PhD THESIS - Calcium Carbonate Bio-precipitation in Gelling Environments via Counter-diffusion

PhD THESIS C a l c i u m C a r b o n at e B i o - p r e c i p i tat i o n i n Gelling Environments via Counter-diffusion

María Sancho Tomás

Laboratorio de Estudios Cristalográficos, Instituto Andaluz de Ciencias de la Tierra, CSIC-UGR Dipartimento di Chimica «Giacomo Ciamician», Università di Bologna 2014

PhD THESIS C a l c i u m C a r b o n at e B i o - p r e c i p i tat i o n i n Gelling Environments via Counter-diffusion

María Sancho Tomás -Supervisor: Jaime Gómez-Morales Co-supervisor: Giuseppe Falini Tutor: Juan Manuel García Ruiz -Laboratorio de Estudios Cristalográficos, Instituto Andaluz de Ciencias de la Tierra, CSIC-UGR

Dipartimento di Chimica «Giacomo Ciamician», Università di Bologna 2014

5

Acknowledgements

First and foremost, I would like to thank my two supervisors, Dr. Jaime Gómez Morales and Prof. Giuseppe Falini for giving me the opportunity to be part of both research groups and their professional influence on my development as a scientist. Thanks also to Juan Manuel García Ruiz for introducing me to the LEC group. Besides, I would like to thank CSIC for the JAE-Pre research contract within the “Junta para la Ampliación de Estudios” co-funded by the European Social Fund (ESF).

Quiero agradecer a Sander, José, Poto, Miguel, Gloria, Pilar, Raquel, Mayte, Gavi y Emilio el apoyo científico durante estos cinco años. Sin estas personas este trabajo no sólo no hubiera sido el mismo sino que también hubiera tenido mucho menos color y alegría.

Grazie a Simona, Damiano, Miki, Erik e Guido per l’auto e per avere reso così facile ogni giorno al laboratorio. Volevo ringraziare anche “casa Bentivogli” e “casa Oriani” dove ho incontrato le mie storiche amiche italiane, ancora così vicine, perché la vita nella “Bologna sotterranea” non sarebbe stata possibile senza nessuna di loro.

Cinco años dan para mucho y en Granada, más. Gracias familia Alkaparreña y resto de Hortigas, con las verduras diurnas y nocturnas, bailando al son de DJ Diva & Bambino. También a mi equipo de Rugby Veleta, con esas mujeres increíbles de color morado. A Varela y sus alrededores Realejenses, con LaModaKinKomoda y las Perras de Laika, a Pink Floyd y al avión número 13 de Puppa y Do, a Bonsai y a la futura Tropicalia, y a las pachangas de baloncesto con todos sus dedos rotos. Y sobre todo, a la Reina del Realejo, a Conxa.

Y a pesar de que la vida da muchas vueltas y los años pasan, hay personas que ahí siguen.

Por ello, gracias a mis estrellas marinas del balón de baloncesto; a mi gente de parque, risas y actividades culturales varias de los martes; y a las chicas de la salita con sus cafecicos tertulianos.

Por último, gracias a la familia y su luz, a Mario, Teresa, Juanjo, Daniel, Montse, Pascual, Carmen, Adolfo y Marifrán. A Pepe, Joaquina, Isabel, Joaquín, Luisa, Paco y a los eventos Caspolinos “Always look on the bright side of life”. A Szuka buceadora. A mi abuelo, Ramón Tomás Valero, el mejor del mundo entero. Y sin dudar, a mi madre y mi hermana, por ser esas alas de mariposa que mueven el mundo.

A mi padre, con amor.

7

00 Index

index

Symbols and abbreviations

12

4.2. Experimental section

74

List of figures

13

4.3. Results

74

List of tables

17

4.4. Discussion

82

1. Introduction and general objectives

4.5. Conclusions

84

21

1.1. Biomineralization

21

5. Influence of soluble organic matrix from scleractinian corals

87

1.2. Crystallization

25

5.1. Introduction

87

1.3. Calcium carbonate, CaCO3

34

5.2. Experimental section

88

1.4. Additives

39

5.3. Results

90

1.5. Objectives and chapters summary

45

5.4. Discussion

100

2. Experimental and theoretical approach

49

5.5. Conclusions

103

2.1. Organic additives

49

107

2.2. Agarose

50

6. Influence of soluble organic matrix from nacre and sea urchin spines

2.3. Crystallization set-up

52

6.1. Introduction

107

2.4. Theoretical considerations

53

6.2. Experimental section

108

2.5. Characterization of CaCO3 precipitates

6.3. Results

109

57

6.4. Discussion

120

3. Influence of charged polypeptides

61

6.5. Conclusions

123

3.1. Introduction

61

127

3.2. Experimental section

62

7. Fluorescence to monitor pH-changes in counter-diffusion experiments

3.3. Results

63

7.1. Introduction

127

3.4. Discussion

68

7.2. Experimental section

128

3.5. Conclusions

70

7.3. Results and discussion

132

4. Influence of diffusing Mg2+ ions together with entrapped polypeptides

7.4. Conclusions

137

73

8. Summary and Conclusions

141

4.1. Introduction

73

9. References

147

Symbols and abbreviations

Symbol

12

Description

tw

Waiting time

xo

The starting point of precipitation

Δ

Crystal growing space

xcat

The boundary of Δ closer to the cationic reservoir

xan

The boundary of Δ closer to the anionic reservoir

Δcat

The space from the starting point xo to xcat

Δan

The space from the starting point xo to xan

dc

Crystallization density

Abbreviation

Description

CDS

Counter-diffusion system

CaCO3

Calcium carbonate

NaHCO3

Sodium bicarbonate

CO32-

Carbonate ions

CaCl2

Calcium chloride

Ca2+

Calcium ions

MgCl2

Magnesium chloride

Mg2+

Magnesium ions

Mg2+/ Ca2+

Molar ratio of magnesium ions / calcium ions

SOM

Soluble organic matrix

pAsp

poly-L-aspartic acid

pGlu

poly-L-glutamic acid

pLys

poly-L-lysine

Beu

Balanophyllia europaea

Lpr

Leptopsammia pruvoti

Npo

Nautilus pompilius

Pli

Paracentrotus lividus

BeuSOM

Soluble organic matrix from Balanophyllia europaea

LprSOM

Soluble organic matrix from Leptopsammia pruvoti

NpoSOM

Soluble organic matrix from Nautilus pompilius

PliSOM

Soluble organic matrix from Paracentrotus lividus

c

Concentration of 50 μg/mL

5c

Concentration of 250 μg/mL

Ksp

Solubility product constant

S.I.

Saturation index

OM

Optical microscope

SEM

Scanning electron microscope

XRD

X-ray diffraction

FTIR

Fourier Transform Infrared spectroscopy

List of figures

13 1.1

Schematic of biologically induced and biologically controlled mineralization

23

1.2

Solubility phase diagram

26

1.3

Free energy for homogeneous nucleation based on cluster size

28

1.4

Classification of the types of faces in the crystal

30

1.5

Schematic illustration of the Terrace-Ledge-Kink model that represents different stages of the crystal growth in solution

30

1.6

Diagram of classical and non-classical crystallization

31

1.7

Picture of the U-tube

34

1.8

Schematic representation of the calcite and aragonite structure

35

1.9

Species concentration of the three inorganic forms of dissolved CO2 in water

37

1.10

Diagram of carbon exchange between atmosphere and oceans

38

1.11

Diagram of a polyp sits on the top of its skeletons

41

1.12

Underwater in situ camera and SEM pictures of B. europaea and L. pruvoti

42

1.13

Composite image illustrating a site view of the organism and a SEM image of the fractured shell of N. pompilius

43

Composite image illustrating a site view of the organism and a SEM image of the spine P. lividus

44

2.1

Structures of pAsp, pGlu and pLys

50

2.2

Agarobiose unit and agarose formation mechanism consists in two steps

51

2.3

Camera picture and schematic illustration of the U-tube set-up

52

2.4

Simulation of Ca2+, HCO3- and CO32- concentration, saturation index and amount of precipitated calcite as a function of the position (x) at different times

54

2.5

Graphical representation of the qualitative scale defined to estimate the crystallization density

56

3.1

Graphical view of the measured parameters in absence and in presence of the entrapped polypeptides pLys, pGlu and pAsp

63

3.2

XRD patterns and FTIR spectra of CaCO3 precipitated in absence and in the presence of pLys, pGlu or pAsp

66

3.3

Optical microscope pictures of the calcium carbonate crystals precipitated in the absence and in the presence of entrapped pLys, pGlu and pAsp

67

3.4

Morphology of calcium carbonate crystals formed in the agarose viscous sols in absence and in the presence of entrapped pLys, pGlu and pAsp

68

4.1

Camera pictures of the crystal growing spaces in the absence and in the presence of entrapped pLys, pGlu and pAsp and Mg2+/Ca2+ equal to 0, 1, 3 and 5

75

4.2

Graphical representations of the parameters measured in the absence and in the presence of entrapped pLys, pGlu and pAsp and Mg2+/Ca2+ equal to 0, 1, 3 and 5

76

4.3

Optical microscope pictures of calcium carbonate in the absence and in the presence of entrapped pLys, pGlu and pAsp and Mg2+/Ca2+ equal to 0, 1, 3 and 5

78

4.4

Scanning electron micrographs of CaCO3 precipitates in the absence and in the presence of entrapped pLys, pGlu and pAsp and Mg2+/Ca2+ equal to 0, 1, 3 and 5

79

4.5

FTIR spectra of CaCO3 precipitated in the absence and in the presence of entrapped pLys, pGlu and pAsp and Mg2+/Ca2+ equal to 0, 1, 3 and 5

80

4.6

XRD patterns of the CaCO3 precipitated in the absence and in the presence of entrapped pLys, pGlu and pAsp and Mg2+/Ca2+ equal to 0, 1, 3 and 5

81

5.1

Amino acid composition and FTIR spectra of the SOMs from B. europaea and L. pruvoti

90

5.2

Graphical representation of the parameters measured in the absence and in the presence of SOMs from B. europaea and from L. pruvoti

91

1.14

14

5.3

Optical microscope images of crystal growing spaces in the absence and in the presence of SOMs from B. europaea and from L. pruvoti

93

5.4

Optical microscope images of crystals in the absence and in the presence of SOMs from B. europaea and from L. pruvoti

94

5.5

XRD patterns of calcium carbonate precipitates in the absence and in the presence of SOMs from B. europaea and from L. pruvoti

95

5.6

SEM micrographs showing the morphology of CaCO3 crystals in the absence of SOMs

96

5.7

SEM micrographs showing the morphology of crystals formed in the presence of SOMs from B. europaea

97

5.8

SEM micrographs showing the morphology of crystals formed in the presence of SOM from L. pruvoti

97

5.9

Low-magnification SEM micrographs of calcium carbonate precipitates in the absence and in the presence of SOMs from B. europaea and from L. pruvoti

98

5.10

FTIR spectra and calcite/aragonite mass ratio of CaCO3 precipitated adding Mg2+ in the absence and in the presence of BeuSOMs and LprSOM

100

6.1

Graphical representation of the measured parameters in the absence and in the presence of SOMs from N. pompilius and from P. lividus

110

6.2

Optical microscope pictures of crystal growing spaces in the absence and in the presence of SOMs from N. pompilius and from P. lividus

112

6.3

Optical microscope images of precipitates in the absence and in the presence of SOMs from N. pompilius and from P. lividus

113

6.4

SEM pictures showing the morphology of calcium carbonate crystals precipitated in the absence of SOMs

114

6.5

SEM pictures showing calcium carbonate crystals precipitated in the presence of entrapped NpoSOM

115

6.6

SEM pictures showing calcium carbonate crystals formed in the presence of entrapped PliSOM

117

6.7

XRD patterns of calcium carbonate precipitates in the absence and in the presence of SOMs from N. pompilius and from P. lividus

118

6.8

FTIR spectra of calcium carbonate precipitated adding Mg2+ into the cation reservoir in the absence and in the presence of NpoSOMs and PliSOM

118

6.9

Raman spectra of CaCO3 particles formed in the presence of diffusing Mg2+ in the absence and in the presence of NpoSOM and PliSOM

119

Amino acid composition and FTIR of NpoSOM and PliSOM

120

6.10

15

16

7.1

Jablonski diagram showing the stages involved in the process of fluorescence

127

7.2

The U-tube set-up used for experiments carried out with the counter-diffusion technique and fluorescence video frame showing a bottom view of the same tube

129

7.3

Absorbance and emission spectra of the pH sensor SF versus pH

130

7.4

Scheme of instrumental setup for fluorescence video acquisition

131

7.5

Dependence of fluorescence intensity of pH sensor SF

131

7.6

Contour maps showing the variations of pH along the U-tube versus time in the presence of SOM

132

7.7

Carbonate speciation as a function of pH showing the common pH range in nature

133

7.8

Fluorescence images at four different times in the absence and in the presence of SOM

134

7.9

Optical Microscopy images of calcite crystals at different times forming in absence and in presence of SOM

136

List of tables

1.1

3.1

PROPERTIES OF THE DIFFERENT CaCO 3 POLYMORPHS .................................................

36

SUMMARY OF RESULTS OBTAINED FROM EXPERIMENTS OF CaCO 3 PRECIPITATION IN THE ABSENCE OF POLYPEPTIDES AND IN THE PRESENCE OF pLys, pGlu OR pAsp ...........

65

CRYSTALLIZATION PARAMETERS MEASURED AND CHARACTERIZATION OF THE OBTAINED

4.1

PRECIPITATES IN EXPERIMENTS OF C a CO 3 PRECIPITATION USING CHARGED POLYPEPTIDES AND DIFFUSING Mg 2+ IONS ...........................................................................

77

17 SUMMARY OF DATA FROM PRECIPITATION EXPERIMENTS OF C a CO 3 IN THE ABSENCE

5.1

AND IN THE PRESENCE OF SOM FROM B. europaea OR L. pruvoti, ENTRAPPED IN AGAROSE VISCOUS SOL OR GEL AND IN THE PRESENCE OF Mg 2+ IN THE CATIONIC RESERVOIR ............................................................................................................

92

SUMMARY OF DATA OBTAINED FROM PRECIPITATION EXPERIMENTS OF C a CO 3 IN

6.1

THE ABSENCE AND PRESENCE OF SOM FROM N. pompilius OR P. lividus, AND IN THE ABSENCE AND PRESENCE OF Mg 2+ IN THE CATIONIC RESERVOIR .....................

111

01 Introduction and general objectives

1. Introduction and general objectives 1.1 Biomineralization

Over the last 3500 Myr or so, first prokaryotes and then eukaryotes developed the ability to form minerals. From that time, organisms from many different phyla evolved the ability to form the almost 70 different biogenic minerals known to date.1 Biominerals are composite materials made of a mineral and an organic component, the so-called organic matrix.

The most common minerals are calcium phosphate and calcium carbonate salts. They give structural support to, for example, endoskeletons in mammals and birds and exoskeletons or shells of many different organisms. Other examples include silica or elements rarely presented in organisms, such as iron and gold. The organic component of biominerals is composed of proteins, polysaccharides, glycoproteins, lipids, etc. In general, biominerals are used for many different purposes depending on its nature, i.e. protection, mechanical strength, cutting, ions storage, optical, gravity or magnetic receptor.

Biomineralization refers to the process by which organisms form biominerals. The formation of these hybrid organic-inorganic composites is a highly regulated process at different levels, from the nanoscopic to the macroscopic scale, and leads to complex and unusual morphologies endowed of physical properties in many cases unparalleled yet by their synthetic counterparts. This huge control over mineral nucleation, growth and organization of the mineral crystals, leading in most cases to hierarchical structures, is a source of inspiration for materials scientists and find applications in other disciplines as, for example, Biomedicine or Palaeontology. Interest in understanding and elucidating the mechanisms of biomineralization is not new but because of its potential applications in different fields, it has grown tremendously in the last years.

21

Chapter 1

1.1.1 Biomineralization processes Biomineralization processes can be divided according to their degree of biological control over mineral deposition. Lowenstam (1981)2 classified them as “biologically induced mineralization” and “organic-matrix-mediated mineralization”, popularized by Mann (1983)3 with the more appropriate term “biologically controlled mineralization”. The former correspond to “the processes that are not specifically designed for mineralization” and the later to “the processes in which a specific machinery is setting up for the purpose of biomineralization”.

One of the main goals of biomineralization research is to understand the control that organisms have developed over the creation of biominerals making them so different from their inorganic counterpart.

22

1.1.1.1 Biologically induced mineralization Precipitation of minerals is produced by the interactions between biological activity and environment (Fig. 1.1, upper-left). In biologically induced mineralization, a minor perturbation, such as metabolic by-products or cell surface alterations, is able to act as a causative agent for nucleation and mineral growth, without control over their shape or type. The characteristics of these processes are the following: (i) mineralization occurs in an open environment and not in a delineated space; (ii) precipitation occurs without macromolecular machinery; and (iii) it depends on the environmental conditions, i.e. same organism in different environmental conditions could form different minerals. The organisms that predominantly follow this kind of processes are archaea, bacteria and fungi.

1.1.1.2 Biologically controlled mineralization In biologically controlled mineralization, organism cells are in charge of nucleating, growing, and shaping as well as of locating the formed mineral. Almost all processes occur in an isolated space, but the degree of control in those environments varies from species to species. They can be classified, based upon the precipitation zone respect to which mineralization cells are responsible, in extra-, inter- or intracellular.4 Even if not all organisms can be categorized in one of them, understanding and identifying these simple processes leads to a better knowledge of the biomineralization mechanism.

Biologically controlled extracellular mineralization. In this case, the site of mineralization is located outside the cells, where a macromolecular framework has been previously produced and secreted by them. There are two types of biologically controlled extracellular mineraliza-

tion; i) cells secrete ions through the membrane (Fig. 1.1, lower-left-a) or ii) ions are concentrated in vesicles before leaving the cells (Fig. 1.1, lower-left-b). Mollusc shell nacreous layer, Scleractinian corals, bones or teeth are some examples of biologically controlled extracellular mineralization.

Biologically controlled intercellular mineralization. This biomineralization type occurs when the epithelial surfaces of cells are responsible for the nucleation and growth through a preferential orientation (Fig. 1.1, upper-right). It appears in unicellular organisms that form communities, such as some calcareous algae.

Biologically controlled intracellular mineralization. In this case, cells exert a high degree of control in the biomineral formation. Mineralization occurs inside vesicles within the cells, and their delimited space is responsible for final shape and composition of the biomineral.

Here, a single crystal unit nucleate and growth in a vesicle, which can assemble a structure before leaving the cell. This is the case of the Haptophyte algae (Fig. 1.1-lower-right-3a).

Otherwise, this single unit can leave the cell and interact with an extracellular organic matrix to form higher ordered architectures, such as those observed in some foraminifera (Fig. 1.1-lower-right-3b). In other cases, as occurs in magnetosomes-producing bacteria, the biomineral formed inside vesicles may remain within the cell (Fig. 1.1-lower-right-b).

Figure 1.1 Schematic of biologically induced mineralization (upper-left); biologically controlled extracellular mineralization (lower-left); schematic of biologically controlled intercellular biomineralization (upper-right); biologically controlled intracellular mineralization (lower-right). Modified from [4], with permission from Mineralogical Society of America.

23

Chapter 1

1.1.2 Organic Matrix

24

Biominerals mainly differ from their inorganic counterparts in their unusual morphologies and in the presence of an organic matrix. According to Krampitz et al.,5 the criteria to be considered a real matrix are “existence of the matrix before formation of the mineral, participation of the matrix in its formation, and spatial inclusion of the mineral by the matrix”.

Depending on its localization, this organic matrix could be considered as intercrystalline, which lies between two adjacent single crystals or intracrystalline, which lies inside the single crystals.6 As we have shown above, organisms use organic macromolecules to control the formation of the mineral itself. The strategies in play to achieve this objective comprise: confining a space, forming an organic matrix framework, controlling ion input, constructing a nucleation site, controlling crystal orientation, growth and aggregation and terminating crystal growth.7 The content of the organic matrix depend of the type of biomineral and contains mostly proteins, with high contents of acid aminoacids, and glycosaminoglycans.8 Also sulphated polysaccharides have been found in biominerals9 and more recently, lipids10 that may be involved in the ionic trafficking.11 As it is difficult to observe biomineral formation in vivo, the organic component is usually extracted and used it to perform crystallization experiments in vitro. The aim is to identify and characterize possible roles of those macromolecules in the formation of the biomineral.

Extraction procedures usually divide the organic matrix into two components: a soluble one, named soluble organic matrix (SOM), which is considered to act as functional macromolecules that “control” the mineralization process, and the insoluble component or insoluble organic matrix, which act as framework macromolecules that provide a structure in which the mineral forms.

SOM is a mixture of proteins, polysaccharides and lipids. These proteins are rich in acid aminoacids, overall aspartic and glutamic residues.12 Besides, some of these proteins may have phosphorylated serine and threonine residues and frequently contain covalently bound sulphated polysaccharides. Glycosylation, when a polysaccharide is joined to polypeptide sequences, is a common phenomenon and may provide sites of high anionic charge. Since the first experimental evidences of SOM,13 researches have been trying to characterize and isolate the different SOM components and establish their possible role in mineral formation. This is the case of some mollusc shells that secret a SOM that is responsible for the polymorph selection.14 Even though the characterization of SOM has been focused on some isolated proteins, only a few of them have been identified so far. For example PIF, an aspartic-rich protein which is essential for the formation of aragonite in mollusc shell nacre,15 or a negatively-charged amino acid protein encoded in a starmaker gene, which is necessary for aragonite polymorph selection in the inner ears of zebrafish.16 The Laboratory of Crystal-

lographic Studies in Granada (Spain) provides another example that showed that a group of acidic proteins are essential for the formation of the eggshell by hens. Thus, when using uterine fluids from hens for precipitating CaCO3, it was observed that only a few of the total proteins contained in such a fluid (ovocleidin-17, ovocalyxin-32, osteopontin and ovocleidin-116), showed a strong affinity for CaCO3 surfaces being removed from the fluid during the precipitation process. The finding suggested that these proteins might have an active role on CaCO3 growth, aggregation, and inhibition. Other proteins, as the couple ovocleidin-17 and ovocalyxin-21, with very different isoelectric points, were responsible of buffering the pH of the environment, thus favouring crystal growth.17 In contrast to SOM, framework macromolecules are quite hydrophobic and often crosslinked. One of the most studied insoluble matrices is the one found in the mollusc shell nacre, where ß-chitin is the most ordered component observed by X-ray diffraction and silk-like fibroin, the most abundant.18 More recently, silk-like fibroin was suggested to present gel-like properties.19 Interestingly, by using “proto-polyp” cells from Stylophora pistillata corals it was demonstrated that the precipitation of aragonite could occur in a gelling environment.20 These are a few examples of the quite new tendency of thinking that the formation of a biomineral might occur in a gelling environment.

1.2 Crystallization In this text we are going to differentiate between “classical” and what is called “non-classical crystallization”. “Classical” pathway involves the attachment of molecules or ions forming a nucleus that then grows to form a crystal. “Non-classical” pathway involves processes of particle self-assembly (oriented attachment or mesocrystals) or also precursor transitions with a precursor (metastable or liquid), which will be eventually transformed in a more ordered structure.21

1.2.1 Classical crystallization Classical crystallization has been the subject of hundreds of studies (see for instance [22]) and although it is not the specific subject of inquiry in this dissertation we believe that is convenient to introduce some relevant concepts. The main character of crystallization must be introduced, that is the supersaturation ratio. Such a quantity will play a crucial role when it comes to form crystals. Its definition is just the ratio between the actual ionic activity product and the solubility product. When that quotient is less than one the system is said to

25

Chapter 1

be undersaturated, in case to equal one the system is in equilibrium, otherwise it will be supersaturated. In order to grow crystals the system must be supersaturated and the logarithm of “S� will be the driving force.22a A typical solubility diagram for crystallization in solution (Fig. 1.2) shows the solubility curve and the supersolubility curve. The region in between is called metastable zone. In a given crystallization system, once the supersaturation limit is exceeded, it is possible to spontaneously create nuclei of the new phase. Consequently, these nuclei grow in the metastable zone accompanied of a decrease of supersaturation up to it equals one, that is, to the point at which the actual ionic activity product equals the solubility product.

26

Once the system is set at a given supersaturation value the molecules will start to aggregate.

The higher the supersaturation is, the faster these molecular clusters will be formed and the lower the energy required to form them will become. When the aggregates (or simply clusters) reach a critical size, they are called nuclei. Then, the nucleation process has a 50% chance to take place and, hence, to start the crystal growth. However, it is highly probable that molecules cannot be ordered whilst they are clustering when the supersaturation is large enough and so, an amorphous phase will appear.

Figure 1.2 Solubility phase diagram. Concentration versus X. X is any parameter that influences solubility, such as pH or temperature. The solubility curve is marked with a solid line. Above this, the metastable zone is the zone where crystals can grow but rarely nucleate (in green colour). Spontaneous nucleation occurs when the supersaturation exceed the supersolubility curve (dashed line) in zone of yellow colour. Amorphous precipitation occurs at very high supersaturations where particles formed disordered aggregates (red colour). Reproduced from [23].

As previously mentioned, the driving force can be expressed in terms of the supersaturation ratio. Actually, the driving force is just the difference between the chemical potential of the supersaturated solution and that of the final equilibrium solid (or solution).

∆μ = kB T ln

a = kB T lnS aeq

(eq.1.1)

where kB is the Boltzmann constant, T is the absolute temperature, a is the activity of the solute in the supersaturated solution and aeq is its activity at equilibrium.

Nucleation is the stage covering the formation of nuclei or clusters of structural particles (atoms, ions, molecules), which act as seeds for the formation of new crystals. Nucleation can be homogeneous or heterogeneous. Homogenous nucleation occurs when the cluster forms without influence of foreign particles or without contact with a surface. When nucleation is influenced by surfaces or dispersed components, such as seeds or dust, the process is called heterogeneous nucleation.

The creation of particles of a new phase is associated to a change in the Gibbs free energy.

This change is caused by two factors: the volume term (∆GV) and the surface term (∆GS), such as:

∆G=∆GV+ ∆GS

(eq.1.2)

The volume term (eq. 1.3) is negative since the new phase is the stable one having associated thus a lower energy than the initial metastable state, whereas the surface term (eq. 1.4) is positive because of the creation of a new surface, hindering the formation of a new phase.

For a spherical cluster of radius r, both components can be expressed as:

(

∆GS = 4πr2 σ

4πr3 3Ω

(

∆GV =

∆μ

(eq.1.3)

(eq.1.4)

27

Chapter 1

where Ω is the molecular volume and σ, the surface tension. Considering thermodynamic equilibrium, equation 1.4 allows us to determine the critical radius, r*:

28

∂∆G ∂r

(

(

= 0 → r * =

2Ωσ ∆μ

(eq.1.5)

The critical cluster (with a critical size) is the one that defines the border between the growth and the shrink zones for clusters. If the radius is smaller than the critical one (r < r*), the cluster will be unstable and will dissolve because its growth implicates an increase of ∆G. However, if the radius is larger (r > r*), the growth of the cluster will continue (Fig. 1.3).

Figure 1.3 Free energy for homogeneous nucleation based on cluster size.

Reproduced from [24], with permission from John Wiley and Sons.

Theoretical studies of homogenous nucleation22b, 22c showed the importance of the kinetics in the nucleation phenomena. On the basis of the classical nucleation theory (CNT) and simplifying such studies, the nucleation rate, J, can be defined as the number of clusters formed by per unit time and per unit volume in a supersaturated system:

J = N* S* Ď&#x2026;A

(eq.1.6)

where N* is the number of critical clusters per unit of volume in the solution, S* is the surface of the critical cluster, and Ď&#x2026;A the rate of molecule attachment to the critical cluster per unit of surface.

Additives can completely hinder the formation of clusters. A few amount of additive molecules can be adsorbed on the surface of the new cluster, reducing the nucleus surface available for further growing units, and thus lowering the rate of molecule attachment per unit of surface. Therefore, the nucleation rate will decrease. The final effect of additives on the nucleation rate is determined by the competition between the molecule attachment rate to the cluster surface and the thermodynamic term. One of the common ways to study the nucleation kinetics is through the measurement of the induction time by several techniques: visual observation, turbidity, conductivity, pH, etc. The induction time is the time that elapses from the creation of the supersaturation in a solution till the formation of first detectable crystals. Since measuring the induction time depends on the employed technique, the waiting time (tw) is usually used instead. The waiting time reports for the period elapsing from the set up of the experiment up to detection of first formed crystals.

The second stage of crystallization is crystal growth. After the nucleation stage, nuclei grow to become a macroscopic crystal. Shape of crystals is related to the unit cell structure, but sometimes a mineral has altered morphologies because of different rates of growth in its diverse faces. How fast a face growth usually depends on its surface energy.

That is, faster growing faces possess high surface energy, making them grow faster than the others and therefore, they will vanish in the final shape of the crystal.25 Therefore the type of faces in a crystal is critical for the growth rate. Faces can be divided into three types:26 flat faces (F), stepped faces (S), or kinked faces (K) (Fig. 1.4). Relating to the attachment of new growth units, the difference lies in the number of bonds that can form when a new unit attach to the crystal.

29

Chapter 1

Figure 1.4 Classification of the types of faces in the crystal. F, refers to flat faces; S, to stepped faces; and K, kinked faces. From [23].

30

The growth mechanisms of a flat surface proceed in two stages. First, building units transfer from solution to crystal surface; and second, they incorporate into the crystal lattice (Fig, 1.5). The later includes: adsorption to the surface, surface diffusion, step diffusion and incorporation into the crystal lattice.27 The slowest step will determine the crystal growth rate.

Figure 1.5 Schematic illustration of the Terrace-Ledge-Kink model that represents different stages of the crystal growth in solution.

Growth unit adsorbs to the surface or directly attach to a kink position. After being adsorb to the surface, it migrates across the surface up to the step and diffuse through the step, where it locates a kink site. Then, it incorporates to the crystal lattice or desorbs from the surface.

Reproduced from [23].

There are usually three mechanisms of crystal growth: continuous growth, surface nucleation and spiral growth. A crystal with rough faces (K), allows units to incorporate directly, without need of surface diffusion step. This type of growth mechanism is called continuous growth. On the other hand, if crystals present flat faces (F) the dominant growth mechanism is called surface nucleation. When screw dislocations are present due to defects on the surface, units incorporate forming a spiral-staircase, being spiral growth the dominant mechanism.

1.2.2 Non-classical crystallization In contrast to the classical way, the non-classical crystallization process is characterized because the decrease of Gibbs free energy of the system is not produced by growth but for ordered aggregation of individual crystals that fuse together to form a single crystal. Two particles became closer to each other attracted by van der Walls forces, eventually this approach could lead to a direct contact and therefore allow the particles to fuse together “exterminating” two surface energies (Fig. 1.6), a process called oriented attachment.

31

Figure 1.6 Diagram of classical (a) and non-classical crystallization: (b) oriented attachment, (c) mesocrystal formation, and (d) amorphous particles or liquid precursors. Reproduced from [28], with permission from Springer.

Oriented attachment was first observed in TiO2 particles in hydrothermal conditions29 (Fig. 1.6-b) where individual units of TiO2 crystallographically aligned. The term mesocrystal is the abbreviation of “mesoscopically structured crystal” and is defined as “colloidal crystal that is built up from individual nanocrystals and are aligned in a common crystallographic register”.24 The mesocrystal pathway involves temporary stabilization of individual crystal units. Then, these units become one by the mechanism of oriented attachment and diffract

Chapter 1

as a single crystal. Sea urchin spines and nacre are in focus on being mesocrystals (Fig. 1.6-c).30 The amorphous precursor pathway is a great challenge of biomineral formation, since it can be sculpted as an artwork to create such a complex and amazing morphologies. Besides, it could be useful as a mineral store, keeping particles at high concentration in the mineralization place (Fig. 1.6-d). In the same way, liquid precursors, also named “polymer-induced liquid precursors (PILP)”,31 are important candidates for the production of complex morphologies. Liquid precursors appear as liquid-liquid separation from a saturated solution and are considered to be a prior-phase to an amorphous solid phase due to its higher labile properties (Fig. 1.6-d).

The PILP pathway was first shown in vitro using CaCO3.

32

1.2.3 Gel Crystallization Gels suppress convection and sedimentation, creating a diffusive scenario for the transport of reactant species during a crystallization process.32 In a gelling media the ionic transport is controlled mainly by diffusion according to Fick’s First Law.33 Fick’s First Law says that the diffusive flux (J) through a surface is directly proportional to the concentration gradient:

J = D

∂c ∂x

(eq.1.11)

where J is the flux of ions or molecules across a gel section (mol/m2.s); ∂c/dx is the concentration gradient (mol/m3) in the normal direction of a section (m) and D is the diffusion coefficient (m2/s). If the length of the diffusion path is long enough, the reactants flux is variable with time. Therefore, concentrations of reactants are variable not only with the position in the column, x, but also with time. Then, deriving from Fick’s First Law and the mass conservation Law, Fick’s Second Law is obtained:

∂c ∂2c = D ∂x ∂x2

(eq.1.12)

The solution of Fickâ&#x20AC;&#x2122;s Second Law depends on the initial and boundary conditions. For example, assuming that: (i) the diffusion coefficient, D, is constant (independent on solution concentrations or time); (ii) the diffusion medium is one-dimension (ions diffuse only in x-direction); (iii) the length of the gel column is semi-infinite (for all time of the course of the experiment); (iv) the concentration of reactants in the reservoirs are equal to the initial concentration for all time during the experiment and, (v) the concentration of reactants in the column gel is zero at the start of the experiment; the solution of the Fickâ&#x20AC;&#x2122;s Second Law based on an error function is obtained (e.g. [34]):

c = co erf c

x 2 Dt

(eq.1.13) 33

Generally, the use of gels as crystallization media has been widespread in many different fields ranging from optimization of crystal size and crystal quality,32a, 35 simulation of microgravity environments,36 analogous of geological environments,37 production of biomimetic materials38 and to simulate biomineralization environments.39 All these studies allow the controlled crystallization of a plethora of minerals, among them calcium phosphates39c, 39e, 40 and calcium carbonates38a, 39a, 39b along with the study of the effects of inorganic and organic additives on these crystallization processes.35b, 41 Using gels, we can control the pore volume distribution by varying the entanglement parameters, such as pH in silica gels42 or the concentration of ions43 and macromolecules44 in polysaccharide gels. Concretely, in this thesis we have chosen agarose as an ion diffusion and crystallization medium, mainly because the structure of an agarose gel is almost independent of the pH and the presence of foreign ions and it is a natural polysaccharide widely found in nature (-- see section 2.2). Besides, agarose forms gels with different strengths depending on its concentration.44 Thus, making a perfect candidate to explore the influence of the degree of entanglement by using two different concentrations, 0.2 % a proper gel, or 0.1 %, viscous sol (or partial gel).

1.2.4 Counter-Diffusion System (CDS) In order to simulate the biomineralization processes in a scenario with physical and ion-transport characteristics closer to that occurring in vivo, counter-diffusion in gels, also called double-diffusion, has emerged as a powerful platform to perform experiments in vitro.39c-e

Chapter 1

Among different tools to perform counter-diffusion experiments the simplest one is the U-tube (Fig. 1.7). It contains a gel column of length L and two reservoirs containing the reactant solutions, situated at both ends. These two solutions diffuse against each other through the gel, generating a continuous gradient of concentrations of both reagents in the gel column, and therefore a continuous variation in space and time of the ionic activity product and supersaturation. Thus, precipitation will occur at the time and location of the gel column where the critical supersaturation value for nucleation is reached.45 Under the counter-diffusion configuration the supersaturation threshold to trigger nucleation is a function of the rate of development of supersaturation and, at the same time, the equivalence rule must be satisfied in addition to reaching the supersaturation threshold.46

34

Figure 1.7 Picture of the U-tube where two solutions counter-diffuse through an agarose gel column containing entrapped additives and yielding a precipitate.

1.3 Calcium carbonate, CaCO3 1.3.1 CaCO3 Phases The precipitation of CaCO3 is of great interest not only in the biomineralization field but also in geochemistry (it is one of the most widely existing minerals in nature) and chemical engineering, showing a variety of industrial applications. It is employed as a pigment in paints and filler in plastics, rubber, paper, food and pharmaceutical industry or in the plastisol industry for car under-body paints, among many others applications.47

CaCO3 has three crystalline polymorphs and a metastable form that are amorphous calcium carbonate (ACC), vaterite, aragonite and calcite (table 1.1).

Calcite is the most stable form at ambient temperature and atmospheric pressure. Aragonite is less stable than calcite and then, less common in nature. Calcite has a rhombohedral structure, where calcium ions and planar carbonate groups form alternate layers and are orientated perpendicular to the c-axis. Each calcium ion is surrounding by six carbonate groups, in which one oxygen of each group is the first neighbour of calcium.7 On the other hand, aragonite has orthorhombic structure (Fig. 1.8). As in calcite, calcium ions and carbonate groups form alternating layers, but carbonate layers are parallel to the a-axis. Besides, the coordination number of calcium is nine, being an oxygen atom the closest neighbour of calcium.

35

Figure 1.8 Schematic representation of the calcite (left) and aragonite (right) structure. Calcium ions are in yellow colour, and in red and blue the oxygen and carbon atoms that form the carbonate ions. Reproduced from [48], with permission from Royal Society of Chemistry.

Amorphous calcium carbonate (ACC) is highly soluble, has a low density and rapidly transform to more stable forms such as calcite or vaterite when it is not stabilized. The known stable forms of ACC are hydrated; however, transient forms of ACC do not contain water.49 ACC has been proposed as the precursor of CaCO3 biominerals since it has been identified in different phyla: echinoderms;50 molluscs;51 crustaceans,52 chordates,53 annelids54 or porifera.55

Chapter 1

Table 1.1 Properties of the different CaCO 3 phases.

Reproduced from [ 56].

PROPERTY

36

CALCITE

ARAGONITE

VATERITE

ACC

Formula

CaCO3

CaCO3

CaCO3

CaCO3nH2O 0<n>1

Solubility Product 25ºC (mol/L)

10-8.48

10-8.34

10-7.91

10-6.40

Space Group

R3 c

Pmcn

P63/mmc

-----

Point Group

3 2/m

2/m 2/m 2/m

6/m 2/m 2/m

-----

Crystal System

Trigonal

Orthorhombic

Hexagonal

-----

Lattice Constants (Å)

a = b = 4.99 c = 17.062 γ = 120º

a= 4.959 b= 7.964 c= 5.738 a= β= γ =90º

a = b = 4.13 c = 8.490 γ = 120º

-----

Density (g/cm3)

2.71

2.93

2.65

1.6

Abundance

V. common

common

rare

common

1.3.2 Carbon Cycle CaCO3 biominerals play an important role in the carbon cycle. When CO2 dissolves in seawater, autotroph organisms (such as phytoplankton) fix and produce organic molecules from inorganic carbon. This organic form of carbon then enters the marine food chain.

Alternatively, marine organisms such as corals, crustacean, coccoliths or echinoderms, form their skeletons or shells transforming dissolved CO2 in CaCO3. When these organisms die, their deposits fall to the bottom to form part of the seabed.

Figure 1.9 Species concentration of the three inorganic forms of dissolved CO2 in water. The green arrow delimits the pH range of seawater, of 7.6 to 8.2. Reprinted from [57] with permission from AAAS.

Dissolved CO2 in seawater is also acting as a buffer by keeping the pH between 7.6 and 8.2.

When CO2 dissolves in seawater, produces protons (H+) and changes the equilibrium among the negatively charged carbonate species (Fig. 1.9). Initially CO2 reacts with water to form carbonic acid, H2CO3 (eq. 1.7) that then dissociates into one bicarbonate ion (HCO3-) and one H+ (eq. 1.8), thus, decreasing the pH. Some of released H+ would react with carbonate molecules (CO32-) to form HCO3- (eq. 1.9). How efficient is the carbonate buffer has great relevance in formation and dissolution of the CaCO3 used by organisms to form their skeletons and shells (eq. 1.10). As the CaCO3 solubility increases when temperature decreases and pressure increases, there is a threshold point below which CaCO3 begins to dissolve in the sea floor (Fig. 1.10). Nowadays, as the partial pressure of CO2 in atmosphere is increasing due to the anthropic activities, the final CO32- concentration dissolved in seawater is decreasing. Thus, it will inhibit the formation of mineral CaCO3, promoting its dissolution

37

Chapter 1

and making that the CaCO3 deposits from death organisms begin to dissolve at shallower seafloor regions.

[CO2 ] + [H2 O] [H2 CO3 ]

[H2 CO3 ]

[H+ ] + [HCO3- ]

[H+ ] + [CO32- ]

[HCO3- ]

pK1 (25 °C) = 6.35

(eq.1.7)

(eq.1.8)

pK2 (25 °C) = 10.33

(eq.1.9)

(eq.1.10)

Mineral formation →

38

[Ca2+ ] + [CO32-]

[CaCO3 ]

← Mineral dissolution

Figure 1.10 Diagram of carbon exchange between atmosphere and oceans. There are three forms of inorganic carbon in seawater, aqueous CO2, bicarbonate (HCO3-) and carbonate (CO32-). Dissolved CO2 transform into organic carbon by autotroph organisms and into CaCO3 by biomineralization organisms. On the other hand, calcification and respiration release CO2 into the oceans, which part of them may be coming back to the atmosphere. The carbonate saturation horizon indicates the point in which the critical CO32concentration below which CaCO3 begin to dissolve. Reproduced from [58].

1.4 Additives Precipitation processes are strongly affected by the presence of additives, which cannot only promote, slow down or inhibit nucleation and growth, but also affect the selective formation of a polymorph or induce a morphological change.7, 59 An evaluation of their effects on nucleation and growth processes is of key importance for a wide range of applications. Particularly, in the case of calcium carbonate (CaCO3), for their use as anti-scaling agents60 or as molecular models for understanding biomineralization processes.1a Additives action can be exerted in the stage of nucleation or growth, or in both stages.

1.4.1 Magnesium ions, Mg2+ Magnesium ions are one of the major constituents of the seawater; in fact the Mg2+/Ca2+ in oceans is around 5. They play a key role in carbonate modification,61 in cellular processes,62 and Mg2+ play a role in the stability of carbonate-mineral formation.63 The role(s) of Mg2+ in biomineralization processes has been ever since studied in depth. In fact it is still being studied.

Mg2+ inhibit the crystal growth of inorganic calcite and provoke the precipitation of almost completely aragonite when the concentration ratio of Mg2+/Ca2+ is above 4.64 The denser structure of aragonite blocks the entrance of Mg2+ in their cation positions. On the other hand, Mg2+ may enter in the calcite structure, but their hydration sphere could inhibit further growth of calcite, and then favour the precipitation of aragonite. It is also known that Mg2+ stabilize ACC in the presence of some additives.49, 51a Interestingly, the composition of biogenic calcium-magnesium carbonate is though to occur in a complete spectrum of compositions between calcite and dolomite. For example, a maximum of 40% MgCO3 was found in echinoids65 whereas, incorporating Mg2+ to calcite in vitro at physiological conditions has not been an easy task (e.g. 14% of Mg2+ content in calcite has been reported by using alcohols).66 In a biological system, Mg2+ could be dehydrated or partially dehydrated before adsorption to the crystal surface. This might be conceivably achieved through chelating Mg2+ by some macromolecules and then, partially dehydrated Mg2+ would enter into the crystal lattice.

39

Chapter 1

1.4.2 Coral biomineralization

40

Coral reefs are probably the most spectacular biomineralization product on Earth. These ecosystems are built from accumulated skeletal fragments of innumerable generations of organisms, such as symbiotic foraminifera or calcareous algae, but overall coral. These reefs represent some of the biggest biodiversity ecosystems in our world. Of particular interest are the Scleractinian coral as they are the biggest source of biogenic CaCO367 and are among the fastest marine mineralizing organisms.68 There are two groups of Scleractinia coral: colonial and solitary. The former are found in clear, shallow tropical water and are primary reef-builders; and the latter are found in all regions of the oceans and do not build reefs.

Despite their great contribution to oceanic biomineralization (see review [69] and references therein) and their important role in the carbon cycle,70 many aspects of coral mineralization mechanisms are still a source of discussion.

A coral is a polyp (or colony of polyps) protected by a CaCO3 skeleton. Polyps look like a bag whose walls are made by two layer of cells, an ectodermal layer, and an endodermal layer, separated by a network of collagen, called mesoglea (Fig. 1.11). Polyps may be symbiotic or asymbiotic. In symbiotic corals, polyps host unicellular symbiotic photosynthetic algae, called zooxhantellae, that provide energetic support to the coral and are thought to facilitate the mineralization process. Concretely, the photosynthetic products, such as polysaccharides, lipids or oxygen, could have a role: as precursors for the organic matrix in the skeleton formation; in the carbon cycle; increasing cell metabolism and thus accelerating the calcification process; or removing the substances that would otherwise inhibit calcification. Still, the mechanisms proposed to explain the relationship between zooxanthellae and calcification remains unclear.

Figure 1.11 Diagram of a polyp sits on the top of its skeletons. Reproduced from National Ocean Service (NOAA).71

41

Each polyp sits atop of the skeletal surface and produces a skeleton of CaCO3 crystals in the form of aragonite and an organic component.8a, 72 Scleractinia skeletons (Fig. 1.12) are composed of groups of needle-like aragonite crystals that radiate out from the centre of calcification,72 a region rich in sulphur.9c, 73 This structural organization is controlled by specific macromolecules and is only slightly affected by external environmental parameters. But still since 1882, when von Koch74 demonstrated that extracellular calcification occurs outside the ectodermal cell layer, the discussion of how much the calcification control is exerted by biological or by environmental conditions has remained open. In particular, corals mineralize at the interface between the polypâ&#x20AC;&#x2122;s calicodermic tissue and the skeleton. This region is extremely rich in glycoproteins and glycosaminoglycans able to bond water molecules, thus the coral mineralization site was suggested to have gelling features.75 An amorphous organic membrane was observed between the calicodermis and the skeleton and it was postulated that the site of mineralization is a colloidal gel matrix.76

Chapter 1

42 Figure 1.12 Underwater in situ camera pictures of B. europaea (Beu1) and L. pruvoti (Lpr1). SEM cross-section images of the skeleton of B. europaea (Beu2, 3) and L. pruvoti (Lpr2, 3) are also shown. In them the different macroscale organization of the septa and microscale organization of the aragonitic fibers and of the centers of calcification is evident. The arrows indicate centers of calcification surrounded by aragonite fibers.

1.4.3 Nacre biomineralization Nautilus pompilius (Npo) is a cephalopod mollusc and is the most widespread specie of all Nautiluses. Nautiluses are considered as living fossils because they have remained almost unchanged for million years. They live in the Pacific and Indic oceans and have been observed until 500 meters deeper so far. Nautilus shell presents a logarithmic spiral divided into chambers, which are separated by the septa. When the animal grows up, it creates a new, bigger chamber and moves into a larger space, closing the old chamber with a new septum. A tube crosses all the septa by the middle and produces a gas that remains in each chamber that forms the shell, acting as a buoyancy compensator.

Among the different types of mollusc shell structures, nacre is the best studied probably due to its tricky geometry. Its layered structure was already observed in the middle of XIX century77 and the presence of an organic component was established almost 100 years after.78 The nacreous layer of the mollusc Nautilus pompilius (Fig. 1.12) is a representative example of a biologically controlled extracellular biomineralization. First an organic matrix framework is designed, which acts as a site for nucleation and growth of single nano-crystals. Then, the assembly of crystals in an ordered array is guided by the structure of the preformed frame-

work.79 Precisely, nacre is formed by a columnar brick and mortar structure.79-80 Aragonite tablets (bricks) are separated by interlamellar sheets of protein- β -chitin complex (mortar)6, 81 (Fig. 1.13). Each tablet is formed by assembly of nanoparticles that diffract X-ray as a twin single crystal of aragonite. This tablet grows parallel to the organic- β -sheet and perpendicular until achieving the consecutive β -sheet layer. Between two β -sheet layers, crystals grow embedded in a silk fibroin-like gel matrix.19 Figure 1.13 Composite image illustrating a site view of the organism (inset) and a SEM image of the fractured shell of N. pompilius. We can observe the typical organization of tablets in the nacre structure.

The image of the organism was modified from wikicommons.

1.4.4 Sea urchin spines biomineralization Paracentrotus lividus (Pli) is an echinoidea species belonging to the Paraechinidae family. It usually lives in the Mediterranean and eastern Atlantic oceans, between 10 and 20 meters depth in rock pools. The name echinoidea means “like an urchin” and that is because spines cover these animals. Sharply spines of Pli are usually purple and that is why is commonly known as purple sea urchin. The body of these animals is the result of calcareous plates melting into a solid box. It has a radial symmetry, in which five equally portions radiate out from a central axe. This characteristic shell can be divided into two halves: the lower or oral and the upper or aboral hemisphere. The former contains the mouth and is in contact with the substratum. The latter is the anal region, called periproct, and contains the madreporite, which is used to filter water into the water vascular system. A ciliated epidermis covers the outer surface, including the spines. Below that, a connective tissue containing the skeleton appears.

Sea urchin spines are a representative example of biologically controlled intracellular mineralization. Mineral forms in vesicles where nanocrystals self-assemble in a tri-dimensional

43

Chapter 1

way to finally generate single crystals with complex and usually curved surfaces.82 Spines are composed of elongated nanocrystals of Mg-calcite, showed a glassy cleavage (Fig. 1.14) and diffract X-ray as single crystals.78, 82-83 Cells called calcoblasts are associated with the growing surface of the spine, and are responsible for the mineral growth occurring on the top of the spine. Spine growth was first studied by regenerating a spine.84 Then, ACC was observed to form in vesicles, deposited on the tip of the spine and, gradually transformed into calcite.50b

44

Figure 1.14 Composite image illustrating a site view of the organism (inset) and a SEM image of the spine P. lividus. We can observe the glassy cleavage of a single crystal sea urchin spine. The image of the organism was modified from wikicommons.

1.5 Objectives of the Thesis and chapters summary Simulating biomineralization processes in vitro help us to deeply understand the strategies employed by organisms to mineralize their skeletons or to deposit minerals for different purposes in the environment where they live. There are many evidences that among these strategies organisms secrete an organic matrix to influence the precipitation of the mineral; however its role still remains unknown in many aspects, e.g. its influence on the selection of the mineral polymorph, on the inhibition or promotion of the nucleation or growth, or on how this organic matrix influences the hierarchical organization of the mineral crystals.

This PhD Thesis is addressed to investigate the influences of environmental (e.g. pH and Mg2+ ions) and biological factors (SOM extracted from different organisms skeletons) on the precipitation of CaCO3, as a tool to better understand these strategies. In this respect several objectives have to be consecutively achieved:

1. Set-up of an in vitro valid and worthwhile protocol for studying biocrystallization by mimicking as much as possible the biomineralization processes occurring in vivo. This protocol will be based on the counter-diffusion technique and will be implemented in U-tube devices to precipitate CaCO3 in gelling environments. In it we will define different measurable parameters, related to crystallization, which will allow us to compare the influence of the above factors.

2. Test of this protocol to study the influence of: a) charged polypeptides simulating acid and basic proteins, (b) diffusing Mg2+ and entrapped polypeptides, (c) two soluble organic matrices extracted from two different corals, diffusing Mg2+ and degree of entanglement; (d) two soluble organic matrices extracted from nacre and sea urchin spines, and diffusing Mg2+.

3. To explore the use of fluorescence sensors in counter-diffusion system to analyse the pH-evolution accompanying ion diffusion and CaCO3 precipitation.

45

Chapter 1

Beside this introduction the memory is organized in the following chapters:

Chapter 2 describes general experimental work and theoretical considerations concerning the protocol based on a counter-diffusion system. Experimental work shows how we extracted SOMs from Scleractinia corals, Nautilus pompilius and Paracentrotus lividus; the preparation of agarose viscous sols and gels; the U-tube crystallization set-up; the morphologic and polymorphic characterization techniques.

The theoretical considerations section defines and interprets the measurable parameters, which are used in the following chapters.

46

Chapter 3 reports the influence of charged polypeptides, poly-L-aspartate (pAsp), poly-L-glutamate (pGlu) and poly-L-lysine (pLys) on the precipitation of CaCO3. We evaluate their inhibiting or promoting capability over nucleation, growth, morphology and polymorphism of CaCO3.

Chapter 4 explores the role of diffusing Mg2+ in combination with entrapped polypeptides pAsp, pGlu and pLys in the gelling media on CaCO3 precipitation.

We study how different Mg2+ concentrations influence the measurable parameters, polymorphism and morphology of resulting crystals.

Chapter 5 explores the effects of intra-skeletal SOMs extracted from two Scleractinian corals differing in the presence of symbiotic photosynthetic algae, Balanophyllia europaea and Leptopsammia pruvoti, on CaCO3 precipitation in agarose gelling environments. In this chapter we also examine the influence of both Mg2+ and degree of entanglement of the gelling media.

Chapter 6 investigates the influence on CaCO3 precipitation of SOMs extracted from two organisms of different phyla, Nautilus pompilius (mollusca phylum) and Paracentrotus lividus (echinodermata phylum) also in combination with diffusing Mg2+.

Chapter 7 shows the preliminary results of a study aimed to reveal the pH evolution accompanying ion diffusion and precipitation of CaCO3 in a counter-diffusion system.

In it we have developed an instrumental set-up combining counter-diffusion and optical fluorescence.

02 Experimental and theoretical approach

2. Experimental and theoretical approach 2.1 Organic additives

We have used two kinds of organic additives, synthetic polypeptides as model macromolecules of both acidic and basic proteins and soluble organic matrix (SOM) extracted from different organisms. Organic components were extracted from three different phyla: two scleractinian corals, namely Balanophyllia europaea (zooxanthellated) and Leptopsammia pruvoti (azooxanthellated), which belong to Cnidaria phylum; Nautilus Pompilius, which belongs to Mollusca phylum; and Paracentrotus lividus, which belongs to Echinodermata phylum. Synthetic polypeptides were poly-L-aspartic acid (pAsp), poly-L-glutamic acid (pGlu) and poly-L-lysine (pLys).

2.1.1 Extraction of the soluble organic matrix (SOM) Fragments of the coral skeletons (see further details of corals collection in chapter 5 –section 5.2), fragments of the septa of Nautilus pompilius (Npo) and spines of Paracentrotus lividus (Pli) were suspended (1% w/v) in a sodium hypochlorite solution (3% v/v) to remove traces of organic material eventually not removed by the first treatment (in the case of coral skeletons) or for the possible organic material traces from the body of organisms (in the other two cases). The samples were air dried for one night and ground in a mortar up to obtain a fine and homogeneous powder. Five mL of milli-Q water, in which 2.5-3 g approx. of powdered biominerals were dispersed, were poured into a 40 cm-long osmotic tube for dialysis (MWCO = 3.5 kDa; CelluSep®, MFPI). The sealed tube was placed into 1 L of 0.1M CH3COOH (Riedel de Haen) solution under stirring. The decalcification proceeded for 72 hr. At the end, the tube containing the dissolved OM was dialysed against milli-Q water (resistivity 18.2 MΩ cm at 25 °C; filtered through a 0.22 μm membrane) until the final pH was about 6. The obtained aqueous solution containing the OM was centrifuged at 6000 rpm for 6 minutes to separate the soluble and the insoluble organic matrix fractions, which were then lyophilized.

49

Chapter 2

2.1.2 Synthetic polypeptides

50

The charged polypeptides used were pAsp, poly-L-aspartic acid as a sodium salt (MW 9600 Da; Sigma-Aldrich); pGlu, poly-L-glutamic acid as a sodium salt (MW 9600 Da; Sigma-Aldrich); and pLys, poly-L-lysine hydrochloride (15000 Da; Sigma-Aldrich). The pAsp and pGlu polyelectrolytes have anionic charge due to the presence of carboxylic groups in their side chains. On the other hand, pLys is a polyelectrolyte with cationic character due to the presence of an extra amino group in its side chain (Fig. 2.1). The isoelectronic point, pI, for the aspartic acid aminoacid is 2.77; for the glutamic acid aminoacid is 3.22 and for lysine aminoacid is 9.74. The pKa for the side chain groups are 3.65 and 4.25 for the carboxyl group of aspartic acid and glutamic acid aminoacids respectively, and 10.53 for the ammonium group of lysine aminoacid.85 Both, anionic and cationic polyelectrolytes will be charge at the experimental pH of agarose (around 6 â&#x20AC;&#x201C; 6.5).

Figure 2.1 Structures of pAsp (left), pGlu (medium) and pLys (right).

Reproduced from [86], with permission from ACS Publications.

2.2 Agarose We use agarose as an ion diffusion and crystallization medium. Most of all, we chose agarose because it is a natural polysaccharide widely found in nature. Its biocompatibility, its high purity and its electroneutrality make agarose a suitable gelling medium-candidate for being mixed with the organic matrix components extracted from organisms.

Agarose is one of the constituents of agar, which is extracted from seaweed. It is a linear polymer made up of repeating units of agarobiose that form double helixes which later

associate through hydrogen bonding to form agarose fibers (Fig. 2.2). Agarose has a melting point temperature about 90 ºC and a gelling temperature about 30-40 ºC, depending on the agarose type. Very few concentration of agarose is needed to form a gel, being the critical concentration c* = 0.12 % (w/v). We used two agarose concentrations, 0.2 % (w/v) and 0.1% (w/v), in order to study the effect of the entanglement degree on CaCO3 precipitation. The former makes a proper gel and the latter, a partial gel or, also called viscous sol.

51

Figure 2.2 Left: Agarobiose unit. Agarose is made of repeating units of agarobiose (β-Dgalactopyranosyl-(1à4) 3,6-anhydro-L-galactose). Right: Agarose formation mechanism consists in two steps. In the first step, agarose fibers form double helix87 (B1) or associate in simple helical chains88 (B2). In the second step, helices associate to form the final agarose network (Cà D). Reproduced from [89].

2.2.1 Preparation of agarose media with entrapped organic additives Firstly, we prepared an agarose solution stock.89 The agarose powder was heated up to 90 ºC for 20 minutes to dissolve completely the agarose powder (Agarose D-5, Hispanagar). Then, the solution was cooled down to about 55 ºC and thereafter mixed with the required volume of heated milli-Q water to obtain a final 0.2% (w/v) or 0.1 % (w/v) agarose solution. Then, the agarose solutions were mixed with the organic additives in different beakers partially submerged in a bath at 50 ºC. In each beaker an amount of pAsp solution, pGlu solution or pLys solution was added to reach a final concentration of 20 μg/mL. In the case of SOMs, a different amount of dissolved BeuSOM, LprSOM, NpoSOM or PliSOM was added to reach a final concentration of 50 μg/mL (c) or 250 μg/mL (5c). The prepared solutions were stirring during 1-2 minutes and transferred to U-tubes using a syringe.

Chapter 2

2.3 Crystallization set-up The experiments were carried out using U-tubes devices (Triana Science & Technology, S.L., see Fig. 2.3). We used tubes of two different lengths, 90 and 45 mm. The longer ones were used in experiments to test the effect of polypeptides on CaCO3 precipitation while the shorter ones were employed for testing the effects of the different SOMs, because of the smaller available amounts of this type of biological material.

52

Figure 2.3 (a) Camera picture of the U-tube set-up used for CaCO3 crystallization experiments in CDS using agarose gels and viscous sols. (b) Schematic illustration of the measured crystallization parameters in the U- tube: xo, the starting point of precipitation; Δ, the crystal growing space; xcat and xan, the boundaries of the crystal growing space close to the cationic and anionic reservoir, respectively; Δcat and Δan, the crystal growing space from the starting point xo to xcat and xan, respectively.

2.3.1 Counter-Diffusion System Reagents from two adjacent side source reservoirs diffuse against each other through the gelling column of the device. Each reservoir (branch of the U-tube) contains 1 mL solution in the bigger tubs, or 0.2 mL in the smaller ones. The cationic reservoir was filled out with 0.5 M solution containing a Mg2+/Ca2+ equal to 0, 1, 3 or 5. These solutions were prepared by

mixing CaCl2.2H2O and MgCl2.6H2O (Sigma-Aldrich). The anionic reservoir, meanwhile, was filled out with a 0.5 M NaHCO3 solution (Fluka Biochemika). The initial pHs of the solutions were: 5.6 for 0.5 M CaCl2; 5.8 for 0.5 M Mg2+/Ca2+ equal to 1; 6.05 for 0.5 M Mg2+/ Ca2+ equal to 3; 6.1 for 0.5 M Mg2+/Ca2+ equal to 5; and 8.1 for 0.5 M NaHCO3. Depending on the set of experiments performed diverse cationic solutions diffused through the U-tube column filled with an agarose gel or a viscous sol (see specific methods section for each chapter).

2.3.2 Measured parameters In the U-tube set-up we measured several parameters: the waiting time (tw) or time that elapsed from the onset of the experiment up to the appearance of the first precipitate (observed under an optical microscope at magnification 4x); the starting point of precipitation (xo) or distance from the cationic reservoir to the place where the first crystals appeared and the crystal growing space (D) or length within the column where precipitates were observed after 14 days from the onset of the experiment. The boundaries of D are xcat in direction toward the cationic reservoir and xan in direction to the anionic deposit. Additionally D can be divided in two regions, Δcat and Δan, which correspond to the spaces from the starting point xo to xcat and to xan, respectively. A detailed theoretical definition of the measured parameters is provided below.

Precipitates were recovered from the column tube and filtered using 0.45µm pore size membranes (Cellu-Sep T1, MWCO: 3,500). The precipitates were washed several times with hot milli-Q water in order to remove agarose and then dried at room temperature. All experiments were performed at room temperature.

2.4 Theoretical considerations on a precipitation experiment in gel using the counter-diffusion system During a CDS experiment the precipitation of a given compound occurs in an almost convection-free environment and, therefore, the transport processes are almost exclusively controlled by diffusion. The reservoirs of the reagents’ solutions act as semi-infinite sources of ions creating time-dependent concentrations gradients along the gel. If the initial concentrations of diffusing reagents in the two reservoirs were the same, we would expect an increase of the ion activity product at the centre of the gel column, eventually reaching the threshold value for nucleation (i.e. critical supersaturation).22a, 32, 37b, 45b

53

Chapter 2

54

Figure 2.4 shows simplified simulations that capture the complex behaviour of simultaneous multicomponent mass transport, chemical speciation and nucleation/growth kinetics in a CDS experiment. The concentration gradients of Ca2+, HCO3- and CO32- are plotted in the top panel, where the initial conditions were 0.5 M of Ca2+ and 0.5 M HCO3-. The evolution of the saturation index (S.I.) in space and time is represented in the middle one. As soon as the IAP in this system overcomes the Ksp of calcite, at a particular region of the gel column, a precipitate in that zone appears and S.I. is reduced to 0. In this particular case, we observe at the bottom panel, representing the amount of precipitated calcite, the first visible maximum produced after 42 h of experiment and located at 53 mm from the Ca2+ reservoir (normalized length = 0.58). At increased periods of time, as the ionic diffusion continues, the precipitated mass of calcite increases, progressively reducing S.I. to 0. The shift of the maximum toward the anion reservoir indicates the unbalanced concentration of the two reactants solutions, being the anionic the limiting one.

Figure 2.4 Simulation of Ca2+, HCO3- and CO32- concentrations (top), saturation index with respect to calcite (middle) and amount of precipitated calcite (bottom) within the tube as a function of the position (x) at different times (from 12 to 72 hours, lines represented each 6 hours). The vertical dashed lines indicate the beginning, middle and end of the tube. The first line in the Saturation Index (S.I.) evolution plot corresponds to 12 hours. The first visible maximum in the precipitated calcite plot corresponds to 42 hours.

This maximum, normalized by the length of the tube, is located around x=0.58. (The simulations showed above were performed by co-author F.O. of [90]. While they are quite didactic to explain our experimental results, it is far from our purposes explaining the complex computational details of such calculations. Experimental details of the simulation are reported in the paper).

2.4.1 Definition and interpretation of the measured parameters A CDS allows us to assess the main parameters involved in a mineral precipitation process (Fig. 2.3). The starting point of crystallization, xo, is defined as the distance from the cationic reservoir to the first observed precipitate. This is the point where the equality range condition has to be fulfilled (the activities of the anions and cations involved in the precipitation have to be similar) and the ion activity product exceeds the critical value needed to yield nucleation and further growth.46 The waiting time, tw, refers to the time that elapses from the onset of the experiment to the observation of the first crystals, i.e. with an optical polarized microscope using a 4x objective in all our experiments.

At constant volume, tw can be considered inversely proportional to the nucleation frequency, which depends on the supersaturation and consequently of the critical ionic activities needed to sustain nucleation and growth. The crystal growing space (Δ) is the length of the region in the gel tube in which crystals are observed at the end of the experiment. We assume that the length of Δ could be unequivocally connected to the sensitivity of nucleation to the equivalence rule hypothesis. Thus, smaller regions imply a lower space where the system fulfils the cation-anion equality and lower ionic activity product needed to sustain nucleation and growth. Δ can be asymmetric around xo, in this case the growing space from xo to the last observed crystals in the direction toward the cationic reservoir (Δcat) differs with respect to the space in direction to the anionic one (Δan). The boundaries of Δ, xcat and xan, represent, respectively the places where the activity of anions in the zone close to the cationic reservoir and the activity of cations in the zone close to the anionic one are the lowest to still sustain nucleation and growth of crystals. The asymmetry of Δ with respect to xo suggests a different range of ionic activity of cations and anions to sustain nucleation and growth. The crystallization density, dc, represents the number of observed crystals by optical microscope per unit of volume. As long as the counting of crystals in the CDS is not a trivial issue, a qualitative scale was defined (Fig. 2.5): low density indicates that crystals are well separated one from another; medium density corresponds to a view in which crystals can still be observed as single units but they partially overlap; high density represents a massive precipitation in which single crystals are not distinguishable anymore. This parameter (dc) is important since it is an indirect measure of the frequency of nucleation, assuming that a strong inhibition of growth is not present. The crystal size was also considered, although in many cases we found it impossible to measure due to an overlap of the particles.

55

Chapter 2

Figure 2.5 Graphical representation of the qualitative scale defined to estimate the crystallization density. Low Density indicates that crystals are well separated the ones from the others (A). Medium Density corresponds to a view in which crystals can still be observed as single unit but they are partially overlapped (B). High Density represents a massive precipitation in which single crystals are anymore distinguishable (C).

56

All these parameters are a function of the boundary conditions, i.e. concentration of the reservoir solutions, nature and degree of entanglement of the gel, length and cross-section of the gel tube, temperature and final time of the experiment. They are supposed to be constant in comparative experiments in which additives are present into the gel or in one of the reservoir solutions.

2.4.2 Role of additives during a CDS precipitation process The comparison among measurable parameters in the absence and presence of additives allows assessment of their promoting or inhibiting effect on the nucleation and/or growth of crystals.

An inhibition or slowdown of the nucleation process produced by adsorption of additives on first formed nuclei (or even on pre-nucleation clusters) of a given mineral phase could cause a rise of tw along to a reduction of D, while xo should remain fixed. As the ion diffusion continues, this process must lead to an increase of the actual ion activity product (i.e. supersaturation) in the bulk of the gel. An increase of the density of crystallization (dc) is also expected since the nucleation rate is an (negative) exponential function of supersaturation.

An inhibition of the growth process should not affect the dc and D values, but the size, growth morphology and aggregation of the crystals. The presence of inhibition on nucleation and growth should cause an increase of tw, a decrease of D with higher dc and leave xo unchanged.

Moreover the crystal morphology should change. The increase of supersaturation at which nucleation and growth can occur in a precipitation process inhibited by soluble additives may also provoke the precipitation of a second phase, usually a polymorph with higher solubility.

2.5 Characterization of CaCO3 precipitates The morphology of the precipitates was study by an Optical Microscopy and Scanning Electron Microscopy, while polymorphism was analysed by X-ray diffraction, Fourier Transform Infrared and Raman Spectroscopy.

2.5.1 Morphological characterization.

Optical Microscopy (OM). Optical microscopy observations were done using a Nikon AZ100 optical microscope connected to a digital camera (Nikon, DS-Fi1).

Scanning Electron Microscopy (SEM). Samples were firstly inspected with a PhenomTM scanning electron microscope, desktop model. In addition, most samples were further characterized by using a GEMINI Carl Zeiss SMT field emission scanning electron microscope.

Samples were carbon-sputtered before observation and acquisition of the micrographs.

2.5.2 Polymorphic characterization X-ray diffraction (XRD). X-ray powder diffraction patterns were collected using a Panalytical X’Pert MPD diffractometer provided of Cu Ka radiation. The diffraction patterns were collected within the 2q range from 10º to 60º. As the amounts of precipitate obtained in some experiments were too small to be analysed by X-ray powder diffraction, therefore, in those cases crystals were ground and mounted on a Bruker X8 Proteum single crystal diffractometer equipped with a Microstar copper rotating anode generator, a κ goniometer, and a SMART 6000 CCD detector. The calculated XRD powder diffraction patterns were obtained after integrating the diffraction frames with the XRD2DSCAN Software.91 The main diffraction patterns of crystalline anhydrous CaCO3 polymorphs are: calcite, 23.1 (012), 29.4 (104), 47.5 (018) and 48.5 (116);92 aragonite 26.2 (111), 27.2 (021) and 45.9 (221);93 and vaterite 24.8 (100), 27.0 (101), 32.7 (102) and 43.8 (110).94 Fourier Transform Infrared (FTIR) spectroscopy. Fourier Transform Infrared spectra were collected using a FTIR Nicolet 380 instrument (Thermo Electron Co.) within the wavelength range from 4000 to 400 cm-1 at a resolution of 4 cm-1. Disks were made by applying a pressure of 48.6 psi by means of a hydraulic pump to a mixture consisting of 1 Mg of sample and 100 Mg of KBr.

57

Chapter 2

The main FITR absorption bands of the CaCO3 anhydrous polymorphs are located at the following wavelengths (in cm-1 units): for calcite, 1429 (antisymmetric, ν3), 1012 (symmetric, ν1), 877 (out of plane bend, ν2) and 713 (in-plane bend, ν4); for aragonite, 1477 (ν3), 1083 (ν1), 858 (ν2) and split peaks at 713 and 700 (ν4); and for vaterite, 1489 and 1432 (ν3), 1083 (ν1), 877 (ν2) and 745 (ν4).95

58

Raman spectroscopy. Raman spectra were collected using a LabRAM HR spectrometer with backscattering geometry (Jobin-Yvon, Horiba, Japan). The excitation line was provided by a diode laser emitting at a wavelength of 532 nm and a Peltier cooled charge-couple device was used as detector. Spectrometer resolution is better than 3 cm-1. The spectra were base-line corrected for clarity. Crystals were poured into a petri dish and observed under the Raman-microscope (10x) before collecting the spectrum of several crystals from each condition (only conditions where Mg2+ were diffusing from cationic reservoir and soluble organic matrix were entrapped into the gel).

Raman peaks at 155, 280, 713 and 1087 cm-1 are assignable to calcite while peaks at 150, 205, 701 and 1085 cm-1, to aragonite. Amorphous phases are detected by a broadening of all peaks, mainly a broad peak around 1085 cm-1 and a broad hump around a 150-300 cm-1.96

03 Influence of charged polypeptides on the precipitation of CaCO3

59

3. Influence of charged polypeptides on the precipitation of CaCO3 3.1 Introduction

CaCO3 is involved in many natural (e.g. biomineralization), environmental (scaling, CO2 cycle), and industrial processes in which it is usually employed as filler or pigment in plastics, rubber, paper, paints, plastisol as well as in pharmaceutical and food industry. Controlling polymorphism, growth rate, morphology, stability and size distribution of CaCO3 is very important in all these fields, and most of the regulation methods lie on the use of biological or synthetic molecules. A huge number of additives have been proposed to efficiently control these parameters (see e.g. reviews [59a, 97]) including Langmuir monolayers (e.g. [98]) self-assembled films (e.g. [99]) macromolecules extracted from biominerals (e.g. [14b, 100]) or synthetic macromolecules (e.g. [101]). However, in spite of the existing variety of additives, researchers are still not able to, for example, exactly reproduce the structure of a biomineral.

Although it is well known that macromolecules are strongly influencing the mechanisms of nucleation or crystal growth or both, the precise mechanism by which bio-macromolecules control the formation of minerals still remains unknown in many aspects.

The composition of the organic matrix extracted from biominerals differs from organism to organism, even though a common feature is that soluble components are acidic in nature, containing a high amount of aspartic and glutamic residues. This is the main reason why some in vitro mineralization studies use poly-aspartate (pAsp) or poly-glutamate (pGlu) as model polypeptides for studying the effect of bio-macromolecules in controlling CaCO3 precipitation. The high amount of carboxylate groups able to bind calcium ions (Ca2+) in solution, or emerging surface >Caδ+ at the crystal faces, is thus its main characteristic. The primary structure of pAsp and pGlu only differs in an additional carbon atom in the pGlu chain. However their secondary structures are different. In the presence of Ca2+, pAsp tends to adopt a β-sheet and random coil structure while pGlu adopts α-helix and also random conformation.102 The disposition of pAsp in plate sheets might favour the interaction between its carboxylate groups and the emerging >Caδ+ of the crystal surfaces, thus, influenc-

61

Chapter 3

ing crystal growth and therefore the crystal morphology by bonding too specific faces.

In the research presented in this chapter, we use charged polypeptides entrapped in agarose viscous sols to study their effect on the precipitation of CaCO3 by the counter-diffusion method. Charged polypeptides are poly-L-aspartate (pAsp), poly-L-glutamate (pGlu) and poly-L-lysine (pLys), being the latter a basic polypeptide that contains an additional amino group in the chain, positively charged, that allows comparing its effect on CaCO3 precipitation with respect to those produced by the negatively charged ones. In short, by determining the precipitation parameters defined in chapter 2 (section 2.4) together with a study of the morphology and polymorphism of the obtained precipitates, our aim is to test the protocol explaining in chapter 2 and to establish the polypeptide capability to inhibit, or eventually promote, the CaCO3 nucleation and/or growth processes.

62

3.2 Experimental section Experiments have been carried out by the CDS technique using U-tubs. A U-tube has a column that is accessible to diffusing reagents from two side reservoirs. The column was filled up with agarose solutions mixed with one of the polypeptides. In all cases, the final concentration of agarose in the column was 0.1 % (w/v) while polypeptide concentration was always 20 μg/mL. Due to the low agarose concentration this medium is categorized as a viscous sol instead of a proper gel. The commercial polypeptides added were poly-L-aspartic acid sodium salt (MW 9600 Da; Sigma-Aldrich), poly-L-glutamic acid sodium salt (MW 9600 Da; Sigma-Aldrich) or poly-L-lysine hydrochloride (15000 Da; Sigma-Aldrich). The cationic reservoir of the U-tube was filled up with 0.5 mol/L CaCl2 solution prepared from

CaCl2. H2O (Sigma-Aldrich), and the anionic branch was filled up with 0.5 mol/L NaHCO3 (Fluka Biochemika).

When first precipitates appeared in each U-tube, we measured the distance from the cationic reservoir to the starting point of precipitation (xo) and, 14 days after, the length of the crystal growing space (D). Then we measured the lengths of both the regions from xo to the cationic (∆cat) and to the anionic (∆an) reservoirs, respectively. Two other important parameters measured in each run were the crystallization density (dc) and waiting time (tw). Precipitates were recovered from the U-tube, and then filtered and washed with deionized water. After drying, the precipitates were characterized by Fourier transform infrared spectroscopy (FTIR) and X-ray powder diffraction. Finally they were inspected under an optical microscope and its growth morphology analysed by a scanning electron microscope.

Further details of the methodology employed or a description of the measured parameters is found in chapter 2.

3.3 Results Figure 3.1 shows pictures of crystal growing spaces (D) of the different U-tubes after the crystallization trials. The measured parameters are also graphically reported and summarized in Table 3.1.

63

Figure 3.1 Left: Graphical view of the measured parameters in absence (A) and in presence of the entrapped polypeptides pLys (B), pGlu (C) and pAsp (D). The length of the U-tubs has been normalized starting from cation reservoir (0) to the anion deposit (1). The real length of the tubs was 90 mm. Red and blue colours indicate the crystallization regions ∆cat and ∆an respectively. Arrows indicate the waiting time (tw, leftupper corner) and the number of replica is shown in the right-upper corner. Horizontal black lines in the middle of each figure and vertical grey lines on the arrow show the standard deviations of the measures.

Crystallization density (see Fig. 2.5) was represented as squares (low), stars (medium) and circles (high). Right: Camera pictures of the crystal growing spaces (∆) after 14 days of crystallization time under conditions of counter-diffusion in the presence of charged polypeptides entrapped in agarose viscous sols. Scale bar: 5 mm.

Chapter 3

64

Reference experiments (4 replica) were carried out in the absence of polypeptides. Calcium chloride and hydrogen carbonate solutions diffused against each other through the agarose viscous sol. In this experiment, the waiting time (tw) was around 2 days (1.8 ± 0.5 days) and the position of the first observed precipitate (xo) was 0.66 ± 0.12. Precipitation evolved asymmetrically with respect to xo and after 14 days, Dcat and Dan were equal to 0.10 and 0.16, respectively, resulting in D of 0.26 (Table 3.1). The precipitate showed a low crystallization density (dc) in the Dcat region and a medium dc in the Dan region. Under this particular condition calcite was the only precipitated phase, as detected by FTIR spectroscopy and XRD (Fig. 3.2). By using these techniques only the characteristic absorption bands at 1422 cm-1 (ν3), 875 cm-1 (ν4) and 712 cm-1 (ν2) and diffraction peaks at 23.0° (012), 29.4° (104), 35.9° (110), 47.5° (018) and 48.5° (116) were observed, respectively.64b, 103 Calcite appeared as single crystals whose sizes oscillated between 200 and 300 mm. They displayed rhombohedral {10.4} faces plus less extended {hk.0} faces (Fig. 3.3 and 3.4).

Table 3.1. Summary of results obtained from experiments of CaCO 3 precipitation by CDS in the absence of polypeptides and in the presence of pLys, pGlu OR pAsp entrapped in the agarose viscous sol. The precipitate features refer to the CaCO 3 polymorph, shape and size of crystals as recorded after removing the precipitate from the agarose.

CRYSTALLIZATION PARAMETERS

∆cat *

xo*

∆an *

PRECIPITATE FEATURES

tw **

dc***

nº

phase#

shape##

size###

Ref.

0.10(0.06) 0.66(0.12) 0.16(0.06)

~2

l,m

4

C

rhomb.

200-300

pLys

0.11(0.08) 0.68(0.09) 0.18(0.06)

~4

l,m,h

3

C,A

r.aggreg.; spher.

200-500 50-100

pGlu

0.06(0.07) 0.66(0.08) 0.06(0.05)

~2

m

4

C

facet rh.

180-250

pAsp

0,02(0.01) 0.64(0.02) 0.03(0.02)

~3

h

4

C

r.aggreg.; pean.

100-250

* Values normalized with respect to the real length of the U-tube from the cation (0) to the anion reservoir (1). Their associated standard deviations are reported in parentheses. ** The tw is measured in days. *** The dc is defined by a qualitative scale: low (l), medium (m) and high (h) (see Fig. 2.5). # Precipitated mineral phase: C and A indicate calcite and aragonite, respectively. ##Morphology of crystals observed by SEM: rhomb. indicates modified rhombohedra; r. aggreg. indicates aggregates of modified rhombohedra; spher. indicates spherulites; facet rh. indicates highly facet modified rhombohedra; pean. indicates peanut-like shape. ### The size of the particles (μm) refers to the longest axis.

65

Chapter 3

66

When pLys was entrapped into the viscous sol, the value of xo was of 0.68 ± 0.09 and D was of 0.29. Dcat (0.11) was shorter than Dan (0.18) and the tw was around 4 days (3.7 ± 0.5). The dc was high in the central region of D, low in the Dcat region and medium in the Dan one. In this case the FTIR spectrum and the X-ray powder diffraction pattern showed additional bands at 1467 cm-1, 855 cm-1 and 700 cm-1 and diffraction peaks at 26.2° (111), 27.2° (021), 38.6° (022) and 45.9° (211), with respect to those from the reference experiment (Fig. 3.2), thus indicating the presence of aragonite plus calcite. Calcite appeared as aggregates of modified rhombohedra crystals having sizes in the range 200-500 mm. The constituting crystalline units of the aggregates showed rough surfaces, almost not displaying the typical {10.4} faces. Aragonite crystallized well embedded in the agarose viscous sol as spherulites terminated by needles. The size of aragonite spherulites, 50-100 µm, suggests that they were formed in the Dan region (Fig. 3.1, right).

Figure 3.2 X-ray powder diffraction patterns (left) and FTIR spectra (right) of CaCO3 precipitated into the agarose viscous sol entrapping pLys (B), pGlu (C) or pAsp (D), and in absence of polypeptides (A). The intensity of the absorption bands at the FTIR spectra and the intensity of the peaks of the XRD diffractograms are both expressed in arbitrary units (a.u.).

In the presence of entrapped pGlu in the agarose viscous sol the xo value was of 0.66 ± 0.08 and the tw was about 2 days (1.8 ± 0.9). D was 0.12 and almost symmetric around xo.

The FTIR spectrum and the XRD pattern showed only the bands and peaks characteristic of calcite (Fig. 3.2). The dc was medium along the D region. Highly faceted rhombohedral calcite crystals whose sizes oscillate between 180 and 250 mm composed the precipitate. The crystals displayed rhombohedral {10.4} together with extended {hk.0} faces.

Figure 3.3 Optical microscope pictures of the CaCO3 crystals precipitated in the agarose viscous sol in absence of additives (A) and in the presence of entrapped pLys (B), pGlu (C) and pAsp (D).

When pAsp was entrapped into agarose viscous sol xo was of 0.64 ± 0.02 and tw about 3 days (2.8 ± 0.9). Dcat was similar to Dan, 0.02 and 0.03, respectively, resulting in a D of 0.05. dc in this short region is high. The FTIR spectra and the XRD diagram show the presence of only calcite. These calcite particles precipitated as irregular aggregates formed by distorted rhombohedra of sizes between 100 and 250 mm grouped in a radial way. The morphology of these single crystals is prismatic. They displayed elongated {hk.0} faces capped by rhombohedral {10.4} surfaces.

67

Chapter 3

Figure 3.4 Morphology of CaCO3 crystals formed in the agarose viscous sols in absence of additives (A, E) and in the presence of entrapped pLys (B, F), pGlu (C, G) and pAsp (D, H). (A-D) Optical microscope pictures of the CaCO3 crystals immersed in the agarose viscous sol and SEM pictures (E-H) of the same crystals after removal from the agarose. The inset in (F) shows a spherulite of aragonite while those in (E), (G) and (H) show high magnification of the calcite crystals. These crystals are representative of the whole sample populations.

68

3.4 Discussion The counter-diffusion technique can be used to monitor the effects of increasing ionic concentration during the precipitation of salts over the time, under conditions controlled mostly by diffusion. By applying the theoretical principles, described in chapter 2 (section 2.4), it tests whether is possible to assess the effects of an additive on the nucleation and growth of a given salt. The influence of additives during precipitation in gels has been considerably investigated, but more attention has been paid to the effects of the temporal increasing ion activities on the morphological evolution of the precipitates.35b, 39a, 39b, 39d, 39e, 41

As we described in chapter 2 (section 2.4), the position of xo has to fulfil the equivalence rule, being a function of the concentration of Ca2+ and CO32- ions. As the pH of the hydrogen carbonate solution was 8.15 ± 0.10, the estimated molar fraction of carbonate ions in the anionic reservoir was about 1.5 % of total carbonate. In all experiments, both in absence or presence of entrapped additives, the pH of the agarose assumed a value of 6-6.5. Agarose viscous sol was not buffered, so its pH changed during the ionic diffusion process and appeared governed by the diffusion of HCO3- and CO32- ions. The fact that the starting point of precipitation was around xo = 0.66 in all experiments is a consequence of the lower concentration of CO32- respect to that of Ca2+ (about 1.5%). As a consequence of the non-stoichiometry of reagents concentration, xo has to shift from the centre of the tube toward the anionic reservoir (limiting reactant) in order to fulfil the equivalence rule. The carbonate speciation effect also justifies the asymmetry of D, being ∆an > ∆cat. It is worth mentioning that the borders of D are determined by the concentration in the reservoirs and the required conditions of supersaturation and critical ions concentration to sustain the precipitation process.

The xo position depends on the speciation and diffusion of the species involved. Since the addition of pAsp, pGlu and pLys to the agarose viscous sol did not significantly change the xo position, we may deduce that both, the speciation and diffusion of carbonate ions was not appreciably affected by these additives. Other physicochemical processes must have produced the same result, for example that these additives would have affected cationic and anionic diffusion coefficients. However, the assumption that a negatively (o positively) charged polypeptide can reduce the diffusion rate of cation and anion in the same way is unlikely.

In the case of pLys, aragonite precipitated together with calcite. As aragonite is only slightly more soluble than calcite, the supersaturation reached with regard to aragonite at the moment of nucleation is very high as well. If pLys inhibits calcite growing, then aragonite should precipitate because the environment is highly supersaturated with regard to aragonite. The observation that pLys does not influence the growing space suggests that the increase of supersaturation is a consequence of growth inhibition of calcite crystals and that the nucleation processes are not influenced by the presence of this additive. Indeed, pLys is a basic polypeptide (pKa ~9) that is positively charged at the pH of the agarose viscous sol. The inhi-

bition process likely takes place after the nucleation event, by electrostatic interaction of the additive with the negatively charged surface of calcite crystals likely formed on the Dan region (in the presence of excess of anions), thus expressing these faces in the growth morphology.

In accordance to this supposed inhibition, a change on the morphology as well as aggregated calcite crystals was observed with this additive.

The inhibition effect of acidic polypeptides and proteins in the nucleation of CaCO3 is well known.17, 86, 104 Here pAsp, and pGlu to minor extent, caused an increase of tw and dc and a

69

Chapter 3

strong decrease of D with respect to the reference. All these observations are in agreement with the strong inhibition of the nucleation of CaCO3. The presence of these two negatively charged additives also caused a modification of the calcite growth morphology and acted as glue producing the crystal aggregation. These findings indicate that these additives also affected the growth process.100

3.5 Conclusions

70

This chapter reports the influence of charged polypeptides, either basic (pLys) or acid (pAsp and pGlu) on the precipitation process of CaCO3 during a CDS experiment. We have assessed the crystallization parameters described in chapter 2 (xo, tw, D, Dan, Dan) and confirmed their validity for studies on the inhibiting or promoting capability of nucleation and/or growth in a mineral precipitation process. As part of this work, it has been found that pLys actually influences the growth mechanism of CaCO3 without affecting the nucleation process; meanwhile pAsp, and to a minor extent pGlu, affect both the nucleation and growth processes of CaCO3.

This work has been published (see M. Sancho-Tomás, S. Fermani, M.A. Durán Olivencia, F. Otalora, J. Gómez-Morales,* G. Falini*, J.M. García-Ruiz. Influence of charged polypeptides on nucleation and growth of CaCO3 evaluated by counter-diffusion experiments. Crystal Growth & Design. 2013, 13, 3884-3891).

04 Influence of diffusing Mg2+ ions together with entrapped polypeptides on CaCO3 precipitation

71

4. Influence of diffusing Mg2+ ions together with entrapped polypeptides on CaCO3 precipitation 4.1 Introduction

The role of Mg2+ on CaCO3 biomineralization processes has been deeply studied, nevertheless their specific mechanism(s) of action or how they are included in biomineral structures by organisms are still not completely known. Mg2+ influence the precipitation of CaCO3, being able to inhibit the growth of calcite, and thus provoking the precipitation of aragonite. When the Mg2+/Ca2+ concentration ratio in a solution is above 4, aragonite is almost the only precipitated phase.64 Therefore, considering that the Mg2+/Ca2+ in oceans is actually around 5, its influence on aragonite formation by organisms should be one of the important factors to be considered. On the other hand, it was shown that formation of magnesium-calcite biominerals occurs via an amorphous precursor phase and that Mg2+ ions may be important in stabilizing it,105 particularly in the presence of organic additives.51a, 106 In some calcitic biominerals, Mg2+ substitute in Ca2+ site, leading to magnesium-calcite mineralized tissues. The content of Mg2+ in calcite covers a wide composition range; for example some mineralized tissues from algae and echinoderms show an isomorphic substitution of 30% of magnesium to calcium107 or even up to 40%.65 This percentage was never reached synthetically at ambient conditions, although using alcohols or molecules bearing carboxylate groups the percentage can increase up to 14%.66, 108 Some studies have shown that macromolecules presented in the biominerals could regulate and control the Mg2+ content in mineralized tissues.105a,109 The aim of the present study is to address the combined effect of acidic or basic polypeptides entrapped in the agarose viscous sol and diffusing Mg2+ from Mg2+/Ca2+ solutions (contained in the cationic reservoir) in CaCO3 precipitation.

73

Chapter 4

4.2 Experimental section Experiments have been carried out by CDS using U-tubs. In each trial the column has been filled with agarose solution mixed with each one of the different polypeptides. The final agarose concentration in the column was 0.1 % (w/v) (viscous sol) while the polypeptide concentration was always 20 μg/mL. The added polypeptides were poly-L-aspartic acid as a sodium salt (MW 9600 Da; Sigma-Aldrich), poly-L-glutamic acid in the form of a sodium salt (MW 9600 Da; Sigma-Aldrich) or poly-L-lysine as a hydrochloride salt (15000 Da; Sigma-Aldrich). The cationic reservoir was filled up with a 0.5 M solution having a Mg2+/Ca2+

74

equal to 0, 1, 3 or 5. These solutions were prepared by mixing MgCl2.6H2O and CaCl2.2H2O (Sigma-Aldrich). The anionic reservoir was filled up with a 0.5 M NaHCO3 solution (Fluka Biochemika).

When the first precipitates appeared in each U-tube, we measured the distance from the cationic reservoir to the starting point of precipitation (xo) and the length of the crystal growing space (D), 14 days after. Then ∆cat, ∆an, dc and the waiting time (tw) were measured.

Precipitates were recovered by the U-tube, filtered and washed with deionized water. After being dried, we characterized the precipitates by FTIR and XRD, and observed the precipitates morphology by an OM and SEM. For further details of the methodology employed or the measured parameters description, see chapter 2.

4.3 Results A set of reference experiments were carried out in absence of entrapped polypeptides in the viscous sol located into the body of the U-tube. Since the research in the presence of entrapped polypeptides, using Mg2+ free solutions, has been already presented in chapter 3, in this chapter we will describe and discuss mainly the results of experiments carried out in the presence of Mg2+. Figure 4.1 shows camera pictures of crystal growing spaces (D) of the different crystallization U-tubes used for the crystallization experiments above reported.

The parameters measured during the crystallization experiment are graphically reported in figure 4.2 and summarized in table 4.1.

75

Figure 4.1 Camera pictures of the crystal growing spaces (ďż˝) obtained after 15 days. From left to right the columns show the precipitates obtained in viscous sol (A, E, I, and M) and in viscous sol entrapping either pLys (B, F, J, and N), pGlu (C, G, K, and O) or pAsp (D, H, L, and P), respectively. From up to down the lines correspond to absence of magnesium ions (A-D) in the cation reservoir and Mg2+/Ca2+ equal to 1 (E-H), 3 (I-L) and 5 (M-N), respectively. (Scale bar of the first row is 5 mm; scale bar of the rest of the pictures is 2 mm).

In the reference experiments in the absence of entrapped polypeptides, at Mg2+/Ca2+ = 1, the precipitation of magnesium-calcite together with aragonite was detected (Table 4.1, Fig. 4.5, Fig. 4.6). The value of xo was 0.68, Dcat = 0.13, Dan = 0.15 and the first crystals were observed after 4 days. Increasing the concentration of Mg2+ to Mg2+/Ca2+ = 3, the precipitation of magnesium-calcite was still detected together with aragonite appeared as predominant phase.

When the Mg2+/Ca2+ was 5 in the cationic reservoir only the precipitation of aragonite was observed. Increasing the Mg2+/Ca2+ from 3 to 5, the xo value shifted from 0.60 to 0.42, Dan was 0.12 in both cases while Dcat were 0.19 and 0.16, respectively. In these last two experimental conditions tw was higher than in the reference experiment and increased when more Mg2+ were added. On the other hand, in the presence of Mg2+ a low precipitation density was always observed, irrespective of the Mg2+/Ca2+ used.

Chapter 4

76

Figure 4.2 Graphical representations of the parameters measured in the precipitation experiments of calcium carbonate by CDS. Red and blue colours indicate the crystallization regions �cat and �an respectively. Arrows indicate the waiting time (tw). Horizontal black lines in the middle of each figure and vertical grey lines on the arrow show the standard deviations of the measures. Crystallization density (further details in figure 2.5) was represented as squares (low), stars (medium) and circles (high). From up to down the lines correspond to Mg2+ -free experiments (A) and Mg2+/Ca2+ equal to 1 (B), 3 (C) and 5 (D), respectively.

The presence of Mg2+ in the cationic reservoir changed the morphology of calcium carbonate crystals precipitated in the agarose media (Fig. 4.3 and Fig. 4.4). In Mg2+-free experiments typical rhombohedra of calcite with sizes around 200-300 mm were observed. As the molar content of Mg2+ was increased, crystals with spherical shapes were predominant. It is worth of note that a wide range of morphologies was observed when Mg2+/Ca2+ = 1. At these conditions, microcrystalline building units grouped into rounded shapes about 100-240 mm precipitated together with small bunches of aragonite fibers (10-20 mm). As the Mg2+/Ca2+ increased from 3 to 5, spherical shapes progressively smaller appeared. In the last conditions we found crystals with hexagonal acicular shapes and sizes around 120-180 mm, which appeared smaller around 50-150 mm and sharper acicular with spherical shapes at the highest Mg2+ concentration.

Table 4.1 Left: Crystallization parameters measured in experiments of CaCO 3 precipitation in CDS using charged polypeptides entrapped in an agarose viscous sol and diffusing Mg 2+ ions. Measured parameters are xo, ∆ cat, ∆ an, t w and d c for a number (nº) of measured-day per day-replicas for each condition. The data are graphically represented in fig. 4.2. Right: Calcium carbonate polymorphism (phase) characterized by FTIR and XRD, morphology (shape) and size distribution of crystals analysed by Optical and Scanning Electron Microscopies. From up to down the lines correspond to Mg 2+-free experiments (A) in the cation reservoir and Mg 2+/Ca 2+ equal to 1 (B), 3 (C) and 5 (D), respectively.

CRYSTALLIZATION PARAMETERS

PRECIPITATE FEATURES

77

(A)

(B)

(C)

(D)

∆cat

xo

∆an

tw

dc

nº

phase

shape

size

Ref.

0.10(0.06)

0.66(0.12)

0.16(0.06)

~2

l,m

4

C

rhomb.

200 - 300

pLys

0.11(0.08)

0.68(0.09)

0.18(0.06)

~4

l,m,h

3

C,A

pGlu

0.06(0.07)

0.66(0.08)

0.06(0.05)

~2

m

4

C

facet rh.

180 - 250

pAsp

0,02(0.01)

0.64(0.02)

0.03(0.02)

~3

h

4

C

r.aggreg.;pean.

100 - 250

Ref.

0.13(0.05)

0.68(0.04)

0.15(0.04)

~4

l

3

MgC,A

r.aggreg.

100-240;10-20

pLys

0.13(0.07)

0.68(0.12)

0.15(0.05)

~3

l,h,m

3

A, (MgC)

r.aggreg.;spher. 150-350; 30-90

pGlu

0.07(0.02)

0.65(0.04)

0.11(0.05)

~4

m

3

MgC,A

r.aggreg.;spher. 150-300; 5-15

pAsp

0.04(0.02)

0.76(0.02)

0.04(0.03)

~4

h

3

MgC

r.aggreg.

40 -150

Ref.

0.19(0.02)

0.60(0.04)

0.12(0.02)

2-3

l

3

A, (MgC)

spher.

120 -180

pLys

0.16(0.07)

0.68(0.10)

0.14(0.09)

~5

l

3

A

spher.

80 - 275

pGlu

0.09(0.01)

0.74(0.04)

0.09(0.07)

~3

m

3

A,**MgC

spher.

80 - 220

pAsp

0.04(0.08)

0.74(0.08)

0.02(0.09)

~6

m

4

A,*MgC

spher.,pean.

70-150;15-20

Ref.

0.16(0.05)

0.42(0.07)

0.12(0.05)

~3

l

3

A

spher.

50 - 150

pLys

0.14(0.04)

0.70(0.04)

0.11(0.03)

~5

l

6

A

spher.

50 - 225

pGlu

0.08(0.04)

0.66(0.03)

0.06(0.05)

2-3

l,m

3

A

spher.

40 - 120

pAsp

0.06(0.14)

0.73(0.14)

0.018(0.16)

~8

l

4

A,**MgC

spher.

25 - 100

r.aggreg.;spher 200-500;50-100

* Precipitated mineral phase: C: calcite; A: aragonite; MgC: Magnesium Calcite; *(**)MgC traces (and very few traces) of magnesium-calcite. (MgC): magnesium-calcite was only detected by X-ray powder diffraction. Morphology of crystals observed by SEM: rhomb. indicates modified rhombohedra; r. aggreg. indicates aggregates of modified rhombohedra; spher. indicates spherulites; facet rh. indicates highly facet modified rhombohedra; pean. indicates peanut-like shape. The size of the particles (µm) refers to the longest axis.

Chapter 4

When pLys was entrapped into the agarose medium, the value of xo remained almost constant, irrespective of the starting Mg2+/Ca2+ (table 4.1 and Fig. 4.2). D did not change significantly with respect to the reference experiments. In conditions of Mg2+-free and Mg2+/ Ca2+=1, Dcat (0.11 and 0.13) were shorter than Dan (0.18 and 0.15). On the other hand, when the Mg2+/Ca2+ increased to 3 or 5, the values of Dcat (0.16 and 0.14) became longer than those of Dan (0.14 and 0.11, respectively). Concerning the tw values, the first condition where we observed precipitates was Mg2+/Ca2+ = 1 (around 3 days) followed of Mg2+/Ca2+ = 0 (around 4 days) and then of Mg2+/Ca2+ = 3 and 5 (a bit longer, around 5 days). After two weeks, the observed crystallization densities (dc) when using starting Mg2+/Ca2+ = 0 and 1 were a mixture of low (closer to the cationic reservoir), high (around xo) and medium (closer to the anionic reservoir). At higher ratios crystals were found more isolated.

78

Figure 4.3 Optical microscope pictures of calcium carbonate crystals immersed in the agarose viscous sol. From left to right the columns represent pure viscous sol (A, E, I, and M) and viscous sols entrapping pLys (B, F, J, and N), pGlu (C, G, K, and O) and pAsp (D, H, L, and P), respectively. From up to down the lines correspond to Mg2+-free (A-D) and Mg2+/Ca2+ equals to 1 (E-H), 3 (I-L) and 5 (M-N), respectively.

In the presence of pLys and absence of Mg2+ a mixture of aragonite and calcite precipitated.

When using Mg2+/Ca2+ = 1 traces of magnesium-calcite together with aragonite were observed (Fig. 4.6); Increasing the starting Mg2+/Ca2+ to 3 and 5, only the precipitation of aragonite occurred. The morphology of the last precipitates seems to follow the same trend of the reference experiments; indeed, more spherical shapes appeared by increasing the Mg2+ concentration. Aggregates of modified rhombohedra whose sizes oscillate between 200-500 mm as well as small hollow microspheres (50-100 mm) precipitated in absence of Mg2+. When Mg2+/Ca2+ = 1, cauliflower-like and hexagonal-acicular aggregates of sizes 150-350 mm together with a small and slightly elongated rhombohedral crystal were observed. Its size is approximately the same of the smaller crystals observed in absence of Mg2+. Aggregates of laminar-flower micro-units with no preferred orientation grown at Mg2+/Ca2+ = 3; the size of those crystals was within the range 80-275 mm. Finally, increasing the Mg2+/Ca2+ to 5, precipitates with smooth acicular shapes were observed (50-225 mm). Looking the fibers in detail (Fig. 4.4, N-inset), we can observe how they are glued, possibly by the action of the agarose or pLys.

Figure 4.4 Scanning electron micrographs of CaCO3 crystals precipitated in the agarose viscous sol. From left to right the columns represent pure partial gel (A, E, I, and M) and partial gel entrapping pLys (B, F, J, and N), pGlu (C, G, K, and O) and pAsp (D, H, L, and P), respectively. From up to down the lines correspond to absence of magnesium ions (A-D) in the cation reservoir and Mg2+/Ca2+ equal to 1 (E-H), 3 (I-L) and 5 (M-N), respectively.

The crystals shown in these pictures are representative of the entire population of the precipitates.

79

Chapter 4

A different scenario appeared when pAsp was entrapped into the viscous sol. The crystal growing space D was drastically reduced while tw increased with respect to the reference experiments. In conditions of Mg2+/Ca2+ = 1, 3 and 5 the values of xo appeared closer to the anionic reservoir (0.76, 0.74 and 0.73, respectively) than in the condition Mg2+-free (0.64). It is important to note that when the Mg2+/Ca2+ was increasing D became more and more asymmetric, with Dcat>Dan. That is, if in absence of Mg2+ Dcat (0.02) was slightly shorter than Dan (0.03), in the condition Mg2+/Ca2+ =1 these spaces were identical (0.04) and at Mg2+/Ca2+ = 3 and 5, the differences were progressively higher, Dcat>>Dan (Dcat = 0.04 and Dan = 0.02 with Mg2+/Ca2+ = 3; and Dcat = 0.06 and Dan = 0.018 with Mg2+/Ca2+ = 5).

80

Figure 4.5 FTIR spectra of CaCO3 precipitated in agarose viscous sols entrapping pLys (blue), pGlu (green) or pAsp (red), and in additive-free experiments (black). Each set of spectra refers to experiments Mg2+free (left-upper corner) or Mg2+/Ca2+molar ratios equal to 1 (right-upper corner), 3 (left-lower corner) or 5 (right-lower corner).

The absorption intensities of the FTIR bands are reported in arbitrary units (a.u.). The salient features of infrared spectra of calcite are 1429, 877 and 713 cm-1; and of aragonite, 1477, 1083, 858, 713 and 700 cm-1.103 The peaks at 1477 and 1429 cm-1 correspond to the ν3 absorption band of carbonate ions; the peaks at 877, 858 correspond to ν2 and 713 and 700 cm-1 correspond to ν4 absorption bands of calcite.

Irregular flower-like aggregates of calcite were observed when Mg2+/Ca2+ = 0 and 1. In the first condition, also peanut-shaped precipitates were observed. Besides, in these cases a high precipitation density was detected, while when increasing Mg2+/Ca2+ to 3 and 5, the precipitation density was lower and lower. Only after Mg2+/Ca2+ = 3, aragonite precipitated but always with traces of magnesium-calcite (Fig. 4.6). Two type of crystals appeared at Mg2+/Ca2+ = 3, shorter acicular spherical shapes of 40-175 mm and crystals with peanuts-like shape of 15-20 mm. The smallest crystals were found at Mg2+/Ca2+ = 5, where smooth rounded shapes of 25-100 mm precipitated.

81

Figure 4.6 X-ray powder diffraction patterns of the CaCO3 precipitated in agarose viscous sols entrapping pLys (blue), pGlu (green) or pAsp (red), and in absence of additive (black). Each set of spectra refers to experiments carried out in Mg2+-free (left-upper corner) or Mg2+/Ca2+ equals to 1 (right-upper corner), 3 (left-lower corner) or 5 (right-lower corner). The main diffraction peaks for calcite, (012), (104), (110), (018) and (116), and aragonite (111), (021), (022) and (221), are indicated according to the reference patterns PDF-calcite 01-083-0587 and PDF-aragonite 01-077-0606. The diffractograms are shifted along the Y-axis that reports intensity in arbitrary units (a.u.).

pGlu behaved similarly to pAsp, but it was less efficient as crystallization processes modifier; D decreased while xo shifted to the anionic reservoir when increasing the Mg2+ concentration. Indeed, when Mg2+/Ca2+ = 1 the first observed crystals were found at xo close to that

Chapter 4

found in absence of Mg2+ (0.65). However, at Mg2+/Ca2+ = 3 the xo shifted to the anionic reservoir (0.74). The difference was that when increased the Mg2+ content, xo appeared again closer to cationic reservoir (0.66); this finding is the same than in the absence of Mg2+.

Additionally, tw was similar to the reference experiments, but precipitation density was quite higher. Specifically, at Mg2+/Ca2+ = 5, a mixed of high and low precipitation density was found and more than that, as in the pLys- Mg2+/Ca2+ = 0 and 1, the highest density was found closer to the anionic reservoir.

82

Modified calcite rhombohedra of sizes about 180-250 mm precipitated when only Ca2+ was added into the cationic reservoir. At Mg2+/Ca2+ = 1, aragonite and magnesium-calcite with wide variety of morphologies were observed: irregular flower-like and acicular spherical shapes of sizes around 150-300 mm precipitated together with much more smaller spherulitic aggregates of acicular crystals of 5-15 mm. Increasing the Mg2+/Ca2+, acicular and smooth spherical shapes were the only crystal actresses of size around 80-220 mm at Mg2+/Ca2+ = 3 condition, and 40-120 mm at Mg2+/Ca2+ = 5 condition. Aragonite was the unique calcium carbonate phase detected at highest Mg2+/Ca2+ added; however traces of magnesium-calcite were identified using Mg2+/Ca2+ = 3.

4.4 Discussion In this research the combined effects of entrapped and diffusing additives on precipitation processes by counter-diffusion systems was studied. This represents an increase of complexity with respect to that presented in chapter 3. To achieve this goal, the precipitation of CaCO3 in the presence of Mg2+ (diffusing additive from a reservoir) together with charged polypeptides (entrapped additive in the agarose medium) was investigated. This choice was dictated by the relevance of these additives in biomineralization and scaling processes.1a, 60 The position of the starting point of precipitation, xo, has to respect the equivalence rule (see chapter 2 â&#x20AC;&#x201C;section 2.4). Thus, it is a function of the concentration of Ca2+ and CO32- ions.

In the presence of Mg2+ the concentration of Ca2+ is reduced in the cationic reservoir, so xo should shifts closer to the cationic reservoir, according to the equivalence rule. Besides, the asymmetry of the D region changed with respect to that found in absence of Mg2+, with Dcat>>Dan. In the equivalence rule the total contribution of cation should be considered, since both Mg2+ and Ca2+ are cations involved in the precipitation process. An important difference between Mg2+ and Ca2+ is in their hydration sphere, being the one of Mg2+ wider than that of Ca2+. This makes the diffusion process of Mg2+ slower than that of Ca2+. The relevance of

this difference increases with the diffusion time (e.g. it is minimal for short diffusion times and significant for long diffusion times), thus it is related with the experimental set up, in particular the length of the diffusing tube. The observed asymmetry in D can be also justified by this observation.

Increasing the concentration of Mg2+ the length of D did not change. It is well known that in the presence of Mg2+ the growth of calcite is inhibited. Then, aragonite precipitates because the environment is highly supersaturated with regard to this phase as well. However the precipitation of aragonite occurs without an overall effect on D, but with a longer tw, thus, suggesting that Mg2+ mainly affects the growth processes of calcite and aragonite, but seems not to influence nucleation. Interestingly, as a consequence of the increase in Mg2+ concentration the aragonite crystals reduced their average needle-thickness. This observation indicates the influence of Mg2+ in the growth of aragonite crystals. Therefore, the function of Mg2+ in the biomineralization processes could be related with acting as crystal morphology modifiers, a function already proposed to take place, for example, during aragonite precipitation in the coralâ&#x20AC;&#x2122;s skeleton.104a As we reported in chapter 3, pLys (positively charged at the working pH) does not influence D, which suggests that a longer tw is a consequence of the growth inhibition of crystals, a situation analogue to that reported for Mg2+. Accordingly, a change on the morphology and aggregation of calcite crystals was observed with this additive. D was never influenced by the diffusion of Mg2+, neither by the entrapped pLys in the agarose medium nor by both additives acting together. This result is expected since both additives seem to affect only the crystal growth process and not the nucleation one. The role of Mg2+ and pLys as aragonite morphology modifiers was thus synergistic. In the presence of high concentration of Mg2+ and entrapped pLys, crystals with two morphologies appeared, the plate-like ones were associated to aragonite. Interestingly, in the presence of entrapped pLys or entrapped pLys combined with diffusing Mg2+, xo remained constant. As pLys is positively charge, it could interact with carbonate ions. Even if we suppose that the effect of the polypeptides is not affecting the diffusion process, this weak interaction, that is not important when Mg2+ is not present, can probably have some role in the precipitation process.

Contrary to what we observed with pLys, the joint effect of diffusing Mg2+ plus entrapped pGlu or pAsp provoked a shift of xo to the anionic reservoir. In this case, it could be due to the combined play of a release of protons to the agarose medium from the acidic polypeptides (pKa of pAsp is 3.65 and pKa of pGlu is 4.25) that affect the speciation of carbonate in favour of HCO3-.

83

Chapter 4

The morphology of the aragonite crystals was influenced by the co-presence of pAsp or pGlu, and Mg2+. Aragonite crystals aggregated forming compact spherulites likely because of the supposed increase of supersaturation at which the precipitation might have occurred.

In the presence of pAsp the crystallization of aragonite is partially inhibited in favour of magnesium-calcite. High magnesium-calcite has a lower stability than aragonite, and thus favoured by the presence of high ionic activity products. On the other hand, only aragonite or calcite was detected, but precursors phases could have precipitated before. Further experiments, considering the use of in situ characterization techniques, need to be done to better explore the potentialities of the counter-diffusion system in the field of bio-crystallization.

84

4.5 Conclusion In the research presented in this chapter the combined role of diffusing Mg2+ and entrapped charged polypeptides during CaCO3 precipitation in the agarose viscous sol by using a counter-diffusion system was explored. Different Mg2+/Ca2+ for each one of the charged polypeptides (pAsp, pGlu and pLys), were used. Comparing the precipitation parameters (xo, tw, D, Dan, Dan) obtained in the different conditions, we found that Mg2+ seem to influence only the growth mechanism of CaCO3 and not the nucleation. This suggests a role as crystal growth modifier. We also detected calcite at high Mg2+/Ca2+ ratio. However, this only occurred in the presence of pAsp, and pGlu in minor extension. It means that both negatively charged polypeptides slightly favour the precipitation of calcite even at high Mg2+/Ca2+.

05 Influence of soluble organic matrix from scleractinian corals on CaCO3 precipitation

85

5. Influence of soluble organic matrix from scleractinian corals on CaCO3 precipitation 5.1 Introduction

Scleractinian corals are the biggest source of biogenic calcium carbonate on Earth67 and are among the fastest marine mineralizing organisms.68 Some corals live in symbiosis with algae, called zooxhantellae, that provide energetic support to them and are thought to facilitate the mineralization process (see e.g. review [111]). Besides, aminoacids composition of the organic matrix is influenced by the presence or not of these symbionts.112 Still, the mechanisms proposed to explain the relationship between zooxanthellae and calcification remains enigmatic.

Scleractinia skeletons (Fig. 1.12, see chapter 1 â&#x20AC;&#x201C;section 1.4.2) are composed of groups of needle-like aragonite crystals that radiate out from the centre of calcification,72 a medium rich in sulphur.9c, 73 Recent research suggested that there is a pathway involving direct seawater transport to the calcifying media in coral, which links the site of calcification to the surrounding ocean.113 Other similar in vivo experiments showed that seawater acidification leads to a gradual relative decrease in pH of the medium in the calcification site, leading to an increasing pH difference between the calcification site and seawater.114 The direct seawater transport to the calcification site implies that the precipitation of aragonite could be due to the high content of Mg2+ in seawater, with respect to calcium ions (Mg2+/Ca2+ equal to 5). However, control of the local saturation state at the nucleation site requires the involvement of biological macromolecules, which are secreted by the calicoblast cells. The role of these macromolecules is also to control the structural and textural organization of the mineral regions of the skeleton. Moreover, their activity could be regulated by the presence of Mg2+.67b, 68-69 Goffredo et al. (2011)104a showed that the intra-skeletal organic matrix from the Mediterranean solitary zooxanthellated coral Balanophyllia europaea favours the precipitation of aragonite and that this occurs through a transient phase of amorphous calcium carbonate likely stabilized by lipids. Furthermore, they showed that the organic matrix molecules also controlled the morphology of the precipitated calcium carbonate crystals. The influence of coral intra-skeletal organic matrix in the precipitation of calcium carbonate has also been demon-

87

Chapter 5

strated for the tropical species Acropora digitifera, Lophelia pertusa and Montipora caliculata.

115 This study highlighted the importance of the low molecular weight macromolecules in the control of calcium carbonate polymorphism. Important recent research has shown that four highly acidic proteins, derived from expression of genes obtained from the common stony coral, Stylophora pistillata, can spontaneously catalyse the precipitation of aragonite in vitro from seawater.116 However, despite these advances, the chemical and physical processes that take place at the nucleation site are still only partly understood.

88

This chapter presents the study of the precipitation of calcium carbonate in a counter-diffusion system (CDS) using a agarose viscous sol or an agarose gel in which the intra-crystalline soluble organic matrix (SOM) extracted from the solitary Mediterranean corals Balanophyllia europaea (BeuSOM), zooxanthellate, and Leptopsammia pruvoti (LprSOM), azooxanthellate, were added. The aim of the present study is to understand the influence of SOM in the precipitation of calcium carbonate in environments with different viscosities and to test the role of diffusing Mg2+.

5.2 Experimental Section Experiments were carried out in CDS by using U-tubs. The column was filled up with agarose solutions mixed with soluble organic matrix (SOM) extracted from two corals, Balanophyllia europaea (Beu) and Leptopsammia pruvoti (Lpr). Agarose final concentration in the column was 0.1 % (viscous sol) or 0.2 % (gel) and SOM was added in two concentrations, 50 μg/mL (c) and 250 μg/mL (5c). The cationic reservoir was filled up with 0.5 mol/L solution containing a Mg2+/Ca2+ of 0 or 3, where the concentration of Ca2+ was reduced to keep constant the ionic strength of the cationic reservoir solution. These solutions were prepared by mixing CaCl2.2H2O and MgCl2.6H2O (Sigma-Aldrich). Anionic reservoir was filled up with 0.5 mol/L NaHCO3 (Fluka Biochemika).

Extraction of soluble organic matrix. Both corals were randomly collected during scuba diving in the North-Western Mediterranean Sea, at Calafuria 43º 27’ N, 10º 21’ E. B. europaea was collected at 6 m depth and L. pruvoti, at 16 m depth.117 After collection, the corals were dipped in a commercial sodium hypochlorite solution for 4 days until the polyp tissue was completely dissolved, then the remaining skeletons were washed with deionised water, dried in an oven at 37 °C for 24 hours and stored. Each skeleton was analysed under a binocular microscope to remove fragment of substratum and calcareous deposits produced by other organisms. Fragments of the coral skeletons were suspended (1% w/v) in a sodium

hypochlorite solution (3% v/v) to remove traces of organic material eventually not removed by the first treatment. The samples were air dried for one night and ground in a mortar up to obtain a fine and homogeneous powder. Five mL of milli-Q water, in which 2.5-3 g approx. of powdered biominerals were dispersed, were poured into a 40 cm-long osmotic tube for dialysis (MWCO = 3.5 kDa; CelluSep®, MFPI). The sealed tube was placed into 1 L of 0.1M CH3COOH (Riedel de Haen) solution under stirring. The decalcification proceeded for 72 hr. At the end, the tube containing the dissolved OM was dialysed against milli-Q water (resistivity 18.2 MΩ cm at 25 °C; filtered through a 0.22 μm membrane) until the final pH was about 6. The obtained aqueous solution containing the OM was centrifuged at 6000 rpm for 6 minutes to separate the soluble (SOM) and the insoluble (IOM) OM fractions, which were then lyophilized.

Chemical analysis of SOMs. SOMs were characterized by their amino acid composition and by using FTIR spectroscopy. SOM material was hydrolysed at 100 ºC for 24 h in 6 M HCl vapour, and analysed using a Dionex BIOLC amino acid analyser. Lyophilized SOMs were hydrolysed with 6N HCl for 24 h at 110 ºC. During hydrolysis, complete or partial destruction of several amino acids occurs: tryptophan is destroyed; serine and threonine are partially destroyed. Sulphur amino acids are altered. Amino acid composition of the hydrolysates was also determined by high performance liquid chromatography (HPLC) using pre-column derivatization with ortho-phtalaldehyde (OPA) for primary amines and fluorometric detection. Fluorescence intensity of OPA-derivatized amino acids was monitored with an excitation wavelength of 330 nm and an emission wavelength of 450 nm.

When first precipitates appeared in each U-tube, we measured xo, tw and, 14 days after, the crystal growing space (Δ). Precipitates were recovered by the U-tube, filtered and washed.

After being dried, we characterized the precipitates by FTIR and XRD, and observed the precipitates morphology by an optical and scanning electron microscopes. Particularly in this research, we did a semi-quantitative analysis. We chose the FTIR absorption bands at 874 cm-1 for calcite and the 856 cm-1 for aragonite, corresponding to an out-of plane bending, ѵ2, of carbonate groups.118 In experiments with Mg2+, these two bands were overlapped.

Thus, these bands were deconvoluted by non-linear fitting. The area ratio between the bands at 874 cm-1 and 856 cm-1 was calculated for each condition experimental condition. Errors were calculated by using the first-order Taylor method for propagating uncertainties considering the standard deviation associated with each area value.119

89

Chapter 5

5.3 Results 5.3.1 Overview on SOMs composition

90

SOMs were characterized by their amino acid composition and FTIR spectroscopy (Fig. 5.1). The amino acid composition was in agreement with that observed in many intra-skeletal acidic macromolecules. It was characterized by a high content of aspartic (and asparagine) and glutamic (and glutamine) residues. The carboxylate bearing residues (Asx and Glx) represented the 52.0 and the 42.5 mol % of residues in BeuSOM and LprSOM, respectively; in the latter a higher content of Ser and Gly (17.4 and 20.9) was present with respect to the former (12.9 and 17.8). The FTIR spectra showed that BeuSOM had, with respect to LprSOM, a lower absorption in the bands in zone 1 and a different structure of the bands in zone 3, which were due to methyl and methylene functional groups and glycosidic ether groups, respectively. The LprSOM also showed a stronger absorption in the bands at 1147 cm-1, which could be associated to the sulphate group.120

Figure 5.1 Left, amino acid composition of the SOMs from B. europaea and L. pruvoti. The amino acid content is reported as mol percentage. Some amino acids were not detected and some chromatographic signal could not be assigned. For this reason the sum of the amino acid percentages is lower than 100. Right, FTIR spectra from the SOMs from B. europaea (Beu) and L. pruvoti (Lpr). In the figure three zones (1-3) are highlighted, they correspond to regions where the main absorption bands due to lipids, proteins and polysaccharides, respectively, are located. The dotted line indicates an absorption band typical of sulphate groups.

5.3.2 CaCO3 precipitation in agarose viscous sols A reference experiment of CaCO3 precipitation was carried out without additives. The first precipitate appeared after a tw of 22 ± 8 hr at xo position equal to 0.62 ± 0.05. The precipitation evolved symmetrically with respect to xo and, after 14 days from the onset of the experiment, the Δ-value was 0.30 ± 0.03 (Fig. 5.2; Table 5.1). Isolated particles were observed in the agarose viscous sol under an optical microscope (Figs. 5.3-A and 5.4-A) and they were identified as calcite by XRD (Fig. 5.5-left). Calcite appeared as crystals from 75 to 200 µm long, displaying rhombohedral {10.4} faces plus less developed {hk.0} faces (Fig. 5.6-A), as already reported.121

91

Figure 5.2 Graphical representation of parameters xo, tw and Δ (Δcat+ Δan) measured in experiments of CaCO3 precipitation by CDS in conditions of SOMs-free (A) and in the presence of SOMs from B. europaea, at concentrations c (B) and 5c (C), and from L. pruvoti, at concentrations c (D) and 5c (E). The left-column refers to the agarose viscous sol experiments, medium-column to the gel experiments and right-column to the viscous sol experiments adding Mg2+ in the cation reservoir. The length of the tubs has been normalized from cation reservoir (0) to anion reservoir (1). The length of the U-tubs was 45 mm. Red and blue colours indicate the crystallization regions Δcat and Δan. Arrows indicate tw (hours, left-upper corner) and the number of replica is shown in the right-upper corner. Horizontal black lines in the middle of each figure and vertical grey lines on the arrow show the standard deviations in the measurements. Crystalline phases are calcite (C), Mg-calcite (MgC) and A (aragonite).

Chapter 5

Table 5.1 Summary of data from precipitation experiments of CaCO 3 by CDS in the absence and in the presence of SOM from B. europaea or L. pruvoti, entrapped in agarose viscous sol or gel and in the presence of Mg 2+ in the cationic reservoir. The precipitation parameters refer to measures of the mineral precipitated in the U-tube: starting point of precipitation ( xo); length of the region around x o (∆); waiting time (t w). The precipitate features refer to the minerals after removal from the agarose media.

VISCOUS SOL

92

ref.

xo *

t w **

�

phase

Beu 5c

Lpr c

Lpr 5c

ref.

Beu c

Beu 5c

VISCOUS SOL (Mg2+/Ca2+=3) Lpr c

Lpr 5c

ref.

Beu c

Beu 5c

Lpr c

Lpr 5c

0.62

0.66

0.69

0.65

0.68

0.64

0.62

0.66

0.71

0.70

0.63

0.72

0.74

0.65

0.73

(0.05)

(0.04)

(0.05)

(0.05)

(0.03)

(0.01)

(0.00)

(0.07)

(0.01)

(0.09)

(0.03)

(0.06)

(0.08)

(0.08)

(0.09)

22

32

52

37

41

31

37

51

53

51

34

50

74

64

73

(8)

(8)

(4)

(15)

(16)

(10)

(9)

(10)

(3)

(10)

(11)

(15)

(31)

(57)

(51)

0.30

0.09

0.10

0.08

0.08

0.33

0.11

0.10

0.08

0.06

0.35

0.13

0.03

0.05

0.04

(0.03)

(0.03)

(0.03)

(0.02)

(0.02)

(0.00)

(0.03)

(0.02)

(0.01)

(0.01)

(0.04)

(0.02)

(0.01)

(0.02)

(0.01)

C

C

C

C

C

C

C

C

C

C

MgC

A

MgC

A

MgC

A

MgC

A

MgC

A

r.ag.

sp.ag.

r.ag.

s.ag.

rhom.

r.ag.

s.ag.

r.ag.

sp.ag.

ac.sp.

sm.sp.

sm.sp.

sm.sp.

sm.sp.

sp.ag.

sp.ag.

sp.ag.

75-

100-

10-90

60-

15-

80-

100-

80-

100-

15-

80-

80-

30-

100-

20-

200

400

300

250

80

150

500

400

300

50

150

150

350

300

300

shape rhom.

size&

Beu c

GEL

* These values are normalized with respect to the length of the U-tube from the cation (0) to the anion reservoir (1).** The value of tw is measured in hours. Precipitated mineral phase: C, MgC and A indicate calcite, Mg-calcite and aragonite, respectively. Shape of crystals observed by SEM: rhom. indicates modified rhombohedra; r. ag. indicates agglomerates of modified rhombohedra; sp. indicates spherulites; ac. sp. indicates acicular spherulites; sm. sp. indicates spherulites with smooth surface; sp. ag. indicates agglomerates of spherulites. & indicates size distribution of precipitates measured along the main axis (µm). Standard deviations are reported within parenthesis.

5.3.3 CaCO3 precipitation in agarose viscous sols containing SOMs BeuSOM was added to agarose solutions at concentrations of c or 5c. Under these conditions, tw was 32 ± 8 hr and 52 ± 4 hr and xo was of 0.66 ± 0.04 and 0.69 ± 0.05, respectively. The precipitation evolved asymmetrically with respect to xo and roughly stopped in the position 0.60 ± 0.01 and 0.64 ± 0.02 in the cationic reservoir direction and in the position 0.70 ± 0.03 and 0.73 ± 0.02 in the anionic reservoir direction, using BeuSOM concentrations equal to c and 5c, respectively (Fig. 5.2; Table 5.1). The optical microscope pictures showed crystalline agglomerates of 100 to 400 mm, when BeuSOM c was used (Fig. 5.3-D). Increasing the concentration of BeuSOM up to 5c a continuum of particles whose sizes vary between 10 and 90 mm was observed (Fig. 5.3-G). SEM images showed that these crystalline particles consisted of interconnected and prismatic-shaped nanoparticles, which formed a microscopically layered structure (Fig. 5.7-B and D). When BeuSOM 5c was used a little distortion of the layered structure and a rounding at the edges of the particles was observed, when compared to those obtained using BeuSOM c.

Figure 5.3 Optical microscope images of crystal growing spaces (Δ) after 14 days, in the absence (A-C) and in the presence of SOMs from B. europaea, at concentration c (D-F) and 5c (G-I), and from L. pruvoti, at concentration c (J-L) and 5c (M-O). The left-column refers to the viscous sol experiments, the medium-column to the gel experiments and the right-column to the viscous sol experiments, adding Mg2+ into the cation reservoir.

93

Chapter 5

Dissolution of LprSOM at concentrations of c and 5c into the viscous sols resulted in tw values of 37 ± 15 hr and of 41 ± 16 hr and xo of 0.65 ± 0.05 and 0.68 ± 0.03 for concentrations c and 5c, respectively (Fig. 5.2 and table 5.1). In both cases the precipitation evolved symmetrically with respect to xo. Under optical microscope (Fig. 5.3-J and M) the precipitates appeared to be formed by agglomerated particles. Increasing concentration of LprSOM from c to 5c caused a decrease in the size of the particles from 60-250 µm to 15-80 µm, respectively, and sharper borders of Δ. The SEM images showed that the precipitates consisted of spherulitic particles (Fig 5.8-A and C; Fig. 5.9) with a textural organization similar to that observed in the presence of BeuSOM c (Fig. 5.7-B). Calcite was the only phase detected by XRD (Fig. 5.5-left).

94

Figure 5.4 Optical microscope images of crystals. The first column corresponds with the agarose viscous sol experiments; the second column with the agarose gel experiments and the third column with the agarose viscous sol adding Mg2+ into the cation reservoir. (A-C) Precipitates in the absence of additives; (D-F) in the presence of B. europaea at concentration equal to c and (G-I) to 5c; (J-L) in the presence of L. pruvoti at concentration equal to c and (M-O) to 5c. Scale bars: 200 µm.

Figure 5.5 XRD patterns of calcium carbonate precipitates. The upper-figure corresponds with the viscous sol experiments, the medium-figure with the gel experiments and the lower-figure with the agarose viscous sol experiments adding Mg2+ into the cation reservoir. The precipitates were obtained in the absence (A) and in the presence of organic macromolecules from B. europaea, at concentrations equal to c (B) and to 5c (C), and from L. pruvoti, at concentrations equal to c (D) and to 5c (E).

95

Chapter 5

5.3.4 CaCO3 precipitation in agarose gels containing SOMs These experiments were carried out to study the influence of increasing the degree of entanglement of agarose molecules in the calcium carbonate precipitation process. Increasing agarose concentration from 0.1% (w/v) to 0.2% (w/v) resulted in longer tw values (Table 5.1). When BeuSOM c was added xo did not differ from the reference experiment, while a significant shift towards the anionic reservoir was observed using LprSOM c (Fig. 5.2). The values of Δ and its evolution with time did not vary using gel instead of viscous sol. Only when using LprSOM c, the precipitation evolved asymmetrically with respect to xo (0.71 ±

0.01), being Δ longer toward the cationic reservoir (0.64 ± 0.02) than toward the anionic one (0.72 ± 0.01).

96

Optical microscope pictures of Δ showed that in gel the differences observed among trials were enhanced with respect to the agarose viscous sol (Fig. 5.3), especially in the presence of LprSOM. Particularly, sharper Δ borders were observed using SOMs 5c. The particles showed morphologies and size distributions similar to those observed in the agarose viscous sol (Figs. 5.6, 5.7 and 5.8). Only when LprSOM 5c was used smaller crystals (15-50 µm) were observed (Fig. 5.9). Calcite was the only phase detected by XRD (Fig. 5.5-middle).

Figure 5.6 SEM micrographs showing the morphology of CaCO3 crystals formed in agarose viscous sol in the absence of SOMs. The first row shows the CaCO3 crystals grown in agarose viscous sol (A, B); the second row, in agarose gel (C, D); and the third row, those formed in agarose viscous sol with diffusing Mg2+ (E, F).

The columns show different magnifications. The morphology of precipitates did not allow to distinguish between the two phases detected by XRD and FTIR analysis in the experiments carried out with added Mg2+.

These micrographs are representative of the whole sample populations.

97 Figure 5.7 SEM micrographs showing the morphology of crystals formed in the presence of SOMs from B. europaea.

The micrographs A, B, E, F, I and J show crystals obtained in the presence of BeuSOM c whereas images C, D, G, H, K and L show crystals obtained in the presence of BeuSOM 5c. The first row (pictures A-D) corresponds to the agarose viscous sol experiments; the second row (E-H), with the agarose gel experiments and the third row (I-L), with the agarose viscous sol experiments with diffusing Mg2+. The micrograph C and the inset show the two different morphologies of the precipitates formed in this condition. The inset of D is a high magnification of the spherulite showed in the inset of C. The observed morphologies of the precipitates obtained in the presence of Mg2+ (third row) did not allow to distinguish between MgC and A, which were detected by XRD and FTIR analysis. These images are representative of the whole sample populations (Fig. 5.9).

Figure 5.8 SEM micrographs showing the morphology of crystals formed in the presence of SOM from L. pruvoti. The micrographs A, B, E, F, I and J show crystals obtained in the presence of LprSOM c whereas images C, D, G, H, K and L show crystals obtained in the presence of LprSOM 5c. The first row

Chapter 5

(pictures A-D) corresponds to the agarose viscous sol experiments; the second row (E-H), with the agarose gel experiments and the third row (I-L), with the agarose viscous sol experiments with diffusing Mg2+. The inset in picture I shows a higher magnification of the peanut-like crystals. The observed morphologies of the precipitates obtained in the presence of Mg2+ (third row) did not allow to distinguish between MgC and A, which were the phases detected by XRD and FTIR analysis. These images are representative of the whole sample populations (Fig. 5.9).

98

Figure 5.9 Low-magnification SEM micrographs of calcium carbonate precipitates. The first column corresponds with the agarose viscous sol experiments; the second column with the agarose gel experiments and the third column with the agarose viscous sol adding Mg2+ into the cation reservoir. (A-C) Precipitates in the absence of additives; (D-F) in the presence of B. europaea at concentrations equal to c and (G-I) to 5c; (J-L) in the presence of L. pruvoti at concentrations equal to c and (M-O) to 5c.

5.3.5 CaCO3 precipitation in viscous sols containing SOMs and diffusing Mg2+ The addition of Mg2+ in the cationic reservoir (Mg2+/Ca2+ = 3) always led to an increase of tw and a shift of xo toward the anionic reservoir. Interestingly, in the presence of LprSOM c the xo value (0.65 ± 0.08) was similar to that obtained in the absence of Mg2+. In the presence of diffusing Mg2+ the length of Δ from xo to the cationic reservoir was greater than toward the anionic one (0.24 ± 0.06 and 0.11 ± 0.03, respectively). The Δ-values were shorter than those observed in the Mg2+ free experiments (Fig. 5.2; Table 5.1).

Mg2+ favoured the precipitation of large rounded and small peanuts-shaped particles (Fig. 5.3-C). The addition of SOMs brought a reduction of dc. Spherical and isolated particles were always observed when SOM c was used, while when using SOMs 5c, the particles were more aggregated and sharp Δ borders were observed, closer to the anionic reservoir (Fig. 5.3). In the latter condition, agglomerates in addition to small particles were obtained. The spherulites observed in the reference experiment were composed of hexagonal needle-shaped microcrystallites (Fig. 5.6-F). Slight morphological changes were observed by adding BeuSOM or LprSOM, while the effect of the SOM concentration was more relevant. Using SOM c the size of crystals was more homogeneous than that obtained using SOM 5c (Fig. 5.9). This observation agrees with the crystal aggregation observed in optical microscope pictures (Figs. 5.3 and 5.4). The presence of SOMs and diffusing Mg2+ made the prismatic crystals thinner, sharper and more co-oriented with respect to those obtained in the reference experiments. Using SOM 5c spherulites displaying rough surfaces (Fig. 5.7-L and Fig. 5.8L) were observed instead of the needle-shaped agglomerates. Aragonite and Mg-calcite were identified by XRD and FTIR spectroscopy in all cases (Figs. 5.5-right and 5.10-left). The quantitative Mg-calcite/aragonite mass ratio (Fig. 5.10-right) was measured analysing the FTIR spectra (Fig. 10). This ratio either increased or decreased with respect to the reference when SOM c or SOM 5c were used, respectively.

99

Chapter 5

100

Figure 5.10 (left) FTIR spectra of CaCO3 precipitated in agarose viscous sol adding Mg2+ into the cation reservoir. The precipitates were obtained in the absence (A) and in the presence of organic macromolecules extracted from B. europaea, at concentrations equal to c (B) and 5c (C), and from L. pruvoti, at concentrations equal to c (D) and 5c (E). (right) Area ratios of the 875 cm-1 and the 855 cm-1 deconvoluted bands of precipitates. This area ratio represents a rough estimation of the Mg-calcite/aragonite mass ratio. The error bars were calculated by using the first-order Taylor method for propagating uncertainties considering the standard deviations associated with each area value.

5.4 Discussion In this chapter, we use the CDS in the presence of SOMs from the solitary Mediterranean coral B. europaea or L. pruvoti, to investigate differences and similarities in calcification between zooxanthellated and azooxanthellated species. To achieve this goal, a series of in vitro crystallization trials, with two concentrations of SOMs, different viscosity of the media and in the presence of diffusing Mg2+, were carried out. In the reference experiments (i.e. the ones without SOMs entrapped in the agarose column) the first precipitates appeared in the same sites (xo), situated in the vicinity of the anionic reservoir (Table 5.1). At xo, the ion activity product has to overcome the critical value needed to induce nucleation and also this point must fulfil the equivalence rule (see chapter 2, -section 2.4). xo is displaced to the right of the U-tube (closer to HCO3-/CO32- reservoir) due to the much lower initial CO32- concentration of this solution compared to that of Ca2+ ions. The data also show that xo is barely affected by the degree of entanglement of the agarose molecules (i.e. no difference was observed between the viscous sol and the gel). This indicates that the diffusion rate of Ca2+ and carbonate species was equally affected by the different porosity of the two media.

Interestingly, the xo values did not change in the experiments in the presence of Mg2+. In this case, this apparent violation of the equivalence rule could be justified considering that Mg2+ can interact with CO32- as Ca2+ does; although with less strength (the solubility of calcium carbonate is lower than that of magnesium carbonate). Thus, the activity of CO32- interacting with Ca2+ is reduced proportionally to the Mg2+/Ca2+ and the precipitation occurred in conditions as if the activity of CO32- was lower. This hypothesis implies a longer tw, as it was indeed observed. A contribution to the increase of tw comes also from the inhibition of calcite growth due to the adsorption of Mg2+ on the calcite nuclei.64 Diffusion of Mg2+ did not change significantly the Δ -values, but affected the symmetry of the growing front. The boundaries of Δ represent the places where the activity of anions -in the zone close to the cationic reservoir- and the activity of cations -in the zone close to the anionic one- are the lowest to still sustain nucleation and growth of crystals. Here, similar Δ-values likely indicate that Mg2+ inhibits only the growth process, as also occur with pLys (chapter 3) or with Mg2+ (chapter 4). Besides, Δ may not be symmetric around xo, in this case the growing space starting from xo up to the last observed crystals in the direction to the cationic reservoir was longer to that in the direction to the anionic one. This asymmetry suggests a different range of ionic activity of cations and anions to sustain nucleation and growth, and also could be a result of the lower activity of Ca2+ compare to that in the reference.

The addition of BeuSOM or LprSOM to the agarose column increased the tw and slightly shifted the xo positions in the direction of the anionic reservoir. These effects were more marked when using the higher concentration of SOMs. The longer tw with respect to the reference experiment most likely indicated an inhibition of the nucleation and/or incipient growth processes. Since in the presence of SOMs the Δ-values were shortened, an inhibition of the nucleation event was evident. Moreover, the morphology of the crystals was influenced by the presence of SOMs, suggesting that an inhibition of the growth process was present as well. The shift of xo, which showed a trend, suggested that the presence of SOMs influenced the speciation of carbonate. Since LprSOM and BeuSOM contain acidic macromolecules characterized in their proteic regions by the presence of high percentage (almost 50 mol %) of aspartic and glutamic residues and glycosylated regions rich in sulphate groups, it can be supposed that their carboxylic group (pKa around 4.5) could release protons in the column, slightly reducing the activity of the carbonate ions in favour of that of hydrogen carbonate, but this was not observed in the presence of charged polypeptides (chapter 3). On the other hand, it is also known that SOM is composed of intrinsically disordered proteins, IDPs.122, 123 The IDPs could locally change their ability to interact with diffusing ions due to their high structural flexibility. It is also note worth that SOMs also contain glycoproteins in which the pKa changes, and therefore the ability to chelate calcium ions, with the degree of grafting.124 Finally, the presence of lipids could have also an important role in stabilizing transient amor-

101

Chapter 5

phous calcium carbonate forms.104a Since SOMs are macromolecular mixtures, to specify a role for each organic component a further detailed study would be required.

We found that when increasing the degree of entanglement and the concentration of SOMs, the Î&#x201D; borders became closer and sharper. In these conditions a lower ionic diffusion rate, a more confined space for the nucleation and growth of crystals and a higher SOMs inhibition effect, were likely present. The fact that the presence of Mg2+ made these borders even sharper suggested that Mg2+ could have a role in confining the crystallization conditions within defined calcium carbonate supersaturation values, as observed in the presence of BeuSOM or LprSOM.

102

In the presence of Mg2+, SOMs provoked co-precipitation of aragonite with Mg-calcite.

Interestingly the Mg-calcite / aragonite mass ratio was altered as a function of the concentration of SOMs; the low concentration (c) favoured the precipitation of aragonite while the high concentration (5c) favoured that of Mg-calcite (Fig. 5.10-right). This effect was more pronounced when using LprSOM. This Janus behaviour -the capability of the same family of molecules to promote and inhibit one phase- has been recently demonstrated for several additives in solution and in solid state.125 Here, this behaviour can be justified in the context of the basic principles of biomineralization. Certain SOMs molecules are able to interact with aragonite crystals, probably on specific crystalline planes. These molecules, when present in low concentration can act as nucleation sites for aragonite and / or inhibition of calcite by Mg2+, thus favouring aragonite precipitation. When the concentration is high they are able to interact with the growing nuclei and / or enhanced magnesium dehydration,126 and thus giving as net result the inhibition of the precipitation of aragonite.

An intriguing effect of the interaction between Mg2+ and SOM was the nano-scale size of crystals. SEM images (Fig. 5.7-L and Fig. 5.8-L) show a granular structure when crystals grew in viscous sols entrapping SOM and Mg2+ were diffusing along the tube. This structure is very similar to that observed for corals by Falini et al.115 in in vitro calcification experiments, by Vandermeulen and Watabe127 and Motai et al.128 in in vivo studies as well as in biominerals from different phyla.129 This granulated texture was identified as formed from amorphous calcium carbonate domains in sea urchin spicule50c and cystoliths.130 Pai and Pillai131, 132 proposed the formation of hollow triangular calcium carbonate forms from amorphous calcium carbonate spherical aggregates stabilized by the change in the conformation of a synthetic polypeptide induced by the presence of Mg2+. These observations suggested a synergic role between SOMs and Mg2+, which starts with a common amorphous precursor that later on transforms to aragonite or Mg-calcite depending on SOM concentration.105a, 106, 133

The stronger effect on crystallization parameters observed by the addition of LprSOM with respect to BeuSOM along to a higher impact on the crystal morphology cannot be only justified on the basis of the different amino acidic composition. Indeed, in several models of biomineralization a more efficient role of acidic macromolecules, as calcium carbonate crystallization modifiers, has been related to a higher content of ionisable functional groups.39d, 39e, 124, 126, 134 The used SOMs also differ in their glycosidic region structures and in the content of lipids. LprSOM has a higher degree of sulphonation along its glycosidic chains (Fig. 5.1, band at 1147 cm-1) and presents a higher content of lipids. These two features entail an additional content of acidic functional groups compared to that due only to proteins. Sulphate groups in corals are mainly localized in the skeletal textural region referred as centre of calcification,129a which represents the zone where the skeleton coral growth starts. Thus, in addition to the above-proposed effects, the favoured precipitation of Mg-calcite in the presence of LprSOM could be also due to the different structure of the polysaccharide chains as well as to a reduced activity of Mg2+ in their presence.11, 133, 135 The diverse distribution of molecules in the two SOMs, and their different impact on the precipitation of calcium carbonate could be related to the presence/absence of zooxanthellae. It has been already shown that zooxanthellate and azooxanthellate corals differ in their average amino acidic composition.112 Here, it was observed that differences are also in the content of lipids and in the structure and functionalization of polysaccharides. Zooxanthellae provide an energetic support to calcification through photosynthesis136 and it has been reported that they may influence the speciation of the inorganic carbon affecting the trafficking of protons around the nucleation site. Thus, it could be speculated that various molecular actors play a different role in the presence of photosynthesis in coral biomineralization.

5.5 Conclusions This chapter presents the study of CaCO3 precipitation in agarose viscous sol and gel hosting SOMs from two corals and diffusing Mg2+. The main results are the following: (i) the molecular composition of the two SOMs has a different impact on the crystallization parameters and morphology of CaCO3; (ii) the viscosity of the gelling media, and thus its porosity, is important in confining the precipitation environment and controlling the ion diffusion; (iii) Mg2+ have a notable role in defining specific, and sharp, limits of supersaturation under which precipitation occurs as well as in phase selection; (iv) phase distribution is affected by SOM concentration. Besides, through the use of the CDS, it was possible to carry out a first study on in vitro biomineralization of a zooxanthellate and an azooxanthellate coral.

103

Chapter 5

This work has been published with graphic design cover in CrysEngComm, Issue 7, 2014 (see M. Sancho-Tomás, S. Fermani, S. Goffredo, Z. Dubinsky, J. M. García-Ruiz, J. Gómez-Morales and G. Falini. Exploring coral biomineralization in gelling environments by means of a counter diffusion system. CrysEngComm, 2014, 16, 1257-1267)

104

06 Influence of soluble organic matrix from nacre and sea urchin spines on CaCO3 precipitation

105

6. Influence of soluble organic matrix from nacre and sea urchin spines on CaCO3 precipitation 6.1 Introduction

The nacreous layer of the mollusc shell Nautilus pompilius and the spines of the echinoderm Paracentrotus lividus (Fig. 1.13 and 1.14) are formed of aragonite and calcite, respectively, the two most common calcium carbonate (CaCO3) biominerals.1a The biomineralization processes of both occur under biological control of specific macromolecules in gelling environments. Although they occur through diverse mechanisms, a common feature is that formation of their crystalline phases takes place through amorphous calcium carbonate precursors.50b, 51c The nacre of N. pompilius is formed by a columnar brick and mortar structure.79-80 Aragonite tablets (bricks) are separated by interlamellar sheets of protein- β -chitin complex (mortar) (Fig. 1.13).6, 81 Each tablet was formed by the 3D registered assembly of nano-crystals that diffracted X-ray as a twin single crystal of aragonite. The nacre is created as follows. First, the organism forms an organic matrix framework that acts as a site for nucleation of the precursor nanoparticles. This framework consists on parallel sheets of protein- β -chitin complex filled with a silk fibroin-like gel matrix.19 The assembly of nano-crystals in an ordered array giving rise to the tablets is guided by this structure;79 see reviews e.g. [80]). The tablets grow parallel to the organic- β -sheet and perpendicular until reaching the consecutive β-sheet layer, embedded in a silk fibroin-like gel matrix.19 Sea urchin spines are composed by the 3D registered assembly of nano-crystals, in this case elongated Mg-calcite. They showed a glassy cleavage (Fig. 1.14) and diffracted X-ray as single crystals.78, 82-83 Spine growth was first studied by regenerating a spine.50b, 84 The formation mechanism of the spines of Paracentrotus lividus is influenced by calcoblasts cells, which are associated with the growing surface of the spine, and are responsible for the mineral growth occurring on the top of the spine. It was observed that an amorphous calcium carbonate (ACC) phase formed in vesicles; then, this phase deposited on the top of the spine and there the ACC precursor gradually transformed into calcite.50b

107

Chapter 6

The interaction between the OM molecules and the mineral phase during the formation of biominerals was suggested three decades ago.100 More recently it has been experimentally proved that nucleation and growth processes of many biominerals occur in sites having the features of a viscous sol (or gelling environments), due to the presence of a high concentration of acidic macromolecules making the OM. 1a, 137 This was proved in vivo not only for nacre80a and sea urchin spine50b, 138 but also for opercolum137a and coral.113-114 In this chapter, we investigated the influence of soluble organic matrix extracted from the aragonitic nacreous layer of Nautilus pompilius (NpoSOM) and from the Mg-calcitic spines of Paracentrotus lividus (PliSOM) on the precipitation of CaCO3 in an agarose gelling environment by CDS. We also study the presence of Mg2+ in an additional set of experiments, allowing Mg2+ to diffuse from the cationic reservoir.

108

6.2 Experimental section Experiments were carried out in CDS by using U-tubs. The column was filled up with agarose solutions mixed with soluble organic matrix (SOM) extracted from the septa of Nautilus pompilius (Npo) and spines of Paracentrotus lividus (Pli). Agarose final concentration in the column was 0.1 % (viscous sol) and SOM was added in two concentrations, 50 μg/mL (c) and 250 μg/mL (5c). Cationic reservoir was filled up with 0.5 mol/L solution containing a Mg2+/Ca2+= 0 or 3, where the concentration of Ca2+ was reduced to keep constant the ionic strength of the cationic reservoir solution. These solutions were prepared by mixing CaCl2.2H2O and MgCl2.6H2O (Sigma-Aldrich). Anionic reservoir was filled up with 0.5 mol/L NaHCO3 (Fluka Biochemika).

Fragments of the septa of Nautilus pompilius (Npo) and spines of Paracentrotus lividus (Pli) were suspended (1% w/v) in a sodium hypochlorite solution (3% v/v) to remove possible organic material traces from the body of organisms. The samples were air dried for one night and ground in a mortar up to obtain a fine and homogeneous powder. Five mL of milli-Q water, in which 2.5-3 g approx. of powdered biominerals were dispersed, were poured into a 40 cm-long osmotic tube for dialysis (MWCO = 3.5 kDa; CelluSep®, MFPI). The sealed tube was placed into 1 L of 0.1M CH3COOH (Riedel de Haen) solution under stirring. The decalcification proceeded for 72 hr. At the end, the tube containing the dissolved OM was dialysed against milli-Q water (resistivity 18.2 MΩ cm at 25 °C; filtered through a 0.22 μm membrane) until the final pH was about 6. The obtained aqueous solution containing the OM was centrifuged at 6000 rpm for 6 minutes to separate the soluble (SOM) and the insoluble (IOM) OM fractions, which were then lyophilized. SOMs were characterized by using FTIR spectroscopy.

When first precipitates appeared in each U-tube we measured xo, tw and the length of the crystal growing space (D) 14 days after. Precipitates were recovered by the U-tube, filtered and washed. After being dried, precipitates were characterized by FTIR and XRD, and observed by an OM and a SEM. Particularly in this chapter, Raman micro-spectroscopy was also used. First, we observed the crystals under the Raman-microscope (10x) and then, collected the Raman spectrum of several crystals from each condition when Mg2+ was added. Raman peaks at 155, 280, 713 and 1087 cm-1 are assignable to calcite while peaks at 150, 205, 701 and 1085 cm-1, to aragonite. Amorphous phases are detected by a broadening of all peaks, mainly a broad peak around 1085 cm-1 and a broad hump around a 150-300 cm-1.49 For further details of the methodology employed or the measured parameters description, see chapter 2.

6.3 Results 6.3.1 CaCO3 precipitation.

A reference experiment of CaCO3 precipitation was carried out without additives. The first precipitates appeared after a tw of 22 ± 8 hr at the position equal to 0.62 ± 0.05. The precipitation evolved symmetrically with respect to xo and after 14 days from the onset of the experiment, Δ was 0.30 ± 0.03 (Fig. 6.1; Table 6.1). Isolated crystals were observed under an optical microscope (Figs. 6.2A and 6.3A) and they were identified as calcite by XRD analyses (Fig. 6.7). Calcite appeared as single crystals of sizes between 75 and 200 μm and displaying rhombohedral {10.4} faces plus less developed {hk.0} m faces (Fig. 6.4A-C), as already reported.121a The reference experiment is the same previously described in chapter 5.

6.3.2 CaCO3 precipitation containing NpoSOM and PliSOM. The NpoSOM was added to the gelling column at concentrations of c or 5c. Using the same experimental set-up of the reference experiment, tw was of 20 ± 10 hr and 24 ± 4 hr, while xo was of 0.69 ± 0.03 and 0.65 ± 0.03, respectively. The evolution of precipitation was slightly asymmetrical with respect to xo and roughly stopped in the position 0.64 ± 0.01 and 0.63 ± 0.01 in the cationic reservoir direction and in the position xo equal to 0.73 ± 0.01 and 0.71 ± 0.03 in the anionic reservoir direction, when using c and 5c NpoSOM, respectively (Fig. 6.1; Table 6.1). In the presence of c NpoSOM, optical microscope pictures showed that in the precipitation zone (Δ) crystalline aggregates with sizes between 80 and 400 μm were formed

109

Chapter 6

(Fig. 6.3C). When 5c NpoSOM was used, the sizes of the observed aggregates oscillate between 40 and 350 μm (Fig. 6.3E).

110

Figure 6.1 Graphical representations of the measured parameters in the precipitation experiments of calcium carbonate carried out by counter-diffusion system. Left-column and right-column refers to the experiments without and with Mg2+ in the cation reservoir, respectively. The parameters were measured in the absence (A) and in the presence of organic macromolecules from N. pompilius, at concentrations equal to c (B) and to 5c (C), and from P. lividus, at concentrations equal to c (D) and to 5c (E). The length of the tubs was 45 mm. Red and blue colors indicate Δan and Δcat. Arrows indicate the waiting time (tw, hours, left-upper corner). The number of replica is shown in the right-upper corner. Horizontal black lines in the middle of each figure show the standard deviations in the measurements. Polymorphs were also indicated as calcite (C), Mg-calcite (MgC) and A (aragonite).

Table 6.1. Summary of the data obtained from precipitation experiments of calcium carbonate by CDS in the absence and presence of SOM from N. pompilius or P. lividus, entrapped in agarose viscous sol in the absence and presence of Mg 2+ in the cationic reservoir. The precipitation parameters refer to measures of the mineral precipitated in the U-tube: starting point of precipitation ( xo); length of the region around xo (∆); waiting time (t w). The precipitate features refer to the minerals after removal from the agarose matrix.

Mg2+/Ca2+=0

ref.

xo *

t w **

Npo c

Npo 5c

Mg2+/Ca2+=3

Pli c

Pli 5c

ref.

Npo c

Npo 5c

Pli c

Pli 5c

0.62

0.69

0.65

0.65

0.66

0.63

0.59

0.69

0.65

0.66

(0.05)

(0.03)

(0.03)

(0.04)

(0.01)

(0.03)

(0.05)

(0.09)

(0.13)

(0.08)

22

20

24

23

22

34

30

36

36

50

(8)

(10)

(4)

(3)

(4)

(11)

(18)

(24)

(15)

(17)

0.30

0.09

0.08

0.11

0.08

0.35

0.10

0.09

0.11

0.08

(0.03)

(0.01)

(0.03)

(0.02)

(0.01)

(0.04)

(0.03)

(0.01)

(0.02)

(0.03)

phase

C

C

C

C

C

MgC A

A

MgC A

A MgC

A MgC

shape

rhom.

r.ag.

r.ag.

r.ag.

r.ag.

sm.sp.

ac.sp.

�

size&

ac.sp. sp.ag.

ac.sp. sm.sp. pe.

sm.sp. sp.ag.

75-

80-

40

100-

75-

80-

100-

50-

80-

150-

200

400

350

550

500

150

200

200

300

500

* These values are normalized with respect to the length of the U-tube from the cation (0) to the anion reservoir (1). The standard deviations in the measurements is reporter in parentheses. ** The tw is measured in hours (standard deviations in the measurements). Precipitated mineral phase: C, MgC and A indicate calcite, Mg-calcite and aragonite, respectively. Shape of crystals observed by SEM: rhom. indicates modified rhombohedra; r. ag. indicates aggregates of modified rhombohedra; sp. indicates spherulites; ac. sp. Indicates acicular spherulites; sm. sp. Indicates spherulites with smooth surface; sp. ag. Indicates aggragates of spherulites; pe. Indicates peanut-shape. & Size distribution of precipitates measured along the main axis (µm).

111

Chapter 6

SEM pictures (Fig. 6.5A-F) showed that in the presence of entrapped NpoSOM only aggregates of rhombohedral crystals of calcite formed and that the aggregation increased with the concentration of NpoSOM. When 5c NpoSOM was present, crystals appeared to be formed by the assembly of elongated sub-micrometer particles (Fig. 6.5F), and this effect was not evident when c NpoSOM was used (Fig. 6.5C).

112

Figure 6.2.

Optical microscope pictures of crystal growing spaces (Î&#x201D;) after 14 days, in the absence (A, B) and in the presence of soluble organic matrix from N. pompilius, at concentration c (C, D) and 5c (E, F), and from P. lividus, at concentration c (G, H) and 5c (I, J). (A), (C), (E), (G) and (I), and (B), (D), (F), (H) and (J) refer to the experiments carried out without and with the presence of Mg2+ diffusing from the cationic reservoir, respectively.

Dissolution of PliSOM into the agarose solutions gave a tw of 23 ± 3 hr and of 22 ± 4 hr, and a xo of 0.65 ± 0.04 and of 0.66 ± 0.01 when using a concentration equal to 50 μg/mL (c) or 250 μg/mL (5c), respectively. The precipitation evolved slightly asymmetrically with respect to xo and stopped in the position 0.60 ± 0.02 and 0.61 ± 0.01 in the cationic reservoir direction and in the position 0.71 ± 0.01 and 0.69 ± 0.01 in the anionic reservoir direction, using PliSOM concentrations equal to c and 5c, respectively (Fig. 6.1; Table 6.1). Under the optical microscope (Figs. 6.2 and 6.3G) the precipitates appeared formed by aggregated particles. The increase of PliSOM concentration from c to 5c provoked a narrowing of Δ and a small decrease of the aggregates size from 100-550 µm to 75-500 µm, respectively.

Figure 6.3 Optical microscope images of precipitates. The first column corresponds with the experiments in the absence of Mg2+ while the second column corresponds with the experiments in the presence of Mg2+ in the cation reservoir. In the absence (A, B) and in the presence of soluble organic macromolecules from N. pompilius, at concentration c (C, D) and 5c (E, F), and from P. lividus, at concentration c (G, H) and 5c (I, J).

Scale bars: 200 µm.

113

Chapter 6

In figure 6.6A-F are reported SEM micrographs showing CaCO3 crystals formed in the presence of entrapped PliSOM, using c PliSOM (Fig. 6.6A-C) and 5c PliSOM (Fig. 6.6D-F).

In these conditions only modified rhombohedral crystals formed. Their aggregation state increased with concentration of PliSOM but the shape and size of the building nano-blocks (Fig. 6.6C and F) seems not to be affected by the PliSOM concentration. Calcite was the only phase detected by XRD analysis (Fig. 6.7).

114

Figure 6.4 SEM pictures showing the morphology of calcium carbonate crystals precipitated in the agarose viscous sol in the absence of SOMs. (A-C) show different magnifications of a calcite crystal grown in absence of Mg2+. In them the rhombohedral {10.4} faces of calcite were indicated together with those {hk.0} due to the interaction with agarose molecules. (D-I) images a different magnification of spherulites formed in the presence of diffusing Mg2+ from the cation reservoir. (D-F) show a spherulite of magnesium calcite. In the image at high magnification (F) the crystallographic faces typical of magnesium calcite crystals are indicated. (G-I) show a spherulite of aragonite. In (I) the hexagonal needle-like crystal of aragonite {001} capped are shown. These pictures are representative of the entire populations of crystals.

Figure 6.5 SEM pictures showing calcium carbonate crystals precipitated in the presence of entrapped NpoSOM. (A-F) images of crystals obtained in the absence of diffusing Mg2+ from the cationic reservoir, using c NpoSOM (A-C) and 5c NpoSOM (D-F). In these conditions only calcite crystals formed. Their aggregation state increased with the concentration of NpoSOM. When 5c NpoSOM was present the crystals appeared formed by the assembly of elongated sub-micrometer particles (F).

(G-O) images of crystals obtained in the presence of diffusing Mg2+ from the cationic reservoir, using NpoSOM. In the presence of c NpoSOM (G-I) only aragonite precipitated.

The needle-like crystals lost the crystalline morphology observed in the absence of SOM and appeared entrapped in the NpoSOM/agarose matrix. When 5c NpoSOM was present magnesium calcite co-precipitated with aragonite and two typologies of spherulite were observed. In one case (J-L) the crystalline units making the spherulite were formed by the assembly of nano-particles and were rhombohedral capped (L). In the other case (M-O) the spherulite was made by the assembly of irregular needle-like shapes that resembled a poor (N), or lost completely (O), the crystalline morphology; in both cases the needle-like shapes were formed by nano-spheroidal particles. The particles showed were representative of the whole sample populations.

115

Chapter 6

6.3.3 CaCO3 precipitation in agarose viscous sols containing SOMs and diffusing Mg2+.

116

Addition of Mg2+ to the cation reservoir (Mg2+/ Ca2+ = 3) always provoked an increase of tw. When there was no added SOM, xo was very similar to that of the previous reference experiment (0.63 ± 0.03) and to that found in the presence of PliSOMs (0.65 ± 0.13 when using c and 0.66 ± 0.08 when the SOM concentration was 5c). Notably, xo value in the presence of c NpoSOM appeared closer to the cationic reservoir (0.59 ± 0.05) while at concentration 5c it shifted toward the anionic reservoir (0.69 ± 0.09), as compared to the xo value of the reference experiment. In the presence of Mg2+, Δ was asymmetric being longer from xo toward the cationic reservoir than toward the anionic one. All the Δ values were similar to those observed in Mg2+-free experiments (Fig. 6.1; Table 6.1).

Diffusion of Mg2+ through the agarose column led to the precipitation of big rounded and small peanuts-shaped particles, which appeared close to each other and partially overlapped (Fig. 6.3B). In figures 6.4D-I images at different magnification of spherulites are shown. Figures 6.4D-F show a spherulite, most probably of magnesium calcite. In the image at high magnification (Fig. 6.4F) the crystallographic faces typical of magnesium calcite crystals are indicated. Fig. 6.4 G-I shows a spherulite, most probably of the aragonite phase. In figure 6.4I the hexagonal needle-like crystals of aragonite {001} capped are shown.

When SOMs were added, we observed spherical and isolated particles and not well-defined Δ borders. Only when using 5c NpoSOM, particles appeared more integrated and defined sharp Δ borders were observed. In this last condition, spherical aggregates with small peanut-shaped particles precipitated. Morphological changes were observed when using NpoSOM or PliSOM with respect to the reference.

Figure 6.6 SEM pictures showing calcium carbonate crystals formed in the presence of entrapped PliSOM. (A-F) images of crystals obtained in the absence of diffusing Mg2+ from the cationic reservoir, using c PliSOM (A-C) and 5c PliSOM (D-F). In these conditions only calcite crystals formed. (G-R) images of particles obtained in the presence of diffusing Mg2+ from the cationic reservoir using PliSOM. In the presence of c PliSOM (G-L) magnesium calcite co-precipitated with aragonite.

Two families (G-I and J-L) of spherulites were observed. Between them differences were clearly observable only at high magnifications. In one case (I) plate-like crystals were observed where in the other one (L) needle-like crystals appeared. Their crystalline morphologies were poor and irregular. When 5c PliSOM was present magnesium calcite and aragonite precipitated. At low magnification all the spherulites showed a similar morphology and were randomly aggregated (M and P). At high magnification at least two different scenarios were observed. In one case the surface (O), or a fractured region of the spherulite (N) showed the presence of nano-spheroidal particles preferential aligned in one direction making needle-like structure. In the other case the presence of nano-particles rhombohedral capped was observed (Q and R). The particles showed were representative of the whole sample populations.

In figures 6.5 G-O images of particles obtained in the presence of diffusing Mg2+ from the cationic reservoir using NpoSOM are shown. In the presence of c NpoSOM (Fig. 6.5 G-I) the needle-like crystals partially lost the crystalline morphology observed in the absence of SOM

117

Chapter 6

and appeared entrapped in the NpoSOM/agarose matrix. When 5c NpoSOM was present two typologies of spherulite were observed. In one case (Fig. 6.5J-L) the crystalline units making the spherulite were formed by the assembly of nano-particles and were rhombohedral capped (Fig. 6.5L). In the other case (Fig. 6.5 M-O) the spherulite was made by the assembly of irregular shapes preferentially aligned along a preferential direction that resembled a poor (Fig. 6.5 N), or lost completely (Fig.6.5 O) crystalline morphology; in both cases the assembled shapes were formed by nano-spheroidal particles.

118

Figure 6.7 XRD patterns of calcium carbonate precipitates. Left-figure corresponds with the experiments in absence of Mg2+; and right-figure, in the presence of Mg2+. Precipitates were obtained in the absence (A) and in the presence of NpoSOM, at concentrations equal to c Îźg/mL (B) and to 5c (C), and from PliSOM at concentrations equal to c (D) and to 5c (E).

Figure 6.8 FTIR spectra of calcium carbonate precipitated in a viscous sol adding Mg2+ into the cation reservoir. The precipitates were obtained in the absence (A) and in the presence of soluble organic macromolecules from N. pompilius, at concentrations equal to c (B) and to 5c (C), and from P. lividus, at concentrations equal to c (D) and to 5c (E).

In figures 6.6 G-R images of particles obtained in the presence of diffusing Mg2+ from the cationic reservoir using PliSOM were shown. Two families (Fig. 6.6G-I and 6.6J-L) of spherulites were observed in the presence of c PliSOM. Between them, differences were clearly observable only at high magnifications. In one case (Fig. 6.6I) plate-like crystals were observed whereas needle-like crystals appeared in the other one (Fig. 6.6L). Their crystalline morphologies were poor and irregular. When 5c PliSOM was present at low magnification all the spherulites showed a similar morphology and were randomly aggregated (Fig. 6.6M and 6.6P). At high magnification at least two different scenarios were observed. In one case the surface (Fig. 6.6O), or a fractured region of the spherulite (Fig. 6.6N), showed the presence of nano-spheroidal particles aligned preferentially in one direction, thus making needle-like structures. In the other case the presence of rhombohedral capped nanoparticles was observed (Fig. 6.6Q and 6.6R).

Figure 6.9 Raman spectra of CaCO3 particles formed in the presence of diffusing Mg2+ from the cationic reservoir. (ref.) indicates particles formed in the absence of soluble organic matrix. (c NpoSOM) and (5c NpoSOM) indicate particles formed in the presence of soluble organic matrix from N. pompilius, at concentrations c and to 5c, respectively. (c PliSOM) and (5c PliSOM) indicate particles formed in the presence of soluble organic matrix from P. lividus at concentrations c and to 5c, respectively. The upper-figure reports spectra from particles in which the main bands were associated to calcite. The reference spectrum of calcite was characterized by bands at 155, 280, 713 and 1087 cm-1. All peaks are broader than the corresponding peaks in reference calcite, implying a disordered structure (Raz et al., 2003). The lower-figure reports spectra from particles in which the main bands were associated to aragonite.

The reference spectrum of aragonite was characterized by bands at 150, 205, 701 and 1085 cm-1. The two families of particles were undistinguishable under optical microscope and each particle was identified as calcitic or aragonitic only after acquiring of the Raman spectrum.

119

Chapter 6

Aragonite and Mg-calcite were identified by XRD and FTIR spectroscopy (Figs. 6.7 and 6.8), except for NpoSOM at concentration c, condition at which only aragonite was detected. Raman spectra showed aragonite and calcite crystals in all conditions (Fig. 6.9). All precipitates were inspected by optical microscopy before collecting the Raman spectrum, but pictures did not reveal morphological differences between phases. Interestingly, few calcite crystals were detected when adding c NpoSOM, in spite that only aragonite was identified by means of XRD and FTIR analyses.

120

Figure 6.10 left Table reporting the amino acid composition (mol %) of NpoSOM and PliSOM according to [121c]. right Fourier Transform Infrared Spectra of NpoSOM and PliSOM. The main absorption bands are indicated.

6.4 Discussion Several studies reported the chemical features of NpoSOM and PliSOM.121b-d They were mainly composed of acidic glycoproteins and/or proteoglycans and had as main feature a high content of acidic residues in their protein regions, about 25 and 30 mol% of Asx and Glx in PliSOM and NpoSOM, respectively (Fig. 6.10).121c The influence of NpoSOM and PliSOM on the in vitro precipitation of CaCO3 was widely explored and many aspects of their role in vivo were clarified. SOMs can control the CaCO3 polymorphism,14a the mechanical properties,139 or as reported for other biominerals also the morphology, size and orientation of the crystallites (e.g. [121d, 140]). In addition, they can also sculpture CaCO3 crystals in diverse way, according to their composition (e.g. [121d]). Also, it was shown that SOM stabilized

the formation of ACC, which was a precursor of the crystalline phases.11, 129b Moreover, it was also shown that they guide the formation of biominerals according to synthetic paths that did not follow the classical theories of nucleation and growth.141 However, despite this qualified and excellent amount of information, some further aspects can be clarified by using the information obtainable from the CDS, in particular those related to the supersaturation conditions for precipitation and the critical concentrations of Ca2+ and CO32- for nucleation and growth. To achieve this goal, by comparative studies we have seen that speciation and further ions diffusion are not roughly affected by the presence of charged polypeptides, at least at the used concentrations (see chapter 3). It might be that water diffusion from the reservoirs to the agarose viscous sol could reduce the sol viscosity upon long periods of time. However, we observed at the end of the experiment that features of the viscous sol did not change.

In the presence of either NpoSOM or PliSOM, and considering that the reproducibility of CDS experiments viscous sols is poor than in gels (Table 6.1), the measured crystallization parameters showed a xo-shift toward the anionic reservoir, a narrowing of Î&#x201D; and almost no variation of tw. The SOMs were rich in carboxylate and sulphate groups and had a strong capability to chelate calcium ions. They could affect the diffusion of calcium ions, or could also change the speciation of inorganic carbon in favour of HCO3-. Only the latter effect agreed with the observed xo-shift, according to previous observations in which the presence of entrapped polypeptides did not roughly affect the ionic diffusion (see chapter 3). This hypothesis was supported also by the observation that the Ca2+/SOM molar ratio is at least > 200, and if we consider the chelating behaviour of SOMs similar to that of poly-aspartate, thus the amount of Ca2+ chelated by SOMs molecules might have been very low.

Being confident on the data, in the presence of NpoSOM or PliSOM, unlike what it was observed using synthetic charged polypeptides (see chapter 3) or intra-skeletal coral macromolecules (chapter 5), the narrowing of Î&#x201D; was not associated to longer tw. In the CDS, xo is the point where the system is supersaturated for nucleation and the equality rule condition is fulfilled (see 2.4.1). Thus, NpoSOM or PliSOM allowed the precipitation of CaCO3 only in specific, and confined, conditions of supersaturation with respect to both calcium and carbonate ions. That means more restrictive conditions of supersaturation exerted by SOMs in the mineralization of the sea urchin spine and nacre than those supposed for the formation of coral skeletons. This hypothesis is supported by palaeontology data, where mineralogical changes in coral fibers have been reported in corals through time whereas no equivalent shift has been cited concerning echinoderm skeletons and nacreous tablets.142

121

Chapter 6

NpoSOM and PliSOM were adsorbed on growing particles and were potentially capable to induce the nucleation (102, 137b and references therein). This capability, and the fact that the precipitation process occurred under control of SOMs, was confirmed by the change of morphology of the primary units forming the precipitates (Figs. 6.5 and 6.6). In them, mainly when we used the highest concentration of SOM, the presence of nano-sized spheroidal grains was observed. The shape of these primary grains changed with the type of SOM, indicating interactions along specific crystallographic planes.100

122

Presence of diffusing Mg2+ had an inhibition effect on the precipitation of CaCO3. Addition of Mg2+ increased the tw; but, interestingly, this value was not significantly affected by the presence of SOMs. It is worth noting that diffusion of both Mg2+ and Ca2+ from the cationic reservoir towards the reaction zone must occur simultaneously. However, although both cations diffused simultaneously, they may not have done with the same rate. The high hydration energy of Mg2+ could hinder its diffusion in comparison to that of Ca2+. Therefore, the Mg2+/Ca2+ in the precipitation region could be smaller than that in the reservoir.

Magnesium calcite and aragonite were detected in the final mineral particles (Fig. 6.7) when Mg2+ diffused in the agarose column. The effect of SOMs concentration on the polymorphic selection showed its capability to promote or inhibit the appearance of one phase; a low concentration of NpoSOM led to the precipitation of aragonite while much more Mg-calcite precipitated at high NpoSOM concentration. A similar effect was observed when we used SOMs extracted from coral skeletons (see chapter 5). PliSOM favoured the precipitation of magnesium calcite with respect to the control, almost independently from its concentration.

An interesting observation was that particles precipitated in the presence of Mg2+ showed always almost similar shapes irrespective from the associated polymorph, as confirmed by the Raman analyses (Fig. 6.9). Moreover, it was observed that the Raman bands widened when compared with the reference material (Fig. 6.9) and that the micro/nano-structure of the spherules revealed that they were formed by the assembly of micro/nanoparticles (Figs.

6.5 and 6.6). It is known that a cooperative effect subsisted between SOM and Mg2+ in stabilizing ACC.105a, 106 Moreover, it was shown that the formation of sea urchin spine and nacre proceeds through CaCO3 polyamorphs.50b, 51c Thus, the above observations could fit with the formation of proto-ACC particles stabilized by magnesium ions and SOM that then evolved in a given crystalline phase depending on the associated SOM.143 Indeed, in the presence of NpoSOM, which was extracted from the aragonitic nacre, only the formation of aragonite was observed; while in the presence of PliSOM, which was extracted from the calcitic sea urchin spine, the formation of magnesium calcite was favoured.

6.5 Conclusions In this research we have explored the effects of SOMs extracted from two biominerals -nacre from N. pompilius and spines from P. lividus- on the precipitation of CaCO3 by using a CDS where the U-tube was filled with an agarose viscous sol. The results show that the presence of these SOMs do not increase the supersaturation needed for precipitation, in contrast to what synthetic polypeptides and SOMs extracted from corals do, but limit the supersaturation conditions to yield precipitation. The experiments carried out in the presence of diffusing Mg2+ suggest the formation of a transient ACC phase only when SOMs are present.

We also found that crystallization of aragonite or calcite depends on the biomineral specie from which the SOM was extracted and also on its concentration. In conclusion, this study suggests that the control exerted by SOMs in the mineralization of the sea urchin spine and nacre is under more restrictive conditions of supersaturation than those supposed for the formation of coral skeletons.

This work has been accepted in European Journal of Mineralogy. See María Sancho-Tomás, Simona Fermani, Jaime Gómez-Morales, Giuseppe Falini, and Juan Manuel García-Ruiz. “Calcium carbonate bio-precipitation in counter-diffusion systems using the soluble organic matrix from nacre and sea urchin spine”, DOI: 10.1127/0935-1221/2014/00262389.

123

07 Fluorescence to monitor pH-changes in counterdiffusion experiments

7. Future perspectives in counter-diffusion studies: monitoring diffusion processes by fluorescence 7.1 Introduction

In this chapter is presented a research where for the first time the pH variations, during precipitation of CaCO3 in a counter-diffusion experiment by using a pH-sensible fluorescence dye, is monitored in real time and with space resolution. First, I will briefly introduce the fundamental principles of the fluorescence to better understand the potentiality of the fluorescence to detect pH-changes in a diffusion-reaction-precipitating system.

Fluorescence is the property of some atoms and molecules, generally polyaromatic hydrocarbons or heterocycles called fluorophores or fluorescent dyes, to absorb light at a particular wavelength and to subsequently emit light of longer wavelength after a brief interval, termed the fluorescence lifetime. Fluorescence is the result of a three-stage process illustrated by the Jablonski diagram shown in Figure 7.1.

Figure 7.1 Jablonski diagram showing the stages involved in the process of fluorescence.

127

Chapter 7

128

During the excitation stage (1) a photon of energy hνEX is supplied by an external source such as an incandescent lamp or a laser and absorbed by the fluorophore, creating an excited electronic singlet state (S1’). The difference with chemiluminescence is that in this last process the excited state is populated by a chemical reaction. The excited-state lifetime (2) is typically 1-10 nanoseconds. During this time, the fluorophore undergoes conformational changes and is also subject to a multitude of possible interactions with its molecular environment. These processes have two important consequences. First, the energy of S1’ is partially dissipated, yielding a relaxed singlet excited state (S1) from which fluorescence emission originates. During the stage of fluorescence emission (3) photon of energy hνEM is emitted, returning the fluorophore to its ground state S0. Due to energy dissipation during the excited-state lifetime, the energy of this photon is lower, and therefore of longer wavelength, than the excitation photon hνEX. The difference in energy or wavelength represented by (hνEX – hνEM) is called the Stokes shift. The Stokes shift is fundamental to the sensitivity of fluorescence techniques because it allows emission photons to be detected against a low background, isolated from excitation photons.

Our study consisted in mixing a fluorescent dye (pH-sensor) with agarose gel, placing the mixture in a U-tube column and used it to measure the pH variations versus time and space accompanying ion diffusion, reaction and precipitation of CaCO3. Two solutions of NaHCO3 and CaCl2 diffused against each other through the gelling column of the U-tube to form CaCO3 precipitates. We developed an instrumental set-up for measuring the fluorescence signal emitted by the fluorophore, which is an indirect measure of pH variations.

Next in this chapter we present the used instrumental setup as well as a description and discussion of the results. In summary, this chapter shows the preliminary experiments combining counter-diffusion and pH fluorescence dyes and the potentiality of this technique to monitor biomineralization processes in vitro.

The work was performed in collaboration with Prof. Marco Montalti and Dr. Damiano Genovese of the Photochemical Nanosciences Laboratory group of University of Bologna.

7.2 Experimental section Experiments were carried out by the counter-diffusion technique using a U-tube system (Fig. 7.2). Reservoirs containing 0.5M solutions of the two reagents, sodium hydrogen carbonate (Fluka Biochemika) and calcium chloride (CaCl2.2H2O, Sigma-Aldrich), were

separated by an agarose gel column, allowing diffusion of two components into the gel from opposite ends. All solutions were prepared with ultrapure water (0.22 μS, 25°C, MilliQ©, Millipore).

Figure 7.2 The U-tube set-up used for experiments carried out with the counter-diffusion technique (a). A fluorescence video frame showing a bottom view of the same tube (b). The red dashed line marks the pixels used to map the pH across the U-tube, located in the center of the U-tube.

7.2.1 Preparation of agarose gelling solutions mixed with pH-fluorescent dye An agarose stock solution of 0.3% (w/v) was heated up to 90 °C for 20 minutes to completely dissolve the agarose powder (Agarose D-5, Hispanagar). The solution was then cooled to 55 °C and diluted with preheated milli-Q water under continuous stirring and constant temperature (55°C) in order to obtain a final agarose concentration of 0.2% (w/v). This solution was divided into three aliquots (samples ref., c and 5c). Soluble organic matrix (SOM) from coral skeleton was added at two different concentrations to samples sample c (50 μg/ml) and 5c (250 μg/ml). The SOM was extracted from powdered coral skeletons of scleractinian coral (Balanophyllia europaea) by a decalcification process (--see section 2.5). Pure agarose was used for reference experiments (sample ref). A small amount (50 μL in 1 mL water) of a stock solution of the pH-fluorescence dye or pH-sensor (8-hydroxypirene-1,3,6-trisulfonic acid trisodium salt (SF); Aldrich H1529) was added to the three samples to reach a final sensor concentration of 6 μmol/L. Then, after thoroughly mixing the warm agarose solutions for 1-2 minutes, they were transferred to three U-tubes and allowed to slowly cool down and jellify. After 24 hours the three U-tubes were placed in a holder and used for fluorescence measurements.

129

Chapter 7

7.2.2 Fluorescence measurements and analysis All quantitative analysis were performed on videos recorded with a custom-assembled instrument represented in figure 7.4. Briefly, the sample tube was excited with a homogeneous illumination beam from a 500W Xenon arc lamp selected with a bandpass filter (bandpass 450nm, Oriel Corp. Scorpio Optics), while the fluorescence signal was selected with a cutoff filter (cut-off 495nm Oriel Corp. Scorpio Optics, to collect fluorescence and get rid of the excitation light, see also figure 7.2) and collected by a CCD camera (Basler scA640-70gc) equipped with a M1214-MP Computer objective (12mm, F/1.4), each pixel yielding a fluorescence intensity signal in a precise spatial point to be elaborated for further analysis. All videos were analysed by means of custom developed Matlab code.

130

Figure 7.3 Absorbance and emission spectra of the pH sensor SF versus pH. Absorbance and emission spectra at acidic pH (pH=5) are marked in red while green curves are the spectra at basic pH (pH=9). Emission spectra are collected by exciting at lexc=450nm.

The blue rectangle indicates the excitation bandpass used for the experiments (450nm), while the green rectangle indicates the spectral region reaching the CCD through a cut-off filter (495nm).

Two references were taken to correct for non-ideal illumination conditions: an average of first 100 frames was taken as the reference for non-homogeneities in the beam profile, and used to normalize the rest of the video frames, assuming that beam profile has no dependence on time. To correct for instabilities of the lamp power (oscillation of total intensity in time), a region in the tube displaying constant fluorescence (very close to the anionic reservoir, constantly basic pH) was taken as the reference for the maximum emission intensity in the rest of the images.

Figure 7.4 Scheme of instrumental setup for fluorescence video acquisition.

All displayed analyses were carried out on the central row of pixels of the U-tube (red dashed line in figure 7.2). Averaging the analysis on several lines or on the whole width of the tube provides only slight noise smoothing, and leads to loss of information and resolution on individual crystallization spots.

Figure 7.5 Dependence of fluorescence intensity of pH sensor SF: experimental points (black dots) and fitting function (dashed red line).

131

Chapter 7

After correction, intensity data were converted into pH. To do so, SF was titrated in PBS upon addition of HCl or NaOH. Fluorescence intensity was monitored with a Perkin Elmer LS50 Fluorometer versus pH, which was measured with a Jenway 4330 pH-meter equipped with a Crison sensor. The dependence of fluorescence intensity on pH was fitted with a sigmoid function (I=a â &#x201E; (1+e-(pH-pKa)/b) and the fitting parameters (a=1.02, b=0.45 and pKa=6.99) were then used in the inverse equation to convert intensity data into pH (Fig. 7.5). Black noise (amounting to <5% of the maximum pixel intensity) was measured and subtracted from intensity data before conversion. As stated before, intensity of every pixel in each frame was normalized respect to the pixels displaying constant high intensity, i.e. the region of the tube closer to the carbonate reservoir.

132

7.3 Results and Discussion Firstly, we must say that the experiments discussed in this section have been repeated several times with different experimental conditions, and in the last round we finally reached the optimal experimental conditions. Further experiments are presently being carried out to prove reproducibility and to extend, in these last conditions, the observations that will be described as follows. We divided this section into three parts: diffusion, nucleation and crystal growth.

Figure 7.6 Contour maps showing the variations of pH along the U-tube (abscissa, only the central part of the tube is displayed, i.e. the central 3.5 cm over the total tube length of 4.5cm) versus time (ordinate axis, top to

bottom). From left to right, U-tubes containing samples in the absence (a), and in the presence of SOM c (b) and 5c (c). Blue and red arrows indicate the reservoir positions of NaHCO3 and CaCl2 respectively.

Dashed lines with white arrows put in evidence the evolution in time of the front of pH~8. Dotted circles mark the regions and the times where nucleation occurs, as proved by following crystal formation (visible as green-blue stripes at the same abscissa, at later times -towards the bottom of the map).

7.3.1 Diffusion The initial pH of the agarose column is around 6 (red-orange in color scale). Once reagents solutions are added into the reservoirs at the corners of the U-tube, diffusion of ions starts and pH increases upon arrival of HCO3-/CO32- ions (color shifts from red to dark blue, starting from the anionic reservoir, left-upper corner of the map). It can be noted that HCO3-/ CO32- ions move toward the cationic reservoir and this movement seems to be slower when SOM is added (Fig. 7.6-b, c), the higher the concentration of SOM, the slower is the carbonate diffusion through the gel at the initial stages (< 6 hours of experiment). This observation could be explained if macromolecules would affect ion diffusion or if the presence of SOM would influence the speciation of carbonate. As we have suggested in previous chapters, carboxylic groups of some acidic macromolecules of SOM may release protons in the gelling medium, reducing the activities of the HCO3-/CO32- ions in favor of those of HCO3-/H2CO3 (see figure 7.7). This fact, together with the diffusion of Ca2+ in the opposite direction, produces an acidification of the gelling column.

Figure 7.7 Carbonate speciation as a function of pH showing the common pH range in nature.

This second hypothesis is more in agreement with experimental data showed in the upper part of Fig. 7.6, obtained at the initial stages. The presence of SOM displays a more acidic

133

Chapter 7

medium (stronger red colour) comparing to that in the reference. After this initial stage HCO3-/CO32- ions diffusion is faster.

7.3.2 Nucleation

134

Crystal nucleation can be identified indirectly by monitoring (i) the trajectory of the front around pH=8 and (ii) the fluorescence signal coming from crystal (which is measurable after few hours from the supposed nucleation). Before the measurable appearance of crystals, we observe a retreat of the high pH region towards the hydrogen carbonate reservoir. Since carbonate ions are “feeding” the new nuclei during their nucleation and growth, their local concentration will decrease, thus, leading to a decrease of pH. In all cases, nucleation occurs closer to the hydrogen carbonate reservoir because of much reduced carbonate concentration respect to that of hydrogen carbonate and Ca2+ (see chapter 2, figure 2.4). We can observe this effect in the dotted circles in Figure 7.6 or in the sequence of images in time in Figure 7.8. Both prove that the region where crystal nucleation occurs shifts toward the anionic reservoir. Following the trajectory of the front of pH~8 (see dashed lines with white arrows in Figure 7.6) we will find these zones, where the supersaturation rule and the equality range condition (see chapter 2, section 2.4.1) are met for CaCO3 nucleation. At the same time, diffusion of HCO3-/CO32- ions continues along the tube (color turns from red/orange to blue till the right edge of the tube).

Figure 7.8 Fluorescence images at four different times (after 25, 33, 46 and 72 hours from experiment starting, from top to bottom) revealing the distribution of pH sensor in the “ON” state, i.e. in basic pH, dispersed in agarose during precipitation of CaCO3. From left to right, U-tubes

containing samples ref (a), c (b) and 5c (c). Blue and red thick arrows indicate the position of the reservoirs of NaHCO3 and CaCl2, respectively. Note that the pH sensor not only maps the pH along the tube, but also serves to identify growing crystals, which are clearly visible as bright spots due to scattering of light and to some accumulation of fluorescent sensor in the crystal structure (red thin arrows). Elaboration of such images versus time provides the pH map in figure 7.5 (see experimental methods for procedure).

7.3.3 Crystal Growth Crystals are visible in the maps as spots with seemingly higher pH (green or blue), an effect that might be due to strong scattering of light from the crystals or to some accumulation of the fluorescence sensor in the crystal structure. It is in fact well known that agarose interferes with calcite121a with possible entrapment into the crystal, yielding higher concentration of fluorescence sensor and thus seemingly higher pH. The addition of SOM to the gel produces three clear effects in the CaCO3 precipitation: (i) precipitation occurs at higher pH, (ii) is slower and (iii) is confined in a much thinner and more well defined region than in the control. When SOM was present, the pH of the crystal growing space (∆) remains at higher pH than in the reference (marked by green colour in b and c, while orange colour emerges in a -Fig. 7.6). In a parallel experiment followed by an optical microscope (Fig. 7.9), we could observe how precipitation limits in the presence of SOM (b and c) are much better defined than in the reference (a).

Very interestingly, monitoring the pH during crystal growth reveals depletion of basic species (CO32-, HCO3-) around the growing crystals, which can be observed as a region of pH depletion, clearly visible in pH maps in figure 7.6, as a spreading orange-red wave of higher pH around the crystals. Such depletion starts from the crystal and “erodes” carbonates (thus high pH) especially where they are not rapidly replaced from a reservoir, thus towards the CaCl2 reservoir, even though in the time of our experiment it never reached the further edge near CaCl2 reservoir.

It is believed that organic matrix has a key role in biomineralization since it slows down the rate of formation and/or growth of mineral phases, allowing new features to appear and to increase the complexity of resulting biomineral structures. This role clearly arises in our experiments, as can be seen by comparing the pH maps in function of the amount of SOM dispersed in agarose: when SOM is not present in the gel, pH is depleted to acidic values down to 6 in the time of the end of the experiment (ca 3 days). For increasing concentration

135

Chapter 7

of SOM (c and 5c) pH is depleted much more slowly, and as a result the pH is more evenly distributed along the U-tube. Thus, while in absence of SOM the strong pH depletion only allows growth of already nucleated crystals, at high SOM concentration growth is slow enough for the pH depletion to be balanced by the diffusion of protons, resulting in a quite even distribution of pH across the U-tube, thus in the appearance of multiple nucleation regions, similar to a Liesegang stripes pattern.

136

Figure 7.9 Optical Microscopy images of calcite crystals at different times forming in the U-tubes in absence of SOM (a) and in presence of SOM in concentrations respectively c (b) and 5c (c). NaHCO3 reservoir is on the left side. The images span around 1 cm in proximity of the first nucleation point, whose position is marked as the origin of the x axis, and correspond to the central part of the fluorescence images reported in figure 7.6 and in the maps in figure 7.5. The plots below each image show the profiles of pixel intensity, yielding an approximate measure of crystal density profiles.

7.4 Conclusions In conclusion, we have done a preliminary study combining two techniques -counter-diffusion and fluorescence- in order to investigate the precipitation of CaCO3 in the presence of coral macromolecules and measure the pH in real time and space, through analysis of the intensity of a fluorescent sensor. Therefore, the use of fluorescence sensors appears as an amazing tool to study the role of additives in mineral precipitation. Further work would consider developing these combined techniques not only for pH, but also for other ions (Ca2+ or Mg2+), or even for different macromolecules implicated in biomineral formation by using specific sensors. Besides, it would be interesting to perform this study in the actual place of mineralization in organisms, since it could be the key for assessing the real role to pH variations, biomineral involved ions and diverse macromolecules functions.

137

08 Summary and Conclusions

139

Chapter 1

140

8. Summary and Conclusions

This PhD project was focused on CaCO3 mineralization in gelling environments (agarose viscous sols or agarose gels) by means of counter-diffusion as a model to study biomineralization processes. The concept behind the choice of this precipitation environments comes from the observation that almost ubiquitary in organisms the mineralization process occurs in sites rich of framework macromolecules having gelling proprieties.

The first part of the PhD Thesis was devoted to the definition of several experimental parameters that can be measured during a precipitation process in the counter-diffusion system. They are associated to the mineral nucleation and/or growth processes. These parameters are the waiting time (tw), the starting point of precipitation (xo) and the crystal growing space (Î&#x201D;). The boundaries of D are xcat in direction toward the cationic reservoir and xan in direction to the anionic deposit. Î&#x201D;cat and Î&#x201D;an correspond to the spaces from xo to xcat and to xan, respectively.

Having defined the crystallization parameters, the counter-diffusion methodology was first proved using synthetic polypeptides such as poly-L-aspartate and poly-L-glutamate.

Both polypeptides were chosen as synthetic models of mineralizing proteins containing high amounts of acidic residues, aspartic and glutamic, which are widely found in the organic matrices of biominerals. For the sake of comparison, poly-L-lysine was used as a basic polypeptidic counterpart. In some of the experiments, Mg2+ were added in the precipitation system because are a major ionic component of seawater. Its effect, combined with those of organic additives, may give hints on the composition of the precipitation fluid at the mineralization sites and its response to changes of environmental factors.

Once established its validity and applicability, the counter-diffusion methodology was applied to study the influence of soluble organic matrices (SOMs) from three biominerals: coral skeletons, nacre and sea urchin spines. These biominerals were chosen because they

141

Summary and Conclusions

are made of aragonite or calcite, the most common polymorphs of CaCO3. Corals exert a low degree of biological control over the formation of their skeleton, which indicates that this process could be also influenced by environmental factors, i.e. seawater composition.

In contrast, a high level of biological control occurs over the formation of nacre in molluscs and sea urchin spines in echinoderms.

The specific conclusions of the research performed in this PhD project can be summarized as follows:

Influence of entrapped synthetic polypeptides.

142

By applying the precipitation protocol and comparing the measured parameters, it was possible to establish the inhibitor or promoter capacity of pAsp, pGlu and pLys over the CaCO3 nucleation and/or growth steps. We observed that pLys only influences the growth mechanisms whereas pAsp and pGlu influence both, the nucleation and growth processes.

Influence of diffusing Mg2+ and entrapped synthetic polypeptides.

By exploring the combined role of diffusing Mg2+/Ca2+ solutions and entrapped charged polypeptides (pAsp, pGlu and pLys) we found that Mg2+ seem to influence only the growth mechanism of CaCO3 and not the nucleation. Thus, organisms might use them as a crystal habit modifier. We also detected calcite at high Mg2+/Ca2+ ratio. However, this only occurred in the presence of pAsp, and pGlu in minor extension. It means that both negatively charged polypeptides slightly favour the precipitation of calcite even at high Mg2+/Ca2+.

Influence of entrapped SOMs from two corals, degree of entanglement and diffusing Mg2+.

The main conclusions of using entrapped SOMs from the solitary Mediterranean corals Balanophyllia europaea (BeuSOM), zooxanthellate, and Leptopsammia pruvoti (LprSOM), azooxanthellate, are the following: (i) the molecular composition of the two SOMs has a different impact on the crystallization parameters and morphology of CaCO3; (ii) the vis-

cosity of the gelling media, and thus its porosity, is important in confining the precipitation environment and controlling the ion diffusion; (iii) Mg2+ have a notable role in defining specific, and sharp, limits of supersaturation under which precipitation occurs as well as in phase selection; (iv) phase distribution is affected by SOM concentration.

Influence of entrapped SOMs from nacre and sea urchin spines, and diffusing Mg2+.

The main conclusions of using entrapped SOMs extracted from biominerals of 2 different phyla -nacre from N. pompilius and spines from P. lividus- on the precipitation of CaCO3 are the following: (i) SOMs do not increase the waiting time with respect to the reference, what suggests a same critical supersaturation for precipitation, in contrast to what synthetic polypeptides and SOMs extracted from corals do. However, they shortened the crystal growing space (Î&#x201D;), indicating a shorter range where the equality rule is fulfilled. The study suggests that the control exerted by SOMs in the mineralization of sea urchin spine and nacre is under more restrictive conditions of supersaturation than those supposed for the formation of coral skeletons. (ii) The results of experiments carried out in the presence of diffusing Mg2+ suggest the formation of a transient ACC phase only when SOMs are present. (iii) The crystallization of aragonite or calcite depends on the biomineral specie from which SOM was extracted as well as on its concentration.

Monitoring pH evolution by using fluorophores Finally and as a further perspective in counter-diffusion studies, we monitored the diffusion process by using a fluorescence pH-sensor. In collaboration with the Photochemical Nanosciences Laboratory group of Bologna University we developed an experimental setup for coupling optical fluorescence and counter-diffusion techniques. This last chapter allows us to move from qualitative estimation of crystallization parameters to experimental quantitative data concerning the variation of pH accompanying ions diffusion in real time and with space resolution during precipitation of CaCO3 in a counter-diffusion experiment. This study supports the hypothesis that coralâ&#x20AC;&#x2122;s SOM really affects the pH of the medium before and after nucleation.

The importance of this Thesis is not limited to the results described above. It shows that the precipitation in gelling environments by counter-diffusion is a powerful tool for understanding the processes of nucleation and growth of ionic crystals in the presence of additives.

Although this methodology has been applied to the study of biomineralization processes, its potentialities are wider and allow scientists to study also the crystallization processes in systems of relevance in the geological and environmental fields and for the design and synthesis of new materials.

143

09 References

9. References

147 1. (a) Lowenstam, H. A.; Weiner, S., On Biomineralization. 1989; (b) Mann, S., Biomineralization: Principles and Concepts in Bioinorganic Materials Chemistry.

Oxford University Press: 2001.

2. Lowenstam, H., Minerals formed by organisms. Science 1981, 211 (4487), 1126-1131.

3. Mann, S., Mineralization in biological systems. In Inorganic Elements in Biochemistry, Springer Berlin Heidelberg: 1983; Vol. 54, pp 125-174.

4. Weiner, S.; Addadi, L., An overview of biomineralization processes and the problem of the vital effect. In Reviews in mineralogy and geochemistry, 2003; Vol. 54, pp 1-29.

5. Krampitz, G.; Graser, G., Molecular Mechanisms of Biomineralization in the Formation of Calcified Shells. Angewandte Chemie International Edition in English 1988, 27 (9), 1145-1156.

6. Watabe, N., Studies on shell formation. XI. Crystal-Matrix relationships in the inner layers of molusk shells. Journal of ultrastructure research 1965, 12, 351-70.

7. Meldrum, F. C., Calcium carbonate in biomineralisation and biomimetic chemistry. International Materials Reviews 2003, 48, 187-224.

8. (a) Young, S. D., Organic material from scleractinian coral skeletonsâ&#x20AC;&#x201D;I. Variation in composition between several species. Comparative Biochemistry and Physiology Part B: Comparative Biochemistry 1971, 40 (1), 113-120; (b) Weiner, S.,

References

Apartic Acid-Rich Proteins: Major Components of the Soluble Organic Matrix of Mollusk Shells. Calcified Tissue International 1979, 29 (2), 163-167; (c) Constantz, B.; Weiner, S., Acidic Macromolecules Associated with the Mineral Phase of Scleractinian Coral Skeletons. Journal of Experimental Zoology 1988, 248 (3), 253-258.

9. (a) Wada, K., Studies on the Mineralization of the Calcified Tissue in Mollusc --X. Histochemical Determination of the Nature of Acid Mucopolysaccharide in Organic Crystals. Bulletin of the Japanese Society of Scientific Fisheries 1964, 4 (30), 993-998; (b) Dauphin, Y., Comparative studies of skeletal soluble matrices from some Scleractinian corals and Mollusc. International Journal of Biological Macromolecules 2001, 28 (4), 293-304; (c) Cuif, J. P.; Dauphin, Y.; Doucet, J.; Sa148

lome, M.; Susini, J., XANES mapping of organic sulfate in three scleractinian coral skeletons. Geochimica et Cosmochimica Acta 2003, 67 (1), 75-83.

10. Farre, B.; Dauphin, Y., Lipids from the nacreous and prismatic layers of two Pteriomorpha mollusc shells. Comparative biochemistry and physiology. Part B, Biochemistry & molecular biology 2009, 152 (2), 103-9.

11. Weiner, S.; Addadi, L., Crystallization Pathways in Biomineralization. Annual Review of Materials Research 2011, 41 (1), 21-40.

12. Carter, P. W.; Mitterer, R. M., Amino acid composition of organic matter associated with carbonate and non-carbonate sediments. Geochimica et Cosmochimica Acta 1978, 42 (8), 1231-1238.

13. Crenshaw, M. A., The soluble matrix from Mercenaria mercenaria shell. Biomineralization 1972, 6, 6-11.

14. (a) Falini, G.; Albeck, S.; Weiner, S.; Addadi, L., Control of Aragonite or Calcite Polymorphism by Mollusk Shell Macromolecules. Science 1996, 271 (5245), 67-69; (b) Belcher, A. M.; Wu, X. H.; Christensen, R. J.; Hansma, P. K.; Stucky, G. D.; Morse, D. E., Control of crystal phase swtiching and orientation by soluble mollusc-shell proteins. Nature 1996, 381 (6577), 56-58.

15. Suzuki, M.; Saruwatari, K.; Kogure, T.; Yamamoto, Y.; Nishimura, T.; Kato, T.; Nagasawa, H., An Acidic Matrix Protein, Pif, Is a Key Macromolecule for Nacre Formation. Science 2009, 325 (5946), 1388-1390.

16. Sรถllner, C.; Burghammer, M.; Busch-Nentwich, E.; Berger, J.; Schwarz, H.;

Riekel, C.; Nicolson, T., Control of Crystal Size and Lattice Formation by Starmaker in Otolith Biomineralization. Science 2003, 302 (5643), 282-286.

17. Hernández-Hernández, A.; Gómez-Morales, J.; Rodríguez-Navarro, A. B.; Gautron, J.; Nys, Y.; García-Ruiz, J. M., Identification of Some Active Proteins in the Process of Hen Eggshell Formation†. Crystal Growth & Design 2008, 8 (12), 4330-4339.

18. Weiner, S.; Traub, W., X-ray diffraction study of the insoluble organic matrix of mollusk shells. FEBS Letters 1980, 111 (2), 311-316.

19. Levi-Kalisman, Y.; Falini, G.; Addadi, L.; Weiner, S., Structure of the nacreous organic matrix of a bivalve mollusk shell examined in the hydrated state using cryo-TEM. Journal of structural biology 2001, 135 (1), 8-17.

20. Mass, T.; Drake, J. L.; Haramaty, L.; Rosenthal, Y.; Schofield, O. M. E.; Sherrell, R. M.; Falkowski, P. G., Aragonite Precipitation by “Proto-Polyps” in Coral Cell Cultures. PLoS ONE 2012, 7 (4), e35049.

21. Colfen, H.; Mann, S., Higher-order organization by mesoscale self-assembly and transformation of hybrid nanostructures. Angewandte Chemie (International ed. in English) 2003, 42 (21), 2350-65.

22. (a) Kashchiev, D., Nucleation. Basic theory with applications. Butterworth-Heinemann: 2000; (b) Volmer, M., Kinetik der phasenbildung. 1939; (c) Becher, R.; Döring, W., Kinetic treatment of grain-formation in super-saturated vapours. Annalen der Physik 1935, 24, 719-752; (d) De Yoreo, J. J.; Vekilov, P. G., Principles of Crystal Nucleation and Growth. Reviews in Mineralogy and Geochemistry 2003, 54 (1), 57-93.

23. Van Driessche, A. E. S. In situ observation of protein crystal growth by advanced optical techniques. Granada, 2007.

24. Cölfen, H.; Antonietti, M., Mesocrystals and nonclassical crystallization. John Wiley & Sons: 2008.

25. Wulff, G., On the question of speed of growth and dissolution of crystal surfaces. Z. Kristallogr 1901, 34, 449-530.

149

References

26. (a) Hartman, P.; Perdok, W. G., On the relations between structure and morphology of crystals. I. Acta Crystallographica 1955, 8 (1), 49-52; (b) Hartman, P.; Perdok, W. G., On the relations between structure and morphology of crystals. II. Acta Crystallographica 1955, 8 (9), 521-524; (c) Hartman, P.; Perdok, W. G., On the relations between structure and morphology of crystals. III. Acta Crystallographica 1955, 8 (9), 525-529.

27. (a) Burton, W. K.; Cabrera, N.; Frank, F. C., The Growth of Crystals and the Equilibrium Structure of their Surfaces. Philosophical Transactions of the Royal Society of London. Series A, Mathematical and Physical Sciences 1951, 243 (866), 299-358; (b) Stranski, I., The theory of crystal growth. Z. Phys. Chem 150

1928, 136 ((3-4)), 259-278.

28. Cölfen, H., Bio-inspired Mineralization Using Hydrophilic Polymers. In Biomineralization II, Naka, K., Ed. Springer Berlin Heidelberg: 2007; Vol. 271, pp 1-77.

29. Penn, R. L.; Banfield, J. F., Morphology development and crystal growth in nanocrystalline aggregates under hydrothermal conditions: Insights from titania. Geochimica et cosmochimica acta 1999, 63 (10), 1549-1557.

30. (a) Rousseau, M.; Lopez, E.; Stempflé, P.; Brendlé, M.; Franke, L.; Guette, A.; Naslain, R.; Bourrat, X., Multiscale structure of sheet nacre. Biomaterials 2005, 26 (31), 6254-6262; (b) Oaki, Y.; Imai, H., Nanoengineering in Echinoderms: The Emergence of Morphology from Nanobricks. Small 2006, 2 (1), 66-70.

31. (a) Gower, L. A.; Tirrell, D. A., Calcium carbonate films and helices grown in solutions of poly(aspartate). Journal of Crystal Growth 1998, 191 (1–2), 153160; (b) Gower, L. B., Biomimetic model systems for investigating the amorphous precursor pathway and its role in biomineralization. Chemical reviews 2008, 108 (11), 4551-4627.

32. (a) Henisch, H. K., Crystal Growth in Gels. 1970; (b) Henisch, H. K., Crystals in Gels and Liesegang Rings. 1988.

33. Fick, A., Ueber diffusion. Annalen der Physik 1855, 170 (1), 59-86.

34. Crank, J., The mathematics of diffusion. Oxford university press: 1979.

35. (a) Van Driessche, A. E. S.; Otálora, F.; Gavira, J. A.; Sazaki, G., Is Agarose an Impurity or an Impurity Filter? In Situ Observation of the Joint Gel/Impurity Effect on Protein Crystal Growth Kinetics. Crystal Growth & Design 2008, 8 (10), 3623-3629; (b) Fernandez-Diaz, L.; Putnis, A.; Prieto, M.; Putnis, C. V., The role of magnesium in the crystallization of calcite and aragonite in a porous medium.

Journal of Sedimentary Research 1996, 66 (3), 482-491.

36. Otálora, F.; Gavira, J. A.; Ng, J. D.; García-Ruiz, J. M., Counterdiffusion methods applied to protein crystallization. Progress in Biophysics and Molecular Biology 2009, 101 (1–3), 26-37.

37. (a) Putnis, A.; Fernandez-Diaz, L.; Prieto, M., Experimentally produced oscillatory zoning in the (Ba, Sr) SO4 solid solution. Nature 1992, 358 (6389), 743-745; (b) García-Ruíz, J.; Mínguez, F., Condiciones de formación del primer precipitado en la técnica del gel de sílice. Estudios geológicos 1982, 38 (1), 3-14.

38. (a) Li, H.; Estroff, L. A., Calcite Growth in Hydrogels: Assessing the Mechanism of Polymer-Network Incorporation into Single Crystals. Advanced Materials 2009, 21 (4), 470-473; (b) Kellermeier, M.; Cölfen, H.; García-Ruiz, J. M., Silica Biomorphs: Complex Biomimetic Hybrid Materials from “Sand and Chalk”. European Journal of Inorganic Chemistry 2012, 2012 (32), 5123-5144.

39. (a) Grassmann, O.; Neder, R. B.; Putnis, A.; Löbmann, P., Biomimetic control of crystal assembly by growth in an organic hydrogel network. American Mineralogist 2003, 88 (4), 647-652; (b) Kosanović, C.; Falini, G.; Kralj, D., Mineralization of calcium carbonates in gelling media. Crystal growth & design 2010, 11 (1), 269-277; (c) Silverman, L.; Boskey, A. L., Diffusion Systems for Evaluation of Biomineralization. Calcified Tissue International 2004, 75 (6), 494-501; (d) AsenathSmith, E.; Li, H.; Keene, E. C.; Seh, Z. W.; Estroff, L. A., Crystal Growth of Calcium Carbonate in Hydrogels as a Model of Biomineralization. Advanced Functional Materials 2012, 22 (14), 2891-2914; (e) Dorvee, J. R.; Boskey, A. L.; Estroff, L. A., Rediscovering hydrogel-based double-diffusion systems for studying biomineralization. CrystEngComm 2012, 14 (18), 5681-5700.

40. Iafisco, M.; Marchetti, M.; Gómez Morales, J.; Hernández-Hernández, M. a.

A.; García Ruiz, J. M.; Roveri, N., Silica Gel Template for Calcium Phosphates Crystallization. Crystal Growth & Design 2009, 9 (11), 4912-4921.

151

References

41. Sánchez-Pastor, N.; Gigler, A. M.; Cruz, J. A.; Park, S.-H.; Jordan, G.; Fernández-Díaz, L., Growth of Calcium Carbonate in the Presence of Cr(VI). Crystal Growth & Design 2011, 11 (7), 3081-3089.

42. Iler, R. K., The chemistry of silica: solubility, polymerization, colloid and surface properties, and biochemistry. 1979.

43. Mitchell, J. R.; Blanshard, J. M. V., RRheological properties of alginate gels.

Journal of Texture Studies 1976, 7 (2), 219-234.

44. Rochas, C.; Lahaye, M., Average molecular weight and molecular weight dis152

tribution of agarose and agarose-type polysaccharides. Carbohydrate Polymers 1989, 10 (4), 289-298.

45. (a) Henisch, H. K.; García-Ruiz, J. M., Crystal growth in gels and Liesegang ring formation: I. Diffusion relationships. Journal of Crystal Growth 1986, 75 (2), 195-202; (b) Prieto, M.; Putnis, A.; Fernández-Díaz, L.; López-Andrés, S., Metastability in diffusing-reacting systems. Journal of Crystal Growth 1994, 142 (1–2), 225-235; (c) Putnis, A.; Prieto, M.; Fernandez-Diaz, L., Fluid supersaturation and crystallization in porous media. Geological Magazine 1995, 132 (01), 1-13; (d) García-Ruiz, J. M., Nucleation of protein crystals. Journal of structural biology 2003, 142 (1), 22-31.

46. (a) Kirov, G., On the diffusion method for growing crystals. Journal of Crystal Growth 1972, 15 (2), 102-106; (b) Pucar, Z.; Pokric, B.; Graovac, A., Precipitation in gels under conditions of double diffusion. Critical concentrations of the precipitating components. Analytical Chemistry 1974, 46 (3), 403-409; (c) Henisch, H. K.; García-Ruiz, J. M., Crystal growth in gels and Liesegang ring formation: II. Crystallization criteria and successive precipitation. Journal of Crystal Growth 1986, 75 (2), 203-211; (d) Prieto, M.; Fernández-Diaz, L.; López-Andrés, L.; López-Andrés, S., Supersaturation evolution and first precipitate location in crystal growth in gels; application to barium and strontium carbonates. Journal of crystal growth 1989, 98 (3), 447-460.

47. Garcia-Carmona, J.; Gomez-Morales, J.; Fraile-Sainz, J.; Rodriguez-Clemente, R., Morphological characteristics and aggregation of calcite crystals obtained by bubbling CO2 through a Ca(OH)(2) suspension in the presence of additives. Powder Technology 2003, 130 (1-3), 307-315.

48. Falini, G.; Fermani, S.; Gazzano, M.; Ripamonti, A., Polymorphism and architectural crystal assembly of calcium carbonate in biologically inspired polymeric matrices. Journal of the Chemical Society, Dalton Transactions 2000, (21), 3983-3987.

49. Addadi, L.; Raz, S.; Weiner, S., Taking advantage of disorder: amorphous calcium carbonate and its roles in biomineralization. Advanced Materials 2003, 15 (12), 959-970.

50. (a) Beniash, E.; Aizenberg, J.; Addadi, L.; Weiner, S., Amorphous calcium carbonate transforms into calcite during sea urchin larval spicule growth. Proceedings of the Royal Society of London. Series B: Biological Sciences 1997, 264 (1380), 461-465; (b) Politi, Y.; Arad, T.; Klein, E.; Weiner, S.; Addadi, L., Sea urchin spine calcite forms via a transient amorphous calcium carbonate phase.

Science 2004, 306 (5699), 1161-1164; (c) Politi, Y.; Metzler, R. A.; Abrecht, M.; Gilbert, B.; Wilt, F. H.; Sagi, I.; Addadi, L.; Weiner, S.; Gilbert, P. U. P. A., Transformation mechanism of amorphous calcium carbonate into calcite in the sea urchin larval spicule. Proceedings of the National Academy of Sciences 2008, 105 (45), 17362-17366; (d) Killian, C. E.; Metzler, R. A.; Gong, Y.; Olson, I. C.; Aizenberg, J.; Politi, Y.; Wilt, F. H.; Scholl, A.; Young, A.; Doran, A., Mechanism of calcite co-orientation in the sea urchin tooth. Journal of the American Chemical Society 2009, 131 (51), 18404-18409.

51. (a) Weiss, I. M.; Tuross, N.; Addadi, L.; Weiner, S., Mollusc larval shell formation: amorphous calcium carbonate is a precursor phase for aragonite. Journal of Experimental Zoology 2002, 293 (5), 478-491; (b) Marxen, J. C.; Becker, W.; Finke, D.; Hasse, B.; Epple, M., Early mineralization in Biomphalaria glabrata: microscopic and structural results. Journal of Molluscan Studies 2003, 69 (2), 113-121; (c) Nassif, N.; Pinna, N.; Gehrke, N.; Antonietti, M.; Jäger, C.; Cölfen, H., Amorphous layer around aragonite platelets in nacre. Proceedings of the National Academy of Sciences of the United States of America 2005, 102 (36), 1265312655; (d) Shechter, A.; Berman, A.; Singer, A.; Freiman, A.; Grinstein, M.; Erez, J.; Aflalo, E. D.; Sagi, A., Reciprocal changes in calcification of the gastrolith and cuticle during the molt cycle of the red claw crayfish Cherax quadricarinatus. The Biological Bulletin 2008, 214 (2), 122-134.

52. (a) Levi-Kalisman, Y.; Raz, S.; Weiner, S.; Addadi, L.; Sagi, I., X-Ray absorption spectroscopy studies on the structure of a biogenic “amorphous” calcium carbonate phase. Journal of the Chemical Society, Dalton Transactions 2000, (21),

153

References

3977-3982; (b) Vinogradov, A. P., The elementary chemical composition of marine organisms. Yale University, 1953; (c) Dillaman, R.; Hequembourg, S.; Gay, M., Early pattern of calcification in the dorsal carapace of the blue crab, Callinectes sapidus. Journal of Morphology 2005, 263 (3), 356-374; (d) Neues, F.; Ziegler, A.; Epple, M., The composition of the mineralized cuticle in marine and terrestrial isopods: a comparative study. CrystEngComm 2007, 9 (12), 1245-1251; (e) Raz, S.; Testeniere, O.; Hecker, A.; Weiner, S.; Luquet, G., Stable amorphous calcium carbonate is the main component of the calcium storage structures of the crustacean Orchestia cavimana. The Biological Bulletin 2002, 203 (3), 269-274; (f) Becker, A.; Bismayer, U.; Epple, M.; Fabritius, H.; Hasse, B.; Shi, J.; Ziegler, A., Structural characterisation of X-ray amorphous calcium carbonate (ACC) in 154

sternal deposits of the crustacea Porcellio scaber. Dalton Transactions 2003, (4), 551-555; (g) Becker, A.; Ziegler, A.; Epple, M., The mineral phase in the cuticles of two species of Crustacea consists of magnesium calcite, amorphous calcium carbonate, and amorphous calcium phosphate. Dalton Transactions 2005, (10), 1814-1820; (h) Chave, K. E., Aspects of the biogeochemistry of magnesium 1.

Calcareous marine organisms. The Journal of Geology 1954, 62 (3), 266-283.

53. (a) Mahamid, J.; Sharir, A.; Addadi, L.; Weiner, S., Amorphous calcium phosphate is a major component of the forming fin bones of zebrafish: Indications for an amorphous precursor phase. Proceedings of the National Academy of Sciences 2008, 105 (35), 12748-12753; (b) Beniash, E.; Metzler, R. A.; Lam, R. S.; Gilbert, P., Transient amorphous calcium phosphate in forming enamel. Journal of structural biology 2009, 166 (2), 133-143.

54. (a) Gago-Duport, L.; Briones, M.; Rodriguez, J.; Covelo, B., Amorphous calcium carbonate biomineralization in the earthwormâ&#x20AC;&#x2122;s calciferous gland: pathways to the formation of crystalline phases. Journal of structural biology 2008, 162 (3), 422-435; (b) Lee, M. R.; Hodson, M. E.; Langworthy, G., Earthworms produce granules of intricately zoned calcite. Geology 2008, 36 (12), 943-946.

55. Sethmann, I.; WĂśrheide, G., Structure and composition of calcareous sponge spicules: a review and comparison to structurally related biominerals.

Micron 2008, 39 (3), 209-228.

56. Voinescu, A.-E. Biomimetic formation of CaCO3 particles showing single and hierarchical structures. 2008.

57. Raven, J.; Caldeira, K.; Elderfield, H.; Hoegh-Guldberg, O.; Liss, P.; Riebe-

sell, U.; Shepherd, J.; Turley, C.; Watson, A., Ocean acidification due to increasing atmospheric carbon dioxide. The Royal Society: 2005.

58. Elderfield, H., Carbonate mysteries. Science 2002, 296 (5573), 1618-1621.

59. (a) Meldrum, F. C.; Cölfen, H., Controlling Mineral Morphologies and Structures in Biological and Synthetic Systems. Chemical Reviews 2008, 108 (11), 4332-4432; (b) Sommerdijk, N. A.; With, G. d., Biomimetic CaCO3 mineralization using designer molecules and interfaces. Chemical reviews 2008, 108 (11), 4499-4550.

60. Peter, K., Calcium Carbonate Scale Control in Industrial Water Systems. In The Science and Technology of Industrial Water Treatment, CRC Press: 2010; pp 39-60.

61. Elderfield, H.; Ganssen, G., Past temperature and δ18O of surface ocean waters inferred from foraminiferal Mg/Ca ratios. Nature 2000, 405 (6785), 442-445.

62. da Silva, J. F.; Williams, R., The inorganic chemistry of life. Clarendon Press, Oxford: 1991.

63. Simkiss, K., The processes of biomineralization in lower plants and animals-an overview. Biomineralization in lower plants and animals 1986, 30, 19-37.

64. (a) Lippmann, F., Sedimentary Carbonate Minerals: Minerals, Rocks and Inorganic Materials. Springer-Verlag Berlin, 1973; (b) Falini, G.; Fermani, S.; Tosi, G.; Dinelli, E., Calcium Carbonate Morphology and Structure in the Presence of Seawater Ions and Humic Acids. Crystal Growth & Design 2009, 9 (5), 2065-2072.

65. Schroeder, J. H.; Dwornik, E. J.; Papike, J. J., Primary Protodolomite in Echinoid Skeletons. Geological Society of America Bulletin 1969, 80 (8), 1613-1616.

66. Falini, G.; Gazzano, M.; Ripamonti, A., Magnesium calcite crystallizatin from water–alcohol mixtures. Chem. Commun. 1996, (9), 1037-1038.

67. (a) Spalding, M. D.; Ravilious, C.; Green, E. P., World atlas of coral reefs. Univ of California Press: Berkeley, 2001; (b) Cohen, A. L.; McConnaughey, T. A., Geochemical Perspectives on Coral Mineralization. Reviews in Mineralogy and Geochemistry 2003, 54 (1), 151-187.

155

References

68. Marshall, A. T.; Clode, P., Calcification rate and the effect of temperature in a zooxanthellate and an azooxanthellate scleractinian reef coral. Coral Reefs 2004, 23 (2), 218-224.

69. Tambutté, S.; Holcomb, M.; Ferrier-Pagès, C.; Reynaud, S.; Tambutté, É.; Zoccola, D.; Allemand, D., Coral biomineralization: From the gene to the environment. Journal of Experimental Marine Biology and Ecology 2011, 408 (1–2), 58-78 (and references therein).

70. Hoegh-Guldberg, O.; Mumby, P.; Hooten, A.; Steneck, R.; Greenfield, P.; Gomez, E.; Harvell, C.; Sale, P.; Edwards, A.; Caldeira, K., Coral reefs under rapid 156

climate change and ocean acidification. science 2007, 318 (5857), 1737-1742.

71. http://oceanservice.noaa.gov/education/tutorial_corals/coral01_intro.html, N. O. a. A. A. N. 72. Barnes, D. J., Coral Skeletons: An Explanation of Their Growth and Structure.

Science 1970, 170 (3964), 1305-1308.

73. Cuif, J. P.; Dauphin, Y., The two-step mode of growth in the scleractinian coral skeletons from the micrometre to the overall scale. Journal of Structural Biology 2005, 150 (3), 319-331.

74. Koch, G. v., Über die Entwicklung des Kalkskeletes von Asteroides calycularis und dessen morphologische Bedeutung. Mittheilungen aus der Zoologischen Station zu Neapel 1882, 3, 284-292.

75. Johnston, I. S., The ultrastructure of skeletogenesis in hermatypic corals. Int Rev Cytol 1980, 67, 171-214.

76. Bryan, W. H.; Hill, D., Spherulitic crystallization as a mechanism of skeletal growth in the hexacorals. Proceedings of the Royal Society of Queensland 1941, 52 (9), 78-91.

77. Carpenter, W. B., On the microscopic structure of shells. London, 1844.

78. Schmidt, W., Die Bausteine des Tierkorpers im polarisierten Licht. Bonn: 1924.

79. Nakahara, H., An electron microscope study of the growing surface of nacre in two gastropod species, Turbo cornutus and Tegula pfeifferi. Venus 1979, 38 (3), 105-211.

80. (a) Heinemann, F.; Launspach, M.; Gries, K.; Fritz, M., Gastropod nacre: Structure, properties and growthâ&#x20AC;&#x201D;Biological, chemical and physical basics. Biophysical chemistry 2011, 153 (2), 126-153; (b) Sun, J.; Bhushan, B., Hierarchical structure and mechanical properties of nacre: a review. Rsc Advances 2012, 2 (20), 7617-7632.

81. (a) Weiner, S.; Talmon, Y.; Traub, W., Electron diffraction of mollusc shell organic matrices and their relationship to the mineral phase. International Journal of Biological Macromolecules 1983, 5 (6), 325-328; (b) Weiss, I. M.; Kaufmann, S.; Heiland, B.; Tanaka, M., Covalent modification of chitin with silk-derivatives acts as an amphiphilic self-organizing template in nacre biomineralisation. Journal of structural biology 2009, 167 (1), 68-75.

82. Towe, K. M., Echinoderm calcite: single crystal or polycrystalline aggregate.

Science 1967, 157 (3792), 1048-1050.

83. Seto, J.; Ma, Y.; Davis, S. A.; Meldrum, F.; Gourrier, A.; Kim, Y.-Y.; Schilde, U.; Sztucki, M.; Burghammer, M.; Maltsev, S., Structure-property relationships of a biological mesocrystal in the adult sea urchin spine. Proceedings of the National Academy of Sciences 2012, 109 (10), 3699-3704.

84. Dubois, P.; Ameye, L., Regeneration of spines and pedicellariae in echinoderms: a review. Microscopy research and technique 2001, 55 (6), 427-437.

85. Carey, F. A., Chapter 27. Amino Acids, Peptides and Proteins. In Organic chemistry, McGraw-Hill: Boston, 2006.

86. Njegic-Dzakula, B.; Brecevic, L.; Falini, G.; Kralj, D., Calcite Crystal Growth Kinetics in the Presence of Charged Synthetic Polypeptides. Crystal Growth & Design 2009, 9 (5), 2425-2434.

87. Rees, D. A., Shapely polysaccharides. The eighth Colworth Medal Lecture.

Biochem. J. 1972, 126 (257).

157

References

88. Foord, S. A.; Atkins, E. D. Y., New x-ray diffraction results from agarose: Extended single helix structures and implications for gelation mechanism. Biopolymers 1989, 28 (8), 1345-1365.

89. Gavira, J. A. Cristalización de proteínas en geles por métodos contradifusivos. 2000.

90. Sancho-Tomás, M.; Fermani, S.; Durán-Olivencia, M. A.; Otálora, F.; Gómez-Morales, J.; Falini, G.; García-Ruiz, J. M., Influence of Charged Polypeptides on Nucleation and Growth of CaCO3 Evaluated by Counterdiffusion Experiments. Crystal Growth & Design 2013, 13 (9), 3884-3891.

158

91. Rodriguez-Navarro, A., XRD2DScan: new software for polycrystalline materials characterization using two-dimensional X-ray diffraction. Journal of Applied Crystallography 2006, 39 (6), 905-909.

92. Wartchow, R., Datensammlung nach der “Learnt profile”-Methode (LP) für Calcit und Vergleich mit der “Background peak background”-Methode (BPB).

Z. Kristallogr 1989, 186, 300-302.

93. Dickens, B.; Bowen, J., Refinement of the crystal structure of the aragonite phase of CaCO3. Journal of Research of the National Institute of Stand. ards. Physics and Chemistry A 1971, 75, 27-32.

94. Kamhi, S. R., On the structure of vaterite CaCO3. Acta Crystallographica 1963, 16 (8), 770-772.

95. Señorale-Pose, M.; Chalar, C.; Dauphin, Y.; Massard, P.; Pradel, P.; Marín, M., Monohydrocalcite in calcareous corpuscles of Mesocestoides corti. Experimental Parasitology 2008, 118 (1), 54-58.

96. Addadi, L.; Raz, S.; Weiner, S., AdV. Mater. 2003, 15, 959.

97. Song, R.-Q.; Colfen, H., Additive controlled crystallization. CrystEngComm 2011, 13 (5), 1249-1276.

98. Mann, S.; Heywood, B. R.; Rajam, S.; Birchall, J. D., Controlled crystallization of CaCO 3 under stearic acid monolayers. Nature 1988, 334 (6184), 692-695.

99. Aizenberg, J.; Black, A. J.; Whitesides, G. M., Control of crystal nucleation by patterned self-assembled monolayers. Nature 1999, 398 (6727), 495-498.

100. Addadi, L.; Weiner, S., Interactions between acidic proteins and crystals: stereochemical requirements in biomineralization. Proceedings of the National Academy of Sciences 1985, 82 (12), 4110-4114.

101. (a) Gower, L. B.; Odom, D. J., Deposition of calcium carbonate films by a polymer-induced liquid-precursor (PILP) process. Journal of Crystal Growth 2000, 210 (4), 719-734; (b) Falini, G.; Fermani, S.; Gazzano, M.; Ripamonti, A., Biomimetic crystallization of calcium carbonate polymorphs by means of collagenous matrices. Chemistry-a European Journal 1997, 3 (11), 1807-1814.

102. Addadi, L.; Moradian, J.; Shay, E.; Maroudas, N.; Weiner, S., A chemical model for the cooperation of sulfates and carboxylates in calcite crystal nucleation: relevance to biomineralization. Proceedings of the national academy of sciences 1987, 84 (9), 2732-2736.

103. Jones, G. C.; Jackson, B., Infrared transmission spectra of carbonate minerals.

Chapman & Hall: 1993.

104. (a) Goffredo, S.; Vergni, P.; Reggi, M.; Caroselli, E.; Sparla, F.; Levy, O.; Dubinsky, Z.; Falini, G., The Skeletal Organic Matrix from Mediterranean Coral Balanophyllia europaea Influences Calcium Carbonate Precipitation. PLoS ONE 2011, 6 (7), e22338; (b) De Yoreo, J. J.; Dove, P. M., Shaping Crystals with Biomolecules.

Science 2004, 306 (5700), 1301-1302.

105. (a) Raz, S.; Weiner, S.; Addadi, L., Formation of highâ&#x20AC;?magnesian calcites via an amorphous precursor phase: possible biological implications. Advanced Materials 2000, 12 (1), 38-42; (b) Loste, E.; Wilson, R. M.; Seshadri, R.; Meldrum, F. C., The role of magnesium in stabilising amorphous calcium carbonate and controlling calcite morphologies. Journal of Crystal Growth 2003, 254 (1), 206-218.

106. Aizenberg, J.; Lambert, G.; Weiner, S.; Addadi, L., Factors involved in the formation of amorphous and crystalline calcium carbonate: a study of an ascidian skeleton. Journal of the American Chemical Society 2002, 124 (1), 32-39.

107. (a) Raz, S.; Hamilton, P. C.; Wilt, F. H.; Weiner, S.; Addadi, L., The transient phase of amorphous calcium carbonate in sea urchin larval spicules: the involve-

159

References

ment of proteins and magnesium ions in its formation and stabilization. Advanced Functional Materials 2003, 13 (6), 480-486; (b) Weber, J. N., The incorporation of magnesium into the skeletal calcites of echinoderms. American Journal of Science 1969, 267 (5), 537-566; (c) Weiner, S.; Levi-Kalisman, Y.; Raz, S.; Addadi, L., Biologically formed amorphous calcium carbonate. Connective Tissue Research 2003, 44 (1), 214-218.

108. Wada, N.; Yamashita, K.; Umegaki, T., Effects of Carboxylic Acids on Calcite Formation in the Presence of Mg2+ Ions. Journal of Colloid and Interface Science 1999, 212 (2), 357-364.

160

109. (a) Segev, E.; Erez, J., Effect of Mg/Ca ratio in seawater on shell composition in shallow benthic foraminifera. Geochemistry, Geophysics, Geosystems 2006, 7 (2), 1-8; (b) Wang, D.; Wallace, A. F.; De Yoreo, J. J.; Dove, P. M., Carboxylated molecules regulate magnesium content of amorphous calcium carbonates during calcification. Proc Natl Acad Sci U S A 2009, 106 (51), 21511-6.

110. (a) Kitano, Y., The Behavior of Various Inorganic Ions in the Separation of Calcium Carbonate from a Bicarbonate Solution. Bulletin of the Chemical Society of Japan 1962, 35 (12), 1973-1980; (b) Given, R. K., Kinetic control of morphology, composition, and mineralogy of abiotic sedimentary carbonates. Journal of sedimentary petrology 1985, 55 (1), 109-119; (c) Falini, G.; Fermani, S.; Gazzano, M.; Ripamonti, A., Structure and morphology of synthetic magnesium calcite. Journal of Materials Chemistry 1998, 8 (4), 1061-1065.

111. Allemand, D.; Ferrier-Pagès, C.; Furla, P.; Houlbrèque, F.; Puverel, S.; Reynaud, S.; Tambutté, É.; Tambutté, S.; Zoccola, D., Biomineralisation in reef-building corals: from molecular mechanisms to environmental control. Comptes rendus de l’Académie des sciences 2004, 3 (6–7), 453-467.

112. Cuif, J. P.; Dauphin, Y.; Freiwald, A.; Gautret, P.; Zibrowius, H., Biochemical markers of zooxanthellae symbiosis in soluble matrices of skeleton of 24 Scleractinia species. Comparative Biochemistry and Physiology, Part A: Molecular & Integrative Physiology 1999, 123 (3), 269-278.

113. Gagnon, A. C.; Adkins, J. F.; Erez, J., Seawater transport during coral biomineralization. Earth and Planetary Science Letters 2012, 329–330 (0), 150-161.

114. Venn, A. A.; Tambutté, E.; Holcomb, M.; Laurent, J.; Allemand, D.; Tambutté,

S., Impact of seawater acidification on pH at the tissue–skeleton interface and calcification in reef corals. Proceedings of the National Academy of Sciences 2013, 110 (5), 1634-1639.

115. Falini, G.; Reggi, M.; Fermani, S.; Sparla, F.; Goffredo, S.; Dubinsky, Z.; Levi, O.; Dauphin, Y.; Cuif, J. P., Control of aragonite deposition in colonial corals by intra-skeletal macromolecules. Journal of Structural Biology 2013, 183 (2), 226-238.

116. Mass, T.; Drake, J. L.; Haramaty, L.; Kim, J. D.; Zelzion, E.; Bhattacharya, D.; Falkowski, P. G., Cloning and Characterization of Four Novel Coral Acid-Rich Proteins that Precipitate Carbonates In Vitro. Current Biology 2013, 23 (12), 1126-1131.

117. http://www.marinesciencegroup.org/ricercheindex.htm, Marine Science Group 118. Adler, H. H.; Kerr, P. F., Infrared study of aragonite and calcite. The American mineralogist 1962, 47, 700-717.

119. Mekid, S.; Vaja, D., Propagation of uncertainty: Expressions of second and third order uncertainty with third and fourth moments. Measurement 2008, 41 (6), 600-609.

120. Parker, F. S., Applications of infrared, Raman, and resonance Raman spectroscopy in biochemistry. Springer: New York, 1983.

121. (a) Li, H.; Xin, H. L.; Muller, D. A.; Estroff, L. A., Visualizing the 3D Internal Structure of Calcite Single Crystals Grown in Agarose Hydrogels. Science 2009, 326 (5957), 1244-1247; (b) Weiner, S.; Hood, L., Soluble protein of the organic matrix of mollusk shells: a potential template for shell formation. Science 1975, 190 (4218), 987-989; (c) Weiner, S., Organization of Organic Matrix Components in Mineralized Tissues. American Zoologist 1984, 24 (4), 945-951; (d) Albeck, S.; Weiner, S.; Addadi, L., Polysaccharides of Intracrystalline Glycoproteins Modulate Calcite Crystal Growth In Vitro. Chemistry – A European Journal 1996, 2 (3), 278-284.

122. Adamiano, A.; Bonacchi, S.; Calonghi, N.; Fabbri, D.; Falini, G.; Fermani, S.; Genovese, D.; Kralj, D.; Montalti, M.; Njegić Džakula, B.; Prodi, L.; Sartor, G., Structural Changes in a Protein Fragment from Abalone Shell during the Precipitation of Calcium Carbonate. Chemistry – A European Journal 2012, 18 (45), 14367-14374.

123. Wojtas, M.; Dobryszycki, P.; Ożyhar, A., Intrinsically disordered proteins in

161

References

biomineralization. In Advanced topics in biomineralization, Seto, J., Ed. InTech: 2012; pp 1-32 (and references therein).

124. Malferrari, D.; Fermani, S.; Galletti, P.; Goisis, M.; Tagliavini, E.; Falini, G., Shaping Calcite Crystals by Means of Comb Polyelectrolytes Having Neutral Hydrophilic Teeth. Langmuir 2013, 29 (6), 1938-1947.

125. Ihli, J.; Kim, Y.-Y.; Noel, E. H.; Meldrum, F. C., The Effect of Additives on Amorphous Calcium Carbonate (ACC): Janus Behavior in Solution and the Solid State. Advanced Functional Materials 2013, 23 (12), 1575-1585.

162

126. Albeck, S.; Aizenberg, J.; Addadi, L.; Weiner, S., Interactions of various skeletal intracrystalline components with calcite crystals. Journal of the American Chemical Society 1993, 115 (25), 11691-11697.

127. Vandermeulen, J. H.; Watabe, N., Studies on reef corals. I. Skeleton formation by newly settled planula larva of Pocillopora damicornis. Mar. Biol. 1973, 23 (1), 47-57.

128. Motai, S.; Nagai, T.; Sowa, K.; Watanabe, T.; Sakamoto, N.; Yurimoto, H.; Kawano, J., Needle-like grains across growth lines in the coral skeleton of Porites lobata. Journal of Structural Biology 2012, 180 (3), 389-393.

129. (a) Cuif, J. P.; Dauphin, Y.; Farre, B.; Nehrke, G.; Nouet, J.; SalomĂŠ, M., Distribution of sulphated polysaccharides within calcareous biominerals suggests a widely shared two-step crystallization process for the microstructural growth units.

Mineralogical Magazine 2008, 72 (1), 233-237; (b) Cartwright, J. H. E.; Checa, A. G.; Gale, J. D.; Gebauer, D.; Sainz-DĂ­az, C. I., Calcium Carbonate Polyamorphism and Its Role in Biomineralization: How Many Amorphous Calcium Carbonates Are There? Angewandte Chemie International Edition 2012, 51 (48), 11960-11970.

130. Gal, A.; Habraken, W.; Gur, D.; Fratzl, P.; Weiner, S.; Addadi, L., Calcite Crystal Growth by a Solid-State Transformation of Stabilized Amorphous Calcium Carbonate Nanospheres in a Hydrogel. Angewandte Chemie 2013, 125 (18), 4967-4970.

131. Pai, R. K.; Pillai, S., Water-Soluble Terpolymer Directs the Hollow Triangular Cones of Packed Calcite Needles. Crystal Growth & Design 2006, 7 (2), 215-217.

132. Pai, R. K.; Pillai, S., Divalent Cation-Induced Variations in Polyelectrolyte

Conformation and Controlling Calcite Morphologies: Direct Observation of the Phase Transition by Atomic Force Microscopy. Journal of the American Chemical Society 2008, 130 (39), 13074-13078.

133. Stephenson, A. E.; DeYoreo, J. J.; Wu, L.; Wu, K. J.; Hoyer, J.; Dove, P. M., Peptides Enhance Magnesium Signature in Calcite: Insights into Origins of Vital Effects. Science 2008, 322 (5902), 724-727.

134. (a) Arias, J. L.; Fernández, M. S., Polysaccharides and Proteoglycans in Calcium Carbonate-based Biomineralization. Chemical Reviews 2008, 108 (11), 44754482; (b) Tester, C. C.; Brock, R. E.; Wu, C. H.; Krejci, M. R.; Weigand, S.; Joester, D., In vitro synthesis and stabilization of amorphous calcium carbonate (ACC) nanoparticles within liposomes. CrystEngComm 2011, 13 (12), 3975-3978.

135. (a) Giuffre, A. J.; Hamm, L. M.; Han, N.; De Yoreo, J. J.; Dove, P. M., Polysaccharide chemistry regulates kinetics of calcite nucleation through competition of interfacial energies. Proceedings of the National Academy of Sciences 2013, 110 (23), 9261-9266; (b) Tao, J.; Zhou, D.; Zhang, Z.; Xu, X.; Tang, R., Magnesium-aspartate-based crystallization switch inspired from shell molt of crustacean. Proceedings of the National Academy of Sciences 2009, 106 (52), 22096-22101.

136. Jokiel, P. L., Ocean Acidification and Control of Reef Coral Calcification by Boundary Layer Limitation of Proton Flux. Bulletin of Marine Science 2011, 87 (3), 639-657.

137. (a) Khalifa, G. M.; Weiner, S.; Addadi, L., Mineral and Matrix Components of the Operculum and Shell of the Barnacle Balanus amphitrite: Calcite Crystal Growth in a Hydrogel. Crystal Growth & Design 2011, 11 (11), 5122-5130; (b) Falini, G.; Fermani, S., The strategic role of adsorption phenomena in biomineralization. Crystal Research and Technology 2013, 48 (10), 864-876.

138. Rao, A.; Seto, J.; Berg, J. K.; Kreft, S. G.; Scheffner, M.; Cölfen, H., Roles of larval sea urchin spicule SM50 domains in organic matrix self-assembly and calcium carbonate mineralization. Journal of Structural Biology 2013, 183 (2), 205-215.

139. Berman, A.; Addadi, L.; Weiner, S., Nature 1988, 331, 546.

140. Hernández-Hernández, A.; Vidal, M. L.; Gómez-Morales, J.; Rodríguez-Navarro, A. B.; Labas, V.; Gautron, J.; Nys, Y.; García Ruiz, J. M., Influence of eggshell

163

References

matrix proteins on the precipitation of calcium carbonate (CaCO3). Journal of Crystal Growth 2008, 310 (7–9), 1754-1759.

141. Gebauer, D.; Cölfen, H., Prenucleation clusters and non-classical nucleation.

Nano Today 2011, 6 (6), 564-584.

142. (a) Cuif, J.-P., Microstructure versus morphology in the skeleton of Triassic scleractinian corals. Acta Palaeontologica Polonica 1980, 25 (3-4), 361-374; (b) Stanley Jr, G. D., The evolution of modern corals and their early history.

Earth-Science Reviews 2003, 60 (3–4), 195-225.

164

143. Gebauer, D.; Gunawidjaja, P. N.; Ko, J. Y. P.; Bacsik, Z.; Aziz, B.; Liu, L.; Hu, Y.; Bergström, L.; Tai, C.-W.; Sham, T.-K.; Edén, M.; Hedin, N., Proto-Calcite and Proto-Vaterite in Amorphous Calcium Carbonates. Angewandte Chemie International Edition 2010, 49 (47), 8889-8891.

PhD THESIS C a l c i u m C a r b o n at e B i o - p r e c i p i tat i o n i n Gelling Environments via Counter-diffusion

María Sancho Tomás