Electron Microscopy. Figure 8 shows optical micrographs of samples, taken in bright-field mode and between crossed polarizers, containing different biopolymers at 0.2% (w/ v) and at pH 10.5. The bright-field images showed that in all cases the majority of the crystals were in the form of aggregates. Single rhombohedra were extremely rare. Crystals formed in the presence of gellan and xanthan were of similar size and were anisotropic, elongated aggregates. Close inspection of the polarizing optical micrographs suggested that these aggregates were stacks of rhombohedra. In contrast, dumb-bell-shaped aggregates were observed in the presence of LM and HM pectin, and a mixture of spherical and ellipsoidal aggregates was observed in the presence of -carrageenan and sodium alginate. The crystalline aggregates formed in the presence of -carrageenan appeared to be the smallest, whereas similarly sized aggregates were formed in the presence of the different pectins 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) -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 bicarbonateTable 3. Induction Times before the Onset of Calcium Carbonate Crystallization in the Presence of Different Biopolymer Additives (0.2% w/v) biopolymer gellan xanthan -carrageenan sodium alginate HM pectin LM pectin induction time (minutes) 1.8 ( 0.5 2.0 ( 0.7 3.0 ( 1.1 82.0 ( 46.6 5.5 ( 1.0 46.0 ( 6.9

Figure 7. Variation of induction time before the onset of crystallization with biopolymer concentration for LM pectin and -carrageenan.

however. Inspection of the polarizing optical micrographs revealed that for the -carrageenan, pectin, and sodium alginate samples Maltese cross patterns were formed, although their appearance did depend on the position of the crystalline aggregate relative to the focal plane of the microscope. Scanning transmission electron micrographs for these crystals are shown in Figure 9. The higher magnification attainable using electron microscopy confirmed the observations made using optical microscopy for crystals grown in the presence of gellan and xanthan. In these cases, the crystal aggregates were indeed

stacks of rhombohedra. The stacks were better defined for the crystals in the xanthan sample. For the pectin, alginate, and -carrageenan samples the ellipsoidal and dumb-bell-shaped aggregates appeared to be formed from interconnected and radially arranged crystals, which was most apparent for the LM pectin sample. The size of the crystals in the sodium alginate sample was greater than for the pectin and -carrageenan samples, which gave the aggregates formed in the presence of sodium alginate a slightly less well ordered appearance. In some cases, shown in Figure 10, for the samples containing pectin and -carrageenan, hollow crystalline shells were observed. The number of these was increased when powdered samples of these crystals were crushed in a press. Figure 11 shows scanning electron micrographs of the surface of the crystals at higher magnification, and therefore in greater 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 -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 concentrations, although the crystal size was similar. For the -carrageenan samples, the crystals appeared to become more dumbbell 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 -carrageenan are shown in Figure 14. Similar crystal morphologies were obtained for all of the concentrations used. Figure 15 shows scanning transmission electron micrographs 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 biopolymers, the 1:2 ratio of calcium chloride/sodium bicarbonate was used because, at pH values greater than 10.5, only carbonate ions are formed, as described in the following reaction. In this case, the hydrogen carbonate ions were in excess and acted as a buffer in the presence of sodium hydroxide to maintain a constant pH value.

Figure 8. Transmission optical micrographs (left-hand side) and corresponding polarizing optical micrographs (right-hand side) of calcium carbonate crystals grown in the presence of 0.2% biopolymer: (a) gellan, (b) xanthan, (c) -carrageenan, (d) sodium alginate, (e) LM pectin, and (f) HM pectin.

surfaces of the crystals grown in the presence of xanthan appeared to be roughened, or composed of many nanocrystallites with sizes of approximately 100 nm, on all of the crystal faces. Crystals grown in the presence of -carrageenan and pectin appeared to be roughened on some of the faces, but other faces were smooth. All of the faces appeared to be smooth for the crystals grown in the presence of sodium alginate. Figures 12 and 13 show bright-field and polarizing optical micrographs of crystals formed in the presence of different

H2CO3 + CaCl2 + 2NaHCO3 w CaCO3 + 2NaCl + HCO3 + H CO32- + 2H+Below pH 10.5, hydrogen carbonate is a weak acid, and it can dissociate and liberate hydrogen ions, leading to a decrease in pH. For the 1:1 ratio of calcium chloride/sodium bicarbonate, 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) -carrageenan, (d) sodium alginate, (e) LM pectin, and (f) HM pectin.

to carbonate ions. The system therefore had no buffering capacity, and the excess hydrogen ions caused the decrease in pH that was measured at all of the pH values at which crystallization was initiated. Since it was desirable to conduct experiments in which the formation of calcium carbonate was as regulated as possible, all of the experiments with biopolymers were performed at pH 10.5, with the 1:2 ratio of calcium chloride/sodium bicarbonate. The induction time that was measured prior to the onset of crystallization in the presence of the biopolymers was indicative of the crystallization inhibition that has been measured in the presence of many polymers, including biopolymers such as

alginate,44,45 -carrageenan,48 and poly(galacturonic acid),44 which is similar to pectin. The current results, which showed that sodium alginate was more effective at inhibiting calcium carbonate crystallization than either pectin or -carrageenan, are consistent with the previous studies on alginate, poly(galacturonic acid) and -carrageenan.44,48 In those studies, it was argued that the main factors that contributed to the crystallization inhibition were charge strength, amount, and polymer conformation. The final factor was proposed because some polyanions, such as chondroitin sulfate,57 were investigated that, despite possessing many anionic 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) -carrageenan and (b) LM pectin.

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

tion. It was concluded that additional factors, such as polymer conformation that may lead to a reduced availability or dissipation of charge, were important. The first factor, strength of charge, explains why -carrageenan had less effect on calcium carbonate crystallization than did HM pectin, despite possessing similar numbers of anionic groups. The lower partial charge on the oxygen atoms of the anionic sulfate groups in -carrageenan compared to the anionic carboxylate groups in pectin will result in a weaker interaction with the calcium ions at the surface of calcium carbonate crystals and therefore a weaker influence on calcium carbonate crystallization. The second factor, number of charges, explains the much smaller induction times for gellan and xanthan compared to pectin and alginate, HM pectin compared to LM pectin, and LM pectin compared to -carrageenan. All of these polymers possess anionic carboxylate groups, but xanthan and gellan contain far fewer carboxylate groups than pectin or alginate. Similarly, HM pectin contains fewer carboxylate groups than LM pectin. The smaller number of charges therefore leads to fewer possible interactions between the polyanion and the surface of the calcium carbonate crystal and weaker interaction overall. The third factor, conformation and availability of charge,

is unlikely to be the reason for the difference in induction time between LM pectin and sodium alginate. Pectin has been reported to be generally more flexible than sodium alginate.58 It might therefore be expected that the carboxylate groups in pectin would be more available for binding to calcium, leading to more efficient calcium sequestration by pectin and hence a longer induction time. However, the reverse was actually observed. It should be noted that the conformation of the polymers used in the current study is not known. In conclusion, the difference between the induction times of LM pectin and sodium alginate remains unknown. The observation that, in the systems containing LM pectin, the pH began to drop prior to the formation of calcium carbonate detected by turbidity, suggests that events began to occur that influenced crystallization prior to the nucleation event itself. One explanation is that the drop in pH demonstrates the interaction of LM pectin with incipient calcium carbonate nuclei, hindering their formation and delaying crystallization until later times. A previous study of calcium carbonate crystallization in the presence of sodium alginate45 has shown the large influence that this polyanion can have on calcium carbonate crystal growth 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 mechanism,59 which has been observed using AFM in the presence of some small molecular weight additives,5,60,61 to one controlled by surface nucleation. The crystal morphologies were all related to the (104) calcite rhombohedra obtained in the control samples, as expected from the XRD data that showed that calcite was the only crystal polymorph present in all cases. The effect of the biopolymers was apparent, however, in the overall morphology of the crystals. Most obvious was the distinction between the different types of biopolymer: those that did not form gels in the presence of calcium ions and those that did. The former biopolymers, gellan and xanthan, formed aggregates of crystals that could be described as roughly stack-like, whereas the latter biolymers, pectin (LM and HM) and alginate, formed aggregates that could be described as rosette-like. Furthermore, the rosette-like aggregates appeared to be hollow shells. -Carrageenan, that forms a gel in the presence of sodium but not calcium ions, formed a morphology that was intermediate between stack- and rosette-like. Stack-like morphologies have been obtained in the presence of hyaluronic acid,49 a carboxylated biopolymer, a mixed solution of a block-copolymer (poly(ethylene glycol)-block-poly(methacrylic acid), PEG-b-PMAA) and a cationic surfactant (cetyltrimethylammonium bromide, CTAB),36 and a watersoluble terpolymer (poly(acrylamide-co-2-acrylamido-2-methyl1-propane sodium sulfonate-co-n-vinyl-pyrrolidone)26 that contained several functional groups known to interact with calcium ions. Although no explanation was given for the formation of the stacks in the case of hyaluronic acid, other than the acid

induced the formation of a monocrystalline aggregate, or for the PEG-b-PMAA CTAB mixture, for the terpolymer it was suggested that the molecule existed in an extended conformation owing to the large number of ionizable sulfate units on the chain. Because the terpolymer also possessed units (>CdO, >SdO, and >NdH) that are known to interact with calcium ions, it was suggested that they nucleated the formation of calcite crystals along the chain, leading to the formation of stacks. A similar explanation may favor the formation of stacks for hyaluronic acid and PEG-b-PMAA CTAB, as well as for xanthan and gellan in the current study. All of these polymers contain ionizable carboxylate groups that can become charged at high pH values as well as interacting with calcium ions. In addition, xanthan and gellan are rather stiff molecules by virtue of being polysaccharides. It is therefore plausible to suggest that, in the high pH conditions used in the present study (>pH10.5), the gellan and xanthan molecules became ionized, extended, and also nucleated calcite crystallization via the interaction between the carboxylate groups on the polymer and the calcium ions in solution, thereby forming stacks of calcite rhombohedra. That there is an interaction between xanthan and calcium carbonate is suggested by the high-resolution SEM image of the calcite crystal surface grown in the presence of xanthan, which shows a high degree of surface roughness consistent with the presence of bound impurities, i.e., xanthan molecules, at the interface. Rosette-like aggregates of calcium carbonate crystals have been observed previously in the presence of biopolymers, such

790 Crystal Growth & Design, Vol. 6, No. 3, 2006

Butler et al.

Figure 13. 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 -carrageenan: (a) 0.2%, (b) 0.3%, (c) 1.0%, (d) 1.8%, (e) 2.0%, (f) 2.5%, and (g) 3.0%.

as heparin,47,49 tobacco mosaic virus,62 gelatin (with added magnesium),63-65 and collagen.15 In the former two cases, it

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 magnesium 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-mercaptophenol colloidal gold seeds leads to the formation of rosettelike 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 morphology 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-bPMAA,23,37,70 PEG-b-PMAA-aspartate (Asp),23 poly(styreneblock-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 copolymer 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 spheres.37 The block copolymer was then proposed to template calcite nucleation and growth on the surface of the core calcium carbonate particle. The outer layer of facetted calcite crystals then grew at the expense of the core, which dissolved to provide material for crystal growth. In a similar manner, surface functionalized gold nanoparticles adsorbed onto vaterite spheres 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) -carrageenan, 0.2%, and (d) -carrageenan, 2.5%.

leading to rosette-shaped crystal aggregates.71 Surface functionalized dendrimer molecules, containing moieties that interact with calcium at the outer surface of the dendrimer, also lead to a spherical overgrowth of calcium carbonate crystals.27,28 Pectin and alginate are biopolymers that interact strongly with calcium ions, to the extent that, at the right concentration, a gel is formed whereby the calcium ions form physical cross-links between different biopolymer chains. It is therefore likely that the addition of calcium chloride to the solution containing pectin or alginate and sodium bicarbonate led to the formation of small regions of gelled biopolymer that contained bound calcium ions that could act as nuclei for calcium carbonate crystal growth. In the same manner as the double hydrophilic block copolymers or colloidal gold particles templated the growth of calcite or aragonite on a particle core, leading to the formation of a crystalline shell, it is proposed that in the current study the pectin or alginate that formed the core also templated the nucleation of the calcite overgrowth. The common feature in all of the systems is that a species is present that contains chemical groups, such as carboxylate groups, that interact strongly with calcium. The difference between the systems is that in the previous studies the crystalline shell grew on a calcium carbonate core that subsequently dissolved, whereas in the present study the interior of the rosette-like aggregate is proposed to contained a biopolymer core. Depending on the amount of free calcium ions present within the biopolymer gel microparticle, it is therefore possible to form calcite within the gel as well as at the surface, which explains why, in some cases, separate calcite aggregates were observed that fitted perfectly within the rosette-like shell. Presumably, crystallization within pectin or alginate would be less likely to occur than in the solution near the surface of the

gel because there will be less free calcium in the interior of the gel that is available to form calcite crystals. An insufficient number of open shells were observed to test this hypothesis, however. Future studies will concentrate on developing an understanding of the hollow calcite shells that are proposed to form in the presence of pectin or alginate. Interestingly, the rosette-like aggregates formed in the presence of LM pectin appeared more ordered than those formed in the presence of HM pectin and alginate, shown both by the appearance of the rosettes in the STEM images and by the presence of Maltese crosses in the rosettes observed between crossed polarizers in the optical microscope. The Maltese cross patterns showed that, for LM pectin, radial growth of the calcite rhombohedra occurred with the crystals all possessing a uniform orientation, i.e., the growth occurred in a directed manner. This marked degree of orientation implies that there was a relation between the underlying LM pectin template and the overgrowth of calcite crystals that formed the rosette-like aggregate. The possibility of such a direct interaction is suggested by studies of calcium carbonate growth on biopolymer films, such as chitosan in the presence of poly(acrylic acid)20,50-55 and cellulose.54,72 In the case of calcite grown on poly(acrylic acid) bound to insoluble chitosan films, a direct match between the separation of the calcium ions in the calcite crystal on the (001), (110), and (104) faces and the spacing of the carboxylate groups on the polymer was used to explain the crystal morphologies obtained.50,51,55 Although detailed crystallographic data are not available from the present study, future investigations on templated growth on biopolymer films will explore the nature of any potential crystallographic templating effect in 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.

Figure 16. Change in pH with time during crystallization in solutions containing LM pectin.

Figure 17. X-ray diffraction patterns of the calcium carbonate crystals grown in the presence of different biopolymers, indexed according to calcite.

It was also observed that, in the present study, the rosettelike aggregates often formed in the shape of dumb-bells. This effect was most noticeable for LM pectin but could also be seen

to some extent for all of the other systems containing rosettelike aggregates, including -carrageenan. In the case of fluoroapatite 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 crystalcrystal 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 aggregates in the presence of double hydrophilic block copolymers,39 enough similarities were observed with the fluoroapatite aggregates to suggest that a similar spherulitic growth mechanism could explain the calcium carbonate dumb-bells in the later stages of aggregate growth. It is likely, therefore, that the dumbbells 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 -carrageenan, an intermediate morphology was obtained that lay between the stack-like morphology observed in the presence of nongelling biopolymers and the rosette-like morphology obtained in the presence of calcium-gelling biopolymers. However, the aggregates formed in the presence of -carrageenan also appeared to be hollow, in at least some cases. 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 -carrageenan is known to form gels in the presence of sodium ions. Sodium ions were present in the current study since the reaction to form calcium carbonate was performed using sodium bicarbonate. Therefore, it is likely that a weak gel of -carrageenan formed. Furthermore, it is known that the cross-links in -carrageenan gels are formed from two chains interacting in a double helix. The presence of a gel therefore provides the template for the formation of a hollow shell of calcite crystals in the same rosette form observed in the presence of pectin and alginate, whereas the presence of extended double-helical regions provides the reason for the extended nucleation in the stack form observed in xanthan and gellan. From the above discussion, it is possible to explain the effect of pH and biopolymer concentration on the crystal morphology. First, the overall similarity in aggregate morphology for the crystals grown in the presence of LM pectin at different pH values is to be expected, since in all cases crystal growth was controlled only by the presence of the nucleating LM pectin gel template and the presence of calcium and carbonate ions in solution, which will be similar at all of the pH values studied. The less-ordered aggregates observed at higher pH values reflect the faster growth rates caused by the higher supersaturation of carbonate ions in solution at higher pH. Second, the overall similarity of the morphologies observed at different biopolymer concentrations can also be explained for similar reasons. That the biopolymers studied interact in solution to a lesser extent than the double hydrophilic block copolymers reported in the literature is shown by the presence of the rhombohedral (104) calcite form in all of the biopolymer systems in the present study, over the range of concentrations used, in the present study. For the block copolymer systems, differences in crystal habit were also observed over particular polymer concentration ranges, indicating a strong interaction of those polymers with particular crystallographic planes that markedly altered their growth rates.23,36 The relationship between crystallization kinetics and the concentration of LM pectin and -carrageenan is less straightforward to explain, however. The nucleating ability of polymers is correlated with the ability of the polymer to bind metal ions that intiates the formation of subcritical nuclei that grow to the critical size required for crystal growth.72 From standard nucleation theory, the induction time, , is related to the solution supersaturation, , by the following equation:

0.4%, however, a decrease in induction time was measured that cannot be explained by the effect of the polymer on supersaturation. Since LM pectin formed the most ordered radially oriented aggregates of crystals and was therefore proposed to act as a direct template for calcite growth, it is possible that the decrease in induction time is indicative of a decrease in surface energy that is increasingly favorable for nucleation despite the reduction in solution supersaturation. Such an effect could occur if, at the higher LM pectin concentrations, the separation between the templating carboxylate groups became increasingly matched to the lattice spacing between calcium ions on the templated crystal plane. For -carrageenan, that does not gel in the presence of calcium ions but is still expected to interact via the sulfate groups, the overall increase in induction time is expected owing to an increase in the amount of bound calcium with increasing polymer concentration. The effect of supersaturation was therefore dominant overall. Interestingly, however, a peak in the induction time was measured, superimposed on the overall increase, at a -carrageenan concentration around 2%, that coincided with the observation of fewer nuclei and larger crystals using optical microscopy. The subsequent reduction in induction time at the -carrageenan concentration of 2.5% indicated that a compensating effect was present at this concentration that served to increase the likelihood of crystallization. As for the LM pectin case, it is possible that at this concentration the average spacing of the sulfate groups provided a better match between the polymer and the incipient calcite crystals, thereby reducing the surface energy of the calcite nuclei and promoting nucleation. Further experiments will be performed on twodimensional biopolymer surfaces to investigate these phenomena. Conclusions The influence on calcium carbonate crystallization of a series of food biopolymers, which contain carboxylic acid or sulfate functional groups, was studied using a variety of techniques, including pH and turbidity measurements, optical microscopy, and scanning electron microscopy. The biopolymers chosen were gellan, xanthan, LM pectin, HM pectin, and sodium alginate, which contain carboxylate groups, and -carrageenan, which contains sulfate groups. In control samples containing no biopolymer, single calcite (104) rhombohedra were formed. In the presence of biopolymers, (104) rhombohedra were formed as aggregates that were either stack-like or rosette-like. Stacks were formed in the presence of nongelling biopolymers, which were nucleated by the carboxylate groups on extended xanthan or gellan chains. Rosettes were formed in the presence of calcium-gelling biopolymers and were proposed to form by the nucleation of calcite on a gelled microparticle template. LM pectin was particularly effective at directing the growth of calcite rosettes and led to aggregates of radially aligned crystals. Evidence was found that the rosettes were hollow. The influence of biopolymer concentration on calcite crystallization was studied for LM pectin and -carrageenan. In the former case, an increasingly favorable influence of the pectin molecules on the surface energy of calcite nuclei was believed 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 -carrageenan concentration generally decreased the solution supersaturation and hence decreased the propensity for calcite formation. At certain higher concentrations, however, it was possible that the -carrageenan

log

(

1 (2.303kBT)3 (log )2

2s2

)

where is a shape factor for the calcite nuclei () 16/3 for spherical shapes), is the molar volume of calcite () 1.89 10-5 m-3), s is the surface energy of the calcite nuclei, kB is Boltzmanns constant, and T is the temperature. Therefore, as the polymer concentration changes the two factors that may be affected are the solution supersaturation, as the polymer binds more calcium from solution, and the nucleus surface energy, as the polymer may become associated with the growing crystal. In the case of LM pectin, it is known that the polymer binds calcium ions, as this is how the pectin gel is formed. At concentrations above 1%, no calcite crystals formed, and the entire solution became a single lump of gel. In this case, the LM pectin had bound all of the available calcium ions leaving none free for crystallization, which is represented in the above equation for the induction time as the limit when supersaturation tended to zero. At the concentrations studied, between 0.2 and

794 Crystal Growth & Design, Vol. 6, No. 3, 2006

Butler et al.(34) Linhai, Y.; Dalai, J. Chin. Sci. Bull. 2004, 49, 235. (35) Sedlak, M.; Antonietti, M.; Colfen, H. Macromol. Chem. Phys. 1998, 199, 247. (36) Qi, L.; Li, J.; Ma, J. AdV. Mater. 2002, 14, 300. (37) Yu, S.-H.; Colfen, H.; Antonietti, M. J. Phys. Chem. B 2003, 107, 7396. (38) Marentette, J. M.; Norwig, J.; Stockelmann, E.; Meyer, W. H.; Wegner, G. AdV. Mater. 1997, 9, 647. (39) Colfen, H.; Qi, L. Chem. Eur. J. 2001, 7, 106. (40) Endo, H.; Schwahn, D.; Colfen, H. J. Phys. Chem. 2003, 107, 7396. (41) Sugawara, T.; Suwa, Y.; Ohkawa, K.; Yamamoto, H. Macromol. Rapid Commun. 2003, 24, 847. (42) Euliss, L. E.; Trnka, T. M.; Deming, T. J.; Stucky, G. D. Chem. Commun. 2004, 15, 1736. (43) Marsh, M. E. in Biomineralization: From Biology to Biotechnology and Medical Applications; Baeuerline, E., Ed.; Wiley-VCH: Weinheim, 2000; pp 251-268. (44) Didymus, J. M.; Oliver, P.; Mann, S.; DeVries, A. L.; Hauschka, P. V.; Westbroek, P. J. Chem. Soc., Faraday Trans. 1993, 89, 2891. (45) Manoli, F.; Dalas, E. J. Mater. Sci.: Mater. Med. 2002, 13, 155. (46) Verraest, D. L.; Peters, J. A.; van Bekkum, H.; van Rosmalen, G. M. J. Am. Oil Chem. Soc. 1996, 73, 55. (47) Falini, G.; Gazzano, M.; Ripamonti, A. J. Cryst. Growth 1994, 137, 577. (48) Yang, L.; Zhang, X.; Liao, Z.; Guo, Y.; Hu, Z.; Cao, Y. J. Inorg. Biochem. 2003, 97, 377. (49) Arias, J. L.; Neira-Carrillo, A.; Arias, J. I.; Escobar, C.; Bodero, M.; David, M.; Fernandez, M. S. J. Mater. Chem. 2004, 14, 2154. (50) Zhang, S.; Gonsalves, K. E. J. Appl. Polym. Sci. 1995, 56, 687. (51) Zhang, S.; Gonsalves, K. E. Mater. Sci. Eng. C 1995, 3, 117. (52) Kato, T.; Suzuki, T.; Amamiya, T.; Irie, T.; Komiyama, M.; Yui, H. Supramol. Sci. 1998, 5, 411. (53) Kato, T.; Suzuki, T.; Irie, T. Chem. Lett. 2000, 2, 186. (54) Hosoda, N.; Kato, T. Chem. Mater. 2001, 13, 688. (55) Kotachi, A.; Miura, T.; Imai, H. Chem. Mater. 2004, 16, 3191. (56) Sugawara, A.; Kato, T. Chem. Commun. 2000, 6, 487. (57) Meyer, H. J. J. Cryst. Growth 1984, 66, 639. (58) Harding, S. E. Prog. Biophys. Mol. Biol. 1997, 2/3, 207. (59) Liu, X. Y.; Boek, E. S.; Briels, W. J.; Bennema, P. Nature 1995, 374, 342. (60) Reyhani, M. M.; Oliveira, A.; Parkinson, G. M.; Jones, F.; Rohl, A. L.; Ogden, M. I. Int. J. Mod. Phys. B 2002, 16, 25. (61) de Yoreo, J. J.; Dove, P. M. Science 2004, 306, 1301. (62) Sinha, A.; Chakraborty, J.; Das, S. K.; Ramachandrarao, P. Curr. Sci. 2003, 84, 1437. (63) Falini, G.; Fermani, S.; Gazzano, M.; Ripamonti, A. J. Chem. Soc., Dalton Trans. 2000, 10, 3983. (64) Falini, G.; Fermani, S.; Gazzano, M.; Ripamonti, A. Chem. Eur. J. 1997, 3, 1807. (65) Falini, G. Int. J. Inorg. Mater. 2000, 2, 455. (66) Bigi, A.; Boanini, E.; Walsh, D.; Mann, S. Angew. Chem., Int. Ed. 2002, 41, 2163. (67) Busch, S.; Schwarz, U.; Kniep, R. Chem. Mater. 2001, 13, 3260. (68) Kunther. J.; Seshadri, R.; Nelles, G.; Assenmacher, W.; Butt, H.-J.; Mader, W.; Tremel, W. Chem. Mater. 1999, 11, 1317. (69) Rautaray, D.; Sinha, K.; Shankar, S. S.; Adyanthaya, S. D.; Sastry, M. Chem. Mater. 2004, 16, 1356. (70) Yu, S. H.; Colfen, H.; Hartmann, J.; Antonietti, M. AdV. Funct. Mater. 2002, 12, 541. (71) Keum, D.-K.; Naka, K.; Chujo, Y. Chem. Lett. 2004, 33, 310. (72) Dalas, E.; Klepetsanis; Koutsoukos, P. G. J. Coll. Int. Sci. 2000, 224, 56a.

conformation became important and favored calcite nucleation via a surface energy decrease in the same way as for LM pectin, as an enhancement in nucleation was observed. Acknowledgment. The authors thank Unilever for permission to publish this paper. References(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) Addadi, L.; Raz, S.; Weiner, S. AdV. Mater. 2003, 15, 959. Colfen, H. Curr. Opin. Coll. Int. Sci. 2003, 8, 23. Mann, S.; Perry, C. C. AdV. Inorg. Chem. 1991, 36, 137. Meldrum, F. C.; Hyde, S. T. J. Cryst. Growth 2001, 231, 544. Orme, C. A.; Noy, A.; Wierzbicki, A.; McBride, M. T.; Grantham, M.; Teng, H. H.; Dove, P. M.; DeYoreo, J. J. Nature 2001, 411, 775. Volkmer, D.; Fricke, M.; Huber, T.; Sewald, N. Chem. Commun. 2004, 16, 1872. Malkaj, P.; Dalas, E. Cryst. Growth Des. 2004, 4, 721. Tong, H.; Ma, W.; Wang, L.; Peng, W.; Hu, J.; Cao, L. Biomaterials 2004, 25, 3923. Manoli, F.; Kanakis, J.; Malkaj, P.; Dalas, E. J. Cryst. Growth 2002, 236, 363. Li, C. M.; Botsaris, G. D.; Kaplan, D. L. Cryst. Growth Des. 2002, 2, 387. Levi, Y.; Albeck, S.; Brack, A.; Weiner, S.; Addadi, L. Chem. Eur. J. 1998, 4, 389. Wierzbicki, A.; Sikes, C. S.; Madura, J. D.; Drake, B. Calcif. Tissue Int. 1994, 54, 133. DeOliveira, D. B.; Laursen, R. A. J. Am. Chem. Soc. 1997, 119, 10627. Jimenez-Lopez, C.; Rodriguez-Navarro, A.; Dominguez-Vera, J. M.; Garcia-Ruiz, J. M. Geochim. Cosmochim. Acta 2003, 67, 1667. Shen, F. H.; Feng, Q. L.; Wang, C. M. J. Cryst. Growth 2002, 242, 239. Raz, S.; Weiner, S.; Addadi, L. AdV. Mater. 2000, 12, 38. Kawaguchi, H.; Hirai, H.; Sakai, K.; Sera, S.; Nakajima, T.; Ebisawa, Y.; Koyama, K. Colloid Polym. Sci. 1992, 270, 1176. Rieger, J.; Thieme, J.; Schmidt, C. Langmuir 2000, 16, 8300. Shakkthivel, P.; Sathiyamoorthi, R.; Vasudevan, T. Desalination 2004, 164, 111. Iwatsubo, T.; Sumaru, K.; Kanamori, T.; Yamaguchi, T.; Sinbo, T. J. Appl. Polym. Sci. 2004, 91, 3627. Hadicke, E.; Rieger, J.; Rau, I. U.; Boeckh, D. Phys. Chem. Chem. Phys. 1999, 1, 3891. Verdoes, D.; van Landschoot, R. C.; van Rosmalen, G. H. J. Cryst. Growth 1990, 99, 1124. Colfen, H.; Antonietti, M. Langmuir 1998, 14, 582. Dalas, E.; Klepetsanis, P.; Koutsoukos, P. G. Langmuir 1999, 15, 8322. Malkaj, P.; Chrissanthopoulos, A.; Dalas, E. J. Cryst. Growth 2002, 242, 232. Pai, R. K.; Hild, S.; Ziegler, A.; Marti, O. Langmuir 2004, 20, 3123. Naka, K.; Tanaka, Y.; Chujo, Y. Langmuir 2002, 18, 3655. Naka, K.; Chujo, Y. C. R. Chimie 2003, 6, 1193. Naka, K. Top. Curr. Chem. 2003, 228, 141. Dimova, R.; Lipowsky, R.; Mastai, Y.; Antonietti, M. Langmuir 2003, 19, 6097. Sedlak, M.; Colfen, H. Macromol. Chem. Phys. 2001, 202, 587. Kaluzynscki, K.; Pretula, J.; Lapienis, G.; Basko, M.; Bartzcak, Z.; Dworak, A. J. Polym. Sci. A, Polym. Chem. 2001, 117, 200. Bolze, J.; Pontoni, D.; Ballauff, M.; Narayanan, T.; Colfen, H. J. Coll. Int. Sci. 2004, 277, 84.

CG050436W