Nucleation of Calcium Carbonate as Polymorphic Crystals in the

8859

2008, 112, 8859–8862 Published on Web 07/04/2008

Nucleation of Calcium Carbonate as Polymorphic Crystals in the Presence of Lipid A-Diphosphate Chester A. Faunce, Hendrik Reichelt, and Henrich H. Paradies* Joule Physics Laboratory, Materials Research Institute, UniVersity of Salford, Manchester M5 4WT, United Kingdom ReceiVed: April 09, 2008; ReVised Manuscript ReceiVed: June 15, 2008

The well-defined structure of lipid A-diphosphate in aqueous solutions provides a way of observing the formation of calcium carbonate crystals. The crystals are either tetrahedral or rhombohedral calcite at a volume fraction of φ ) 5.4 × 10-4 at pH 5.8 or the vaterite polymorph of CaCO3 at a volume fraction of φ ) 7.8 × 10-4 at pH 5.8. In both cases, nucleation, adsorption pH, and the shape-dependent template of lipid A-diphosphate control the formation of the calcite and vaterite. Crystal formation and growth in the presence of biologically active organic molecules are of prime importance in the understanding of nanometric and mesoscopic structures. This is particularly true in the case of microbial-induced metal corrosion and in the fabrication of biochemical materials, as well as for Biofilm (exopolymers), which includes Ca2+dependent biodegradation.1 Recent evidence discloses the importance of calcium carbonate and hydroxyapatite in the formation of nanobacteria2 in a number of medical conditions. These conditions include renal stone formation,3 polycystic kidney diseases,4 rheumatoid arthritis,5 and coinfection with HIV.6 A further inclusion must be the influence in the activity of cationic antimicrobial peptides (CAM) on Gram-negative bacteria.7 When calcium carbonate crystals are constrained at fluid interfaces, interesting structures are encountered.8 For these structures, the spatial arrangement, the distribution, and the nature of charged groups on the template are of significant importance.9 During the nucleation stages, the template is expected to be invariant to shape and conformational changes which facilitate crystal growth.10 Template-mediated nucleation of calcium carbonate is used to investigate proteins, nonfunctional amphiphiles, vesicles, lyotropic crystalline systems,11 and synthetic chiral alanylalanine-derived poly(isocyanide) polymers.12 For lipid A-phosphate materials from E. coli (Figure 1a), it is difficult to achieve controlled oriented nucleation because in aqueous solutions, these molecules are polydisperse in both charge and size. The preparation of lipid A-diphosphate dispersions with low polydispersities in charge and shape was presented in a previous publication.13 Also described previously was the influence of a 10 nM concentration of Ca2+ ions of lipid A-diphosphate in aqueous dispersions at low ionic strength.14 Discussed in the latter paper is the unique role that Ca2+ possesses in influencing the formation of a discrete assembly. Also revealed is the role of both the Ca2+ and Mg2+ ions in the modification of ordering * To whom correspondence should be addressed. Phone: (+44)-161295-4286or (+49)-2371-12305. Fax: (+44)-161-295-5575 or (+49)2371-149705. E-mail: [email protected].

10.1021/jp803067b CCC: $40.75

of lipid A-diphosphate assemblies and how essential these counterions are in limiting bacterial growth.15 Lipid A-diphosphate (Figure 1a), the abundant and conserved component of Gram-negative bacterial lipopolysaccharides (LPS), is a glycolipid found on the surface of Gram-negative bacteria.16 The molecule and its self-assembly17 are of major importance because it provides the host (i.e., humans) with a defense against infection by microorganisms; however, it may also give rise to the development of lethal shock.18 The interplay of CAM with LPS or lipid A-diphosphate in the presence of Ca2+ ions and bicarbonate is not fully understood. However, pH, Ca2+ and Mg2+, and the effective net charges of the CAM and lipid A-diphosphate appear to be crucial for antimicrobial activity and for the host’s defense after bacterial and viral infections.19 Reported in the present paper is the formation of different types of CaCO3 crystals at volume fractions of φ ) 5.4 × 10-4 or 7.8 × 10-4 at a pH of 5.8 or 8.5. Consequently, in order to form the calcite structures, the Ca2+ concentration must exist at a much higher level than that discussed previously14 (1.0-5.0 mM). Nevertheless, the as-prepared colloidal dispersions of lipid A-diphosphate at pH 5.8 and 8.5 for both of the above volume fractions are transparent. The colloidal assemblies of lipid A-diphosphate possess either a spherical shape with a radius of R ) 35.5 nm (MW ) 10.5 × 106 g/mol) at pH 5.8 or a rodlike morphology with L ) 800 nm and a diameter of d ) 2.8 nm (MW ) 5.9 × 106 g/mol) at pH 8.513,20 (see Supporting Information). The lipid A-diphosphate dispersions were introduced into 1 mL of a supersaturated calcium bicarbonate solution (5 mM, 25 °C) and left to crystallize at an appropriate pH. After a slow evaporation, when crystallization subsided, calcite crystals formed, with a spontaneous loss of CO2 at a final pH of 5.8 (Figure 1b). An alternative preparation method is to introduce lipid A-diphosphate into 1 mL of a 1 mM calcium oxide solution, with a final pH of 8.5. Using this method, the crystals grew to form tetrahedral CaCO3, similar to ones reported previously8 (Figure 2b). At a volume fraction of φ ) 7.8 × 10-4 and pH 5.8, the crystals developed in the form of vaterite  2008 American Chemical Society

8860 J. Phys. Chem. B, Vol. 112, No. 30, 2008

Figure 1. (a) Chemical structure of lipid A-diphosphate; (b) calcite crystals grown in the absence of lipid A-diphosphate; bar ) 1 µm; (c) scanning electron micrograph of calcite crystals observed at the interface lipid A-diphosphate-water (φ ) 5.4 × 5.8 × 10-4, pH 5.8); (d) some of the calcite crystals show truncated corners (arrows) and cavities on the (104) plane (cross diameter ) 0.25 µm), revealing the intersection of a plane and the corner of a (104) rhombohedral calcite crystal; bar ) 10 µm; (e) X-ray powder diffraction patterns (Siemens D-5000 diffractometer, Ni filtered Cu KR radiation, λ ) 1.54 nm) of the CaCO3 crystals obtained with lipid A-diphosphate as a template at different pH values. The black trace relates to calcite crystals formed in the presence of 1 mM Ca2+ at φ ) 7.8 × 10-4 and pH 8.5. The red trace corresponds to the vaterite modification of CaCO3 in the presence of the template at pH 5.8 and φ ) 7.8 × 10-4 (20 °C).

(Figure 2c, d). The introduction of lipid A-diphosphate into the crystallization process resulted in the formation of oriented CaCO3 crystals displaying various morphologies. A nucleation density of approximately 500 mm-2 was determined, which was significantly higher than that found in the bulk solutions.4 In fact, a nucleation density of this magnitude is in the range of the Langmuir monolayer template. The most common crystals to appear were either rhombohedral (Figures 1b and 2a) with six faces, each nucleated on the (104) crystal plane, or tetrahedral crystals at pH 8.5 (Figure 2b), crystals that normally form in heated solutions.8,21 Agglomerates of hexagonal plates of vaterite were seen at a higher volume fraction, φ ) 7.8 × 10-4, and at pH 5.8 (Figure 2c). Only rhombohedral calcite crystals were observed in the volume fraction range between φ ) 5.8 and 7.8 × 10-4 Crystallizations of CaCO3 in the presence of the template obtained at φ ) 5.8 × 10-4 and pH 5.8 exhibited only reflections from the (10.4) planes, while at higher volume fractions, the additional reflections of the (10.2), (11.0), (11.3), and (20.2) faces were observed. The FT-IR spectrum (Figure 2e), TEMSADP, and X-ray powder diffraction results (Figure 1e) showed the crystalline materials to be either calcite or vaterite. No effect on molecular weight or polydispersity of the template was observed with crystal formation at either pH. Crystal growth

Letters was seen on the surface, where crystals were entrained by the strong surface tension of the lipid A-diphosphate dispersions. For rhombohedral calcite, crystal growth was along the c-axis, normal to the surface. Thus, the (001) surfaces were the basal planes and displayed smooth rhombohedral side faces with truncated corners and concave depressions. SEM images revealed corners which consisted of three symmetry-related faces at 120° to each other. Other calcite crystals emanated from a flat (001) face of the truncation (Figure 1b). When viewed from the underside, the crystals showed no truncated corners. Exhibited in Figure 2a are layered and terraced calcite crystals and crystals displaying truncated corners on the top, as well as very small rhombohedral crystals. At this volume fraction and pH, the lipid A-diphosphate template characterized the layered structure, the truncated corners on the (104) face, and the symmetry of the crystals. However, the morphology was related to the initial Ca2+ concentration and pH, which exhibited a different crystal form of the CaCO3 (Figure 2b). At [Ca2+] ) 1 mM, tetrahedral crystals appeared at pH 8.5, irrespective of the volume fraction, φ. Note, at pH 8.5, the template had the shape of a negatively charged rod with a length of 800 nm and d ) 2.8 nm.20 These crystals showed three equivalent (104) faces and a fourth face on a different plane, reminiscent of observations made by Wickman and Korley22 in their “w” layer and more recently by Hashmi et al.8 The base of the tetrahedral crystal was an isosceles triangle (Figures 1b and 2b), and an isosceles angle of 76 to 92° was measured from approximately 65 crystals on SEM images. These results confirm that the intersection of either the (010) plane or the (012) plane with the (104) rhombohedron forms an isosceles angle of between 81 or 73°.8,22 This suggests that the bases of these tetrahedral crystals lie on or between the (010) and (012) calcite planes. At [Ca2+] ) 1 mM, no rhombohedral single crystals or dimer of tetrahedral crystals were observed, and no tetrahedral crystal aggregation was encountered. It would appear that the shape and charge asymmetry of the lipid A-diphosphate template at pH 8.5 in an aqueous environment facilitated tetrahedral crystal growth. The asymmetry-initiated tetrahedral crystal growth on only three (104) faces differed considerably from the bulk rhombohedral crystals grown at pH 5.8, which possessed six (104) faces with a neutral surface charge composed equally of Ca2+ and CO32- ions.23 A more loosely packed polymorph structure of vaterite formed at a volume fraction of φ ) 7.8 × 10-4 (pH 5.8). These crystals appeared as hexagonal plates24 or as a mosaic of thin flat hexagons. Agglomerates of hexagonal plates were observed between volume fractions of φ ) 6.5 × 10-4 and 7.0 × 10-4 (Figure 2c) and, in a few cases, were mixed with rhombohedral calcite crystals. The formation of aragonite crystals was not detected. The SADPs from isolated particles exhibited hexagonal symmetry and shape (Figure 2d), indicating that the vaterite particles were single crystals and monolithic (see Supporting Information). The vaterite crystals produced in accordance with this protocol crystallized as hexagonal platelets on the (001) plane rather than in spherical particle clusters, which display all crystallographic directions. There were, however, differences in the intensity distribution of the X-ray powder diffraction lines (Figure 1e) between bulk vaterite and the template-mediated vaterite, for example, the intensity of (112) reflection was much higher than the bulk vaterite (100 vs 50). The magnitude of the (114) reflection from the bulk vaterite crystals was significantly lower than the template-mediated one (65 vs 100). The (300), (118), and the (224) reflections were all greater in intensity than those of the corresponding bulk vaterite crystals. It would appear

Letters

J. Phys. Chem. B, Vol. 112, No. 30, 2008 8861

Figure 2. (a) Scanning electron microscope (SEM) images of rectangular projections and terraces on adjoining (104) planes distal to the truncated corner; (b) SEM images of tetrahedral calcite crystals obtained at [Ca2+] ) 1 mM, pH 8.5, by dipping a TEM grid on top of the crystallized sample; bar ) 1 µm; inset: view of the tetrahedral crystal from above with the isosceles angle (R) of the base; (c) SEM image of agglomerated vaterite crystals obtained from the top of the crystallized sample (φ ) 7.8 × 10-4); bar ) 1 µm; (d) SEM image of a mat of vaterite crystals at pH 5.8; bar ) 10 µm; iInset: selected-area electron diffraction pattern (SADP) of an isolated vaterite crystal; (e) FT-IR spectra (Bruker FTIR spectrometer) of the two polymorphic forms of calcium carbonate obtained. Spectrum 1 was recorded from calcite crystals produced at pH 5.8 from a supersaturated calcium bicarbonate solution at a volume fraction of 5.5 × 10-4, according to Kitano;21 spectrum 2 (solid red line) corresponds to calcite crystals grown at a volume fraction of φ ) 5.4 × 10-4 for 1.0 mM calcium oxide at pH 8.5. Spectra 1 and 2 also show the typical calcite vibrations at 714 and 876 cm-1. Spectrum 3 shows a FT-IR spectrum of vaterite crystals, the IR absorption bands at 1070, 877, and 746 cm-1, as well as a general spectrum displaying the broad carbonate band at 1430 cm-1. Spectrum 3 corresponds to the vaterite crystals developed at a lipid A-diphosphate volume fraction of φ ) 7.8 × 10-4 and pH 5.8.

that the vaterite crystals formed in the presence of the lipid A-diphosphate template developed a preferred orientation because the magnitude of the X-ray powder diffraction intensities was quite different. The FT-IR absorption band at 746 cm-1 was not detected for calcite but was observed for the vaterite form.25 Growth appeared to be restricted in some crystallographic directions, resulting in a hexagonal rather than a spherical form of vaterite.26 In vitro crystallization demonstrated that lipid A-diphosphate induced and stabilized the metastable vaterite phase above a volume fraction of φ ) 7.8 × 10-4 at low ionic strength, pH 5.8, and ambient temperature. The morphology of the vaterite phase did not change at higher volume fractions, unlike the formation of the calcite phase at the significantly lower volume fraction value of φ ) 5.8 × 10-4, at pH 8.5, and in the presence of 1 mM Ca2+. An elemental analysis of the supersaturated bicarbonate solution showed that the Mg2+ concentration was

too low with respect to Ca2+ to allow the nucleation of significant amounts of aragonite. Therefore, in light of the considerable influence of Ca2+ and Mg2+ ions on the structure and activities of lipid A-diphosphate,14,15 the experiments were repeated in the presence of 150 µM Mg2+ in the same crystallization milieu. Even at this Mg2+ concentration, no changes in the above findings were observed. However, recrystallization experiments in the presence of higher amounts of Mg2+ (150 mM) revealed aragonite to be the major polymorph of the precipitated CaCO3.27 The crystalline CaCO3 particles that formed remained well separated between the interfaces of the lipid A-diphosphate assemblies and the aqueous surrounding. Furthermore, the CD spectrum between 210 and 500 nm did not change with time or [Ca2+] at either pH, indicating that no changes occurred in the secondary structures of the template upon crystallization. No contamination of lipid A-diphosphate by the CaCO3 crystals

8862 J. Phys. Chem. B, Vol. 112, No. 30, 2008 was detected using a MALDI-TOF-MS and LC-MS. Interparticle interactions at a well-defined screening length of k-1/2 ∼ 9.5 nm (surface charge density)13 between the template, crystals, and water may have arisen from a balance between long-range attractive and repulsive forces at the above screening length. Roughness-induced capillary effects8,28 and irregular meniscus defects may explain the prevalence of the observed vaterite cluster at high volume fractions in the subphase. Loose aggregates of vaterite formed at high volume fractions by suppressing the growth of calcite. The development of a specific well-formed mineral layer by the various polymorphs of CaCO3 with lipid A-diphosphate as a template may cause direct resistance to viral and bacterial invasion and/or penetration with respect to CAM activity. The CAM activity was not changed in the presence of 10 µM to 5 mM CaCl2 at pH 8.5. However, it was significantly altered in the presence of 10 mM CaCl2 but revealed no further increase in CAM activity.27 Acknowledgment. The authors acknowledge Ms. J. M. Aldred for proofreading and Prof. S. E. Donnelly for critical discussions. Supporting Information Available: SAXS, LS, SADP, FTIR, and experimental procedures. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) (a) Biomineralization: Chemical and Biochemical PerspectiVes; Mann, S., ; Webb, J., ; Williams, R. J. P, Eds.; VCH Publishers: New York, 1989. (b) Paradies, H. H.; Thies, M.; Hinze, U. Rigaku J. 1996, 13, 16. (c) Paradies, H. H. Chemical and Physical-Chemical Aspects of MetalBiofilms. In The Biology of World Resources; Videla, H. A., ; Gaylarde, C. C., Eds.; Cambridge University Press: New York, 1995; Vol. II, pp 197279. (d) Paradies, H. H. Physical Chemistry and Polymer Dynamics of Exopolymers in Solution and Their Interfacial Relation to Surfaces. In 4th Microbial Induced Corrosion; Sequeira, C., ; Tiller, K., Eds.; Materials Institute London: U.K., 2000; pp 170-195. (e) Clancy, S. F.; Tanner, D. E.; Thies, M.; Paradies, H. H. J. Chem. Soc., Chem. Commun. 1997, 2036. (f) Paradies, H. H.; Quitschau, P.; Pischel, I. Z. Phys. Chem. (Munich) 2000, 3, 379. (2) Martel, J.; Young, J. D.-E. Proc. Natl. Acad. Sci. U.S.A 2008, 105, 5549. (3) Kajander, E. O.; Ciftcioglu, N. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 8274. (4) Hjelle, J. T.; Miller-Hielle, M. A.; Poxton, I. R.; KIajander, E. O.; Ciftcioglu, N.; Jones, M. L.; Caughey, R. C.; Brown, P. D.; Darras, F. S. Kidney Int. 2000, 57, 2360. (5) Tsurumoto, T.; Matsumoto, T; Yonekura, A.; Shindo, H. J. Proteome Res. 2006, 5, 1276. (6) (a) Sommer, A. P.; Milamkovits, M.; Mester, A. R. Chemotherapy 2006, 52, 95. (b) ZimmermannK.; ParadiesH. H. U.S. Patent 2000, 6,117,849. (7) Sugiarto, H.; Yu, P. L. Biochem. Biophys. Res. Commun. 2004, 323, 721. (8) Hashmi, S. M.; Wickman, H. H.; Weitz, D. A. Phys. ReV. E 2005, 72, 041605.

Letters (9) (a) Addadi, L.; Weiner, S. Angew. Chem., Int. Ed. Engl. 1992, 153. (b) Addadi, L.; Weiner, S. Proc. Natl. Acad. Sci. U.S.A. 1985, 82, 4110. (c) Lahiri, J.; Xu, G.; Dabbs, D. M.; Yao, N.; Aksay, I. A.; Grovers, J. T. J. Am. Chem. Soc. 1997, 119, 5449. (d) Lakshinarayanan, R.; Vivekanandan, S.; Samy, R. P.; Banerjee, Y.; Cgi-Jin, E. O.; Teo, K. W.; Jois, S. D. S.; Kimi, R. M.; Valinyaveettil, S. J. Am. Chem. Soc. 2008, 130, 4660. (e) Mann, S. Nature 1993, 365, 499. (f) Heywood, B. R.; Rajam, S.; Mann, S. J. Chem. Soc., Faraday Trans. 1991, 87, 735. (g) Fricke, M.; Volkmar, D.; Krill, C. E., III; Kellermann, M.; Hirsch, A. Cryst. Growth Design 2006, 6, 1120. (h) Wang, W.; Wang, G.; Liu, Y.; Zheng, C.; Zhan, Y. J. Mater. Chem. 2001, 11, 1752. (10) Lahiri, J.; Xu, G.; Dabbs, D. M.; Yao, N.; Aksay, I. A.; Grovers, J. T. J. Am. Chem. Soc. 1997, 119, 5449. (11) (a) Young, H.; Coombs, N.; Ozin, G. N. Nature 1997, 386, 692. (b) Yang, H.; Ozin, G. A.; Kresge, C. T. AdV. Mater. 1998, 10, 883. (c) Stupp, S. I.; Braun, P. V. Science 1997, 277, 1242. (d) Hartgerink, J. D.; Benianiash, E.; Stupp, S. I. Science 2001, 294, 1684. (12) Donners, J. J. J. M.; Nolte, R. J. M.; Sommerdijk, N. A. J. M. J. Am. Chem. Soc. 2002, 124, 9700. (13) (a) Faunce, C. A.; Paradies, H. H.; Quitschau, P. J. Phys. Chem. B 2003, 107, 2214. (b) Reichelt, H.; Faunce, C. A.; Paradies, H. H. J. Phys. Chem. B 2008, 112, 3290. (14) Faunce, C. A.; Reichelt, H.; Paradies, H. H.; Quitschau, P.; Zimmermann, K. J. Chem. Phys. 2005, 122, 214727. (15) Ernst, R. K.; Yi, E. C.; Guo, L.; Lim, K. B.; Burns, J. L.; Hacket, M.; Miller, S. I. Science 1996, 286, 1561. (16) (a) Raetz, C. R.; Whitfield, C. Annu. ReV. Biochem. 2002, 71, 635. (b) Christ, W. J.; Hawkins, L. D.; Lewis, M. D.; Kishi, Y. In CarbohydrateBased Drug DiscoVery; Wong, C.-H., Ed.; Wiley-VCH: New York, 2003; Vol. 1, pp 341-355. (c) Aschauer, H.; Gros, A.; Hildebrandt, J.; Schuetze, E.; Stuetz, P. J. Biol. Chem. 1990, 265, 9159. (17) (a) Faunce, C. A.; Reichelt, H.; Paradies, H. H.; Quitschau, P.; Rusch, V.; Zimmermann, K. J. Phys. Chem. B 2003, 107, 9943. (b) Thies, M.; Quitschau, P.; Zimmermann, K.; Rusch, V.; Faunce, C. A.; Paradies, H. H. J. Chem. Phys. 2002, 116, 3471. (18) (a) Alexander, C.; Rietschel, E. T. J. Endotoxin Res. 2000, 7, 167. (b) The release of endotoxin is intimately associated with septic shock. This problem caused more than 21000 mortalities in 1996 in the U.S. alone. See: Morbidity and Mortality Weekly Report; Centre for Disease Control: Atlanta, GA, 1997; Vol. 46, pp 941-944. (19) (a) Wehkamp, J.; Bals, R.; Kreft, B.; Schroeder, J.-M.; Stange, E. F. Dtsch. A¨rzteblatt 2007, 104, A257–262. (b) Mims, C. A.; Nosh, A.; Stephen, J. Mim’s Pathogenesis of Infectious Disease, 5th ed.; Academic Press: London, San Diego, CA, 2001; pp 73-82. (20) Faunce, C. A.; Paradies, H. H. J. Chem. Phys. 2008, 128, 065105. (21) Kitano, Y. Bull. Chem. Soc. Jpn. 1962, 12, 1980. (22) Wickman, H. H.; Korley, J. N. Nature 1998, 393, 445. (23) Lippmann, F. Sedimentary Carbonate Minerals; Springer Verlag: New York, 1973. (24) Meyer, H. J. Z. Kristallogr. 1969, 128, 183. (25) Dupont, L.; Portemer, F.; Figlarz, M. J. Mater. Chem. 1997, 7, 797. (26) (a) Xyla, A. G.; Giannimaras, E. K.; Koutsoukas, P. G. Colloids Surf. 1991, 53, 241. (b) Shaik, A. M. J. Appl. Crystallograph. 1990, 23, 263. (27) Reichelt, H.; Paradies, H. H.; Quitschau, P.; Faunce, C. A. These assemblies comprising of lipid A-diphosphate and calcium carbonate exhibit chemical and physcial survival mechanisms with respect to chronic inflammation, release of endotoxin or significant loss of endotoxicity, and cellular immunity with respect to cytokines. (28) Stamou, D; Duschl, C.; Johannsmann, D. Phys. ReV. E 2000, 62, 5263.

JP803067B