Ca2+ Activity Ratio on the Formation of Crystalline

Jan 24, 2017 - (8, 9) Since then, the presence of an ACC precursor has been observed in the skeletons of sea urchins,(10, 11) freshwater snails,(12, 1...
2 downloads 12 Views 2MB Size
Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Article

Control of Mg2+/Ca2+ activity ratio on the formation of crystalline carbonate minerals via an amorphous precursor Bettina Purgstaller, Florian Konrad, Martin Dietzel, Adrian Immenhauser, and Vasileios Mavromatis Cryst. Growth Des., Just Accepted Manuscript • DOI: 10.1021/acs.cgd.6b01416 • Publication Date (Web): 24 Jan 2017 Downloaded from http://pubs.acs.org on January 25, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Crystal Growth & Design is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

1

Control of Mg2+/Ca2+ activity ratio on the formation of crystalline carbonate

2

minerals via an amorphous precursor

3

Bettina Purgstaller*1, Florian Konrad1, Martin Dietzel1, Adrian Immenhauser2, Vasileios Mavromatis1,3

4

1

Institute of Applied Geosciences, Graz University of Technology, Rechbauerstrasse 12, 8010 Graz, Austria

5

2

Institute of Geology, Mineralogy and Geophysics, Bochum, Universitätsstraße 150, 44801 Bochum, Germany

6

3

Géosciences Environnement Toulouse, CNRS, UMR5563, Avenue Edouard Belin 14, 31400 Toulouse, France

7 8

In the present work the transformation of Mg-rich amorphous calcium carbonate (Mg-ACC) into

9

crystalline carbonate phases has been studied at distinct initial Mg/Ca ratios (1/3 to 1/8) under

10

controlled pH conditions (7.8 to 8.8). The experimental results revealed two pathways of Mg-

11

ACC transformation strictly depending on the prevailing ratio of the Mg2+ to Ca2+ activity,

12

aMg2+/aCa2+, of the reactive solution after Mg-ACC was synthesized: Mg-ACC transformed to (i)

13

Mg-calcite at 5 ≤ aMg2+/aCa2+ ≤ 8 and to (ii) monohydrocalcite at 8 ≤ aMg2+/aCa2+ ≤ 12. Once the

14

reactive solution reaches saturation with ACC it exhibits high supersaturation with respect to

15

crystalline carbonate minerals, which promotes their formation and parallel initiates the

16

dissolution of Mg-ACC. At a high aMg2+/aCa2+ the time period of metastability of Mg-ACC in the

17

reactive solution (tACC) is prolonged due to the inhibition of calcite formation and the slow

18

precipitation kinetics of monohydrocalcite. These observations suggest that the formation

19

behavior of the crystalline phases is the rate-limiting process in the transformation of Mg-ACC.

20 21 22 23 24 25 26 27 28 29

Bettina Purgstaller Institute of Applied Geosciences Graz University of Technology Rechbauerstrasse 12 8010 Graz, Austria Telephone: +43 (0) 316/ 873-6362 e-mail: [email protected]

1 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

30

Control of Mg2+/Ca2+ activity ratios on the formation

31

of crystalline carbonate minerals via an amorphous

32

precursor

33

Bettina Purgstaller*1, Florian Konrad1, Martin Dietzel1, Adrian Immenhauser2, Vasileios

34

Mavromatis1,3

35

1

36

Graz, Austria

37

2

38

150, 44801 Bochum, Germany

39

3

40

Toulouse, France

41 42

*Corresponding author: [email protected]

Institute of Applied Geosciences, Graz University of Technology, Rechbauerstrasse 12, 8010

Institute of Geology, Mineralogy and Geophysics, Ruhr-Universität Bochum, Universitätsstraße

Géosciences Environnement Toulouse, CNRS, UMR5563, avenue Edouard Belin 14, 31400

43

2 ACS Paragon Plus Environment

Page 2 of 32

Page 3 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

44

ABSTRACT

45

The formation of amorphous calcium carbonate (ACC) and its transformation to crystalline

46

phases plays a key role on the formation of carbonate minerals on Earth’s surface environments.

47

Nonetheless, the psychochemical parameters controlling the formation of crystalline CaCO3 via

48

an amorphous precursor are still under debate. In the present study the possibility of crystalline

49

CaCO3 formation via an ACC precursor in the pH range from 7.8 to 8.8 and at initial Mg/Ca

50

ratios from 1/3 to 1/8 is experimentally investigated. The obtained results document that the

51

transformation of Mg-rich ACC (Mg-ACC) to a crystalline phase is strictly controlled by the

52

prevailing ratio of the Mg2+ to Ca2+ activity, aMg2+/aCa2+, of the reactive solution after Mg-ACC

53

was synthesized: Mg-ACC transformed to (i) Mg-calcite at 5 ≤ aMg2+/aCa2+ ≤ 8 and to (ii)

54

monohydrocalcite at 8 ≤ aMg2+/aCa2+ ≤ 12. Our findings suggest that the formation of the

55

crystalline phase induces undersaturation of the reactive solution with respect to the ACC and

56

triggers its dissolution. Thus, the metastability of Mg-ACC in the reactive solution is not

57

determined by its Mg content, but is related to the formation kinetics of the less soluble

58

crystalline phase. The experimental results highlight the importance of prevailing

59

physicochemical conditions of the reactive solution on Mg-ACC transformation pathways.

60 61

INTRODUCTION

62

Calcium carbonate (CaCO3) occurs in three anhydrous crystalline forms, calcite, aragonite

63

and vaterite, and two hydrous crystalline forms, monohydrocalcite (CaCO3·H2O) and ikaite

64

(CaCO3·6H2O). Of these calcium carbonate minerals, the geological occurrence of the less stable

65

vaterite, monohydrocalcite, and ikaite is rare.1-4 In contrast, calcite and aragonite are deposited

66

readily in marine and terrestrial environments.5 Many biogenic and non-biogenic calcites contain 3 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

67

significant amounts of magnesium.6-7 Calcites with more than 4 mol% magnesium are generally

68

referred to as Mg-calcites.

69

As early as the 19th century, it has been observed that biominerals can be formed via an

70

amorphous calcium carbonate (ACC) phase.8-9 Since then, the presence of an ACC precursor has

71

been observed in the skeletons of sea urchins,10-11 freshwater snails,12-13 bivalves,14 brachiopods,15

72

ascidians,16 sponges,17 and crustaceans.18 More recently, amorphous carbonate phases were found

73

in intracellular inclusions in modern microbialites in Lake Alchichica19 and in the mucilaginous

74

matrix embedding planktonic cells in Lake Geneva.20 With respect to abiotic precipitation

75

environments, ACC formation and its transformation to calcite was for instance associated with

76

speleothem formation in caves21 and found to be applicable for medical product development.22-23

77

By mixing concentrated solutions of CaCl2 and Na2CO3, Brečević and Nielsen24 showed

78

that ACC is highly soluble (Ksp-ACC = 10-6.39 at 25°C); and that it rapidly transforms to crystalline

79

calcium carbonate polymorphs. In this context, it has been observed that organic molecules and

80

magnesium ions stabilize both biogenic and inorganic ACC.25-32 In the latter case, the elevated

81

metastability of Mg enriched amorphous calcium carbonate (Mg-ACC) was attributed to the

82

structural water bound to Mg ions that retards the process of dehydration and transformation.

83

Despite the fact that numerous studies have focused on ACC formation, the mechanisms

84

controlling its transformation into, for instance, (Mg-)calcite and/or aragonite is poorly

85

understood. The literature of the last decade has revealed that the crystallization pathways of the

86

amorphous phase depend on physicochemical parameters including temperature,33 pH,32,34

87

solution additives like magnesium ions26,27,32 and organic molecules.27

88

For example, when ACC was synthesized from a solution with a high initial pH (> 9.5), it

89

transformed to calcite via metastable vaterite.32,35 Conversely, by using Mg in the initial solution

90

(Mg/Ca = 1/9) ACC crystallized directly to (Mg-)calcite.32 On the other hand, by mixing 4 ACS Paragon Plus Environment

Page 4 of 32

Page 5 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

91

solutions with molar Mg/Ca ratios ≥ 1/5.6 and CO3/Ca ratios > 1, monohydrocalcite formation

92

was reported to proceed through an ACC precursor.36-38 In this context, it has been suggested that

93

ACC crystallizes through a nucleation controlled dissolution-reprecipitation reaction.35,38,39 In

94

contrast, many studies suggested that the short-range structure or the composition of ACC (e.g.

95

Mg content) controls its crystallization to a distinct crystalline carbonate phase.26, 27, 34

96

Most of the former studies dealing with (Mg-)ACC formation were conducted at solution

97

pH > 9.5,31,32,35,36,39 a value that is significantly higher than the pH of most natural aquatic

98

environments. Moreover, (Mg-)ACC was synthesized by batch methods with CaCl2/MgCl2 and

99

NaHCO3/Na2CO3 solutions, where the pH decreased from ~11 to ~8 during mixing and

100

subsequent (Mg-)ACC crystallization.32,35 In the context of this study, we conducted experiments

101

under controlled physicochemical conditions (constant T and pH) in order to shed light on the

102

role of circum-neutral pH regimes and effects of Mg/Ca ratios during the (trans-)formation of

103

Mg-ACC to crystalline phases. Specifically, the transformation pathways of Mg-ACC were

104

systematically examined from pH 7.8 to 8.8 at distinct Mg/Ca ratios (1/3 to 1/8) and discussed

105

together with our previous study at pH 8.3.40 The aim of the present paper is to provide a more

106

detailed understanding on the mechanisms controlling the formation of crystalline carbonate

107

phases via a Mg-ACC precursor.

108 109

METHODS

110

Experimental setup and analytical procedures

111

The synthesis of a solid Ca1-xMgxCO3 phase was carried out under controlled

112

physicochemical conditions (T and pH = constant) using the experimental setup described in

113

Purgstaller et al.40. Hereafter, only the most important features are described. The precipitation of

114

Ca1-xMgxCO3 was induced by the titration (702 SM Titrino titrator; Methrom) of 50 ml of a 0.6 5 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

115

M (Ca,Mg)Cl2 solution with distinct Mg/Ca ratios (1/3, 1/4, 1/6 and 1/8) at a rate of 2 ml/min into

116

50 ml of a 1 M NaHCO3/Na2CO3 solution at 25.00 ±0.03°C (Easy MaxTm 102; Mettler Toledo).

117

The pH of the reactive solution was kept constant at values of 7.8, 8.3, or 8.8 ±0.1 by the

118

automatic titration of a 1 M NaOH solution (Schott; TitroLine alpha plus), with titration steps of

119

0.1 ml. Experiments conducted with the Mg/Ca ratios 1/4, 1/6 and 1/8 are labeled as A1, A2 and

120

A3 at pH 7.8 and as C1, C2 and C3 at pH 8.8, respectively (Table 1). A replicate experiment of

121

C1 is labeled as C1_2. One experiment was carried out at pH 8.3 with a Mg/Ca ratio of 1/3 (exp.

122

MgCa3, Table 1). Prior to each experiment the pH of the carbonate buffer solution was adjusted

123

to the required experimental pH (±0.07) using a 1 M NaOH or a 1 M HCl solution.

124

During the experimental run, the temporal evolution of mineral precipitation was monitored

125

by in situ Raman spectroscopy (Raman RXN2TM analyzer, Kaiser Optical Systems). After 180

126

min of reaction time, the reactive solution was transferred into a 150 ml glass bottle and placed

127

air-tight on a compact shaker (Edmund Bühler GmbH; KS-15) operating at 150 rpm in a

128

temperature controlled room at 25 ±1°C. In each experiment, 5 ml of the reactive solution were

129

sampled after selected time steps for solid and solution analysis. Immediately after sampling the

130

solids were separated by a 0.2 µm cellulose acetate filter using a suction filtration unit and were

131

analyzed using Attenuated Total Reflectance - Fourier Transform Infrared Spectroscopy (ATR-

132

FTIR; Perkin Elmer Spektrum 100). The solids were subsequently dried at room temperature in a

133

desiccator using silica gel. The mineralogy of the dried precipitates was determined using X-ray

134

diffraction (PANalytical X´Pert PRO) and the mineral phases were quantified using Rietveld

135

refinement (software HighScore Plus, PANalytical, PDF-2 database). Selected precipitates were

136

gold coated and imaged using a scanning electron microscope (SEM, ZEISS DSM 982 Gemini).

137

The aqueous cation concentrations were determined using ion chromatography (Dionex IC S

138

3000, IonPac®, AS19 and CS16 column), with an analytical precision of ±3%. The total 6 ACS Paragon Plus Environment

Page 6 of 32

Page 7 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

139

alkalinity of the solutions was measured by a Schott TitroLine alpha plus titrator using a 0.02 M

140

HCl solution with a precision better than ±2% and a detection limit of 5 x 10-6 M.

141

The Mg content of the solids (in mol%) sampled at a given reaction time was calculated

142

according to the equation:

143

[Mg]solid = [Mg]

144

where [Mg]add and [Ca]add are the concentrations of added Mg and Ca by titration into the reactive

145

solution given in mol L-1 considering at any time volume change in the reaction solutions. [Mg]aq

146

and [Ca]aq are the measured molar concentrations of Mg and Ca in the reactive solutions. The

147

calculated [Mg]solid values obtained from Eq. 1 are in excellent agreement (R² = 0.96) with the

148

Mg contents of the bulk digested solids analyzed by inductively coupled plasma optical emission

149

spectrometry (Perkin Elmer Optima 8300 DV) (Figure S1). The analytical error was calculated to

150

be < ±3% for Ca and Mg.

[Mg]add - [Mg]aq

add -[Mg]aq +

[Ca]add - [Ca]aq

·100

(1)

151 152

Hydrochemical modeling

153

The aqueous speciation and ion activities of the reactive solutions were calculated using

154

the PHREEQC software (minteq.v4 database). The saturation Index (SI) of the reactive solution

155

with respect to ACC was calculated using the equation (aCa2+ ) . (a

CO3 2-

)

156

SIACC = log

157

where a denotes the activity of the ith species in solution and the solubility product (Ksp) is 10-6.39

158

at 25°C. 24

K sp



(2)

159

7 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

160

RESULTS

161

Solid phase characterization

162

The in situ Raman spectra of experiments A1, A2, and A3 carried out at pH 7.8 showed

163

that Mg-calcite is present right after the onset of the experiments within 1 min, as it was indicated

164

by the increase of intensity of the carbonate stretching band (v1) at ~1087 cm-1 (exp. A1, Figure

165

1A) and of the libration mode of Mg-calcite at ~282 cm-1 (see values in Table 2).41 The intensity

166

of the Raman v1 band for Mg-calcite was constant over time as soon as the titration of the

167

(Ca,Mg)Cl2 solution was stopped (exp. A1, Figure 1A). In experiments A2 and A3, aragonite was

168

identified in addition to Mg-calcite after 180 min of reaction time detectable through the presence

169

of the libration mode at ~205 cm-1 in the Raman spectra (Table 2).42 The results obtained by

170

Raman spectroscopy are in good agreement with the ATR-FTIR analysis, showing the

171

characteristic vibration band of Mg-calcite at ~711 cm-1 (v4)43 in all precipitates and the presence

172

of aragonite by the ATR-FTIR bands at ~699 cm-1 (v4) and ~856 cm-1 (v2)44 after 25 min of

173

reaction time in experiments A2 and A3 (Table 2). Moreover, the ATR-FTIR spectra of the

174

precipitates that were sampled between 13 and 25 min and between 13 and 180 min in

175

experiments A2 and A3, respectively, showed vibration bands at ~743 cm-1 (v4) corresponding to

176

vaterite (Table 2). 44 The discrepancy of the analyzed mineralogy at certain reaction times within

177

a single experiment (Table 2) is attributed to the higher sensitivity of the ATR-FTIR analysis of

178

the collected solids compared to that of the in situ Raman spectroscopy of the reactive

179

solution/suspension.

180

Quantitative XRD results (Table 3) showed that the dried precipitates of experiment A1

181

mainly consist of Mg-calcite (96 ±2 wt. %) and traces of aragonite (4 ±2 wt. %). In experiments

8 ACS Paragon Plus Environment

Page 8 of 32

Page 9 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

182

A2 and A3 vaterite (up to 21 wt. %) and aragonite (13-33 wt. %) can be found additionally to

183

Mg-calcite, which is abundant in the bulk precipitate (61-75 wt. %).

184

The SEM images of the precipitates of experiments A1, A2 and A3, sampled after 1 day

185

of reaction time, showed aggregates (~4 µm) of rhombohedral Mg-calcite crystals (Figure 2A-C),

186

which are similar in texture to those obtained in experiment MgCa8 from Purgstaller et al.40

187

(Figure 2F). The Mg-calcite crystals are smaller in size (< 0.1 µm) in experiment A1 compared to

188

those from experiments A2 (~0.3 µm) and A3 (~1 µm). Moreover, in experiment A1 Mg-calcite

189

crystal habits are more elongated (Figure 2A) compared to those from experiments A2 (Figure

190

2B) and A3 (Figure 2C). Finally, in the A2 and A3 experiments, pseudo-hexagonal crystals of

191

aragonite (~1 µm) were observed (exp. A3 in Figure S2A).

192

In experiments C1, C2, and C3 carried out at pH 8.8 the collected in situ Raman spectra

193

revealed the presence of only Mg-ACC within 58, 40, and 36 min of reaction time, respectively.

194

The Raman spectra of Mg-ACC are characterized by a single broad v1 band at 1082 ±1 and by the

195

absence of lattice modes

196

ATR-FTIR analyses of sampled precipitates (exp. C1 in Figure 1B, Table 2). The lack of the v4

197

vibration band and the presence of vibration bands at 860-861 cm-1 and 1073 ±1 cm-1 in the ATR-

198

FTIR spectra are characteristic of Mg-ACC.26 In the experiment conducted with the highest

199

Mg/Ca ratio (exp. C1), a Raman band at ~1068 cm-1 corresponding to the v1 band of

200

monohydrocalcite46 evolved after 58 min of reaction time, while the characteristic band of Mg-

201

ACC at ~1083 cm-1 decreased (Figure 1B). ATR-FTIR spectra of the precipitates collected after

202

180 min of reaction time showed vibration bands at ~697/~726 cm-1, ~755 cm-1, ~871 cm-1,

203

~1068 cm-1 and ~1395/~1474 cm-1 (Figure 1B). These ATR-FTIR bands are in good agreement

204

with wavenumber values for synthetic monohydrocalcite from Neumann and Epple47. Similar in

205

situ Raman and ATR-FTIR results were obtained in the replicate experiment C1_2 (Table S1).

45

(exp. C1 in Figure 1B; Table 2). These results were confirmed by

9 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 32

206

After 1 day of reaction time, the in situ Raman spectra indicated the presence of a second v1 band

207

at ~1100 cm-1 (Table 2), corresponding to nesquehonite46. In contrast to the in situ Raman

208

spectroscopic observations, the presence of nesquehonite was not detectable by ATR-FTIR

209

analysis due to similarities of the vibration bands of monohydrocalcite and nesquehonite at higher

210

wavenumbers in the ATR-FTIR spectra. The XRD results showed that the dried precipitate

211

consists of Mg-calcite at 25 min of reaction time, whereas monohydrocalcite was found in the

212

precipitate sampled at 180 min of reaction time (Table 3). After 1 day of reaction time

213

nesquehonite (14 wt. %) was detected beside monohydrocalcite (86 wt. %, Table 3).

214

In contrast, in experiments conducted with the Mg/Ca ratios 1/6 and 1/8 (exp. C2 and C3),

215

in situ Raman spectroscopy revealed the transformation of Mg-ACC to Mg-calcite (exp. C3 in

216

Figure S3). The formation of Mg-calcite was indicated by the rise of the translation mode of Mg-

217

calcite at ~283 cm-1 (Table 2) and the rise of the intensity of the v1 Raman vibration band at 1088

218

cm-1.40 The results obtained by in situ Raman spectroscopy are in good agreement with those

219

obtained by ATR-FTIR analysis (Table 2). As reported in Table 3 the XRD patterns of all dried

220

precipitates in experiments C2 and C3 showed only Mg-calcite.

221

The monohydrocalcite obtained in experiment C1 forms non-aggregated nano-spheres

222

(exp. C1, Figure 2G), whereas the nesquehonite precipitates of experiment C1 are rod-shaped

223

with individual rods reaching lengths of up to 40 µm (Figure S2B). The Mg-calcites of

224

experiments C2 (Figure 2H) and C3 (Figure 2I) display aggregates composed of nanoparticles (~

225

80 nm) that agglomerated to larger clusters. These particles are significantly smaller in size than

226

the Mg-calcite crystals resulting from experiments conducted at pH 7.8 (Figure 2A-C). The Mg-

227

calcite aggregates however are similar in texture to those observed in experiments MgCa4

228

(Figure 2D) and MgCa6 (Figure 2E) conducted at pH 8.3 by Purgstaller et al.40 where Mg-calcite

229

was formed via Mg-ACC. 10 ACS Paragon Plus Environment

Page 11 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

230

The collected in situ Raman spectra of experiment MgCa3 conducted at pH 8.3 (Mg/Ca =

231

1/3) showed that Mg-ACC was present during the titration of the (Ca,Mg)Cl2 solution (Table 2).

232

After about 50 min of reaction time the presence of a second v1 band at ~1069 cm-1

233

corresponding to monohydrocalcite was observed in the in situ Raman spectra (Table 2). Similar

234

results were obtained by ATR-FTIR analysis, showing the characteristic vibration bands of

235

monohydrocalcite beside the vibration bands of ACC after 60 min of reaction time (Table 2).

236

After 180 min of reaction time only vibration bands of monohydrocalcite and a weak band at

237

1088 cm-1 were observed in ATR-FTIR spectra (Table 2). The XRD pattern showed that the dried

238

precipitates produced in experiment MgCa3 at 25 min of reaction time consist of Mg-calcite

239

(Table 3). In contrast, the precipitate collected at 180 min of reaction time consists of

240

monohydrocalcite (87 wt.%) and minor amounts of Mg-calcite (13 wt.%, Table 3). After 1 day of

241

reaction time nesquehonite (13 wt.%) was found in addition to monohydrocalcite (79 wt. %) and

242

Mg-calcite (8 wt. %, Table 3).

243 244

Chemical evolution of the reactive solutions

245

The initial carbonate solutions are characterized by a carbonate alkalinity of 0.99 ±0.02 M at

246

pH 7.8, 1.02 ±0.01 M at pH 8.3 and 1.15 ±0.02 M at pH 8.8. In all experiments, the carbonate

247

alkalinities decreased to a value of 0.11 ±0.03 M during the titration of the (Ca,Mg)Cl2 solutions

248

(Table S2-S4).

249

The titration of the (Ca,Mg)Cl2 solutions into the carbonate buffer solutions caused instant

250

nucleation/precipitation of Ca1-xMgxCO3 within 1 min of reaction time. The [Ca]aq of the reactive

251

solutions did not exceed 0.007 M during all experimental runs (Table S1-4). After 25 min of

252

reaction time about 98 % of the nCaadd was removed from the reactive solutions due to mineral

253

precipitation. The temporal evolution of the [Mg]aq of the reactive solutions of experiments 11 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 32

254

conducted at pH 8.3 and 8.8 is displayed in Figure 3A-B, while the obtained [Mg]aq values of

255

experiment MgCa3 are displayed in Figure S4A. The [Mg]aq of the reactive solutions increased

256

during the titration of the (Ca,Mg)Cl2 solutions. In experiments A1 to A3 the [Mg]aq remained

257

nearly constant between 25 min and 1 day of reaction time (Figure 3A), whereas in experiments

258

C2, and C3 a significant decrease of the [Mg]aq (up to 0.015 M) was observed for the same time

259

interval (Figure 3B). In experiments C1 and MgCa3 on the other hand, the [Mg]aq remained

260

constant at 0.026 and 0.036 M , respectively, between 25 and 60 min of reaction time, before it

261

increased to about 0.040 M in both experiments (180 min, Figure 3B and Figure S4A).

262

Subsequently, the [Mg]aq decreased again to 0.019 M and 0.035 M in experiments C1 and

263

MgCa3, respectively, between 180 min and 1 day of reaction time. A similar [Mg]aq trend as in

264

experiment C1 was observed in the replicate experiment C1_2 (Figure 4B).

265 266

Temporal evolution of Mg content in precipitates

267

During the titration of the (Ca,Mg)Cl2 solutions, the [Mg]solid increased in all experiments

268

(Figure 3C-D). The highest [Mg]solid content after 25 min of reaction time was 10.2 ±0.5 mol%

269

present in the solids of experiments A1 and C1. In experiments A2 and C2 the precipitates

270

contained 7.6 ±0.6 mol% Mg, whereas those in experiments A3 and C3 contained 6.7 ±1.2 mol%

271

Mg after 25 min of reaction time. In experiment MgCa3 the [Mg]solid obtained after 25 min of

272

reaction time is 11.4 mol%.

273

In experiments A1 to A3 (Figure 3C), the [Mg]solid remained nearly constant between 25 min

274

and 1 day of reaction time, while in experiments C2 and C3 the [Mg]solid increased significantly

275

between 25 and 60 min of reaction time (Figure 3D). In the case of experiments C1 and MgCa3,

276

the [Mg]solid remained constant between 25 and 60 min of reaction time and then decreased from

277

10 to 4 mol% in experiment C1 (180 min, Figure 3D) and from 12 to 9 mol% in experiment 12 ACS Paragon Plus Environment

Page 13 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

278

MgCa3 (180 min, Figure S4B). However, between 180 min and 1 day of reaction time, the

279

[Mg]solid increased significantly in both experiments (Figure 3D and Figure S4B), which is also

280

the case for the replicate experiment C1_2, where the evolution of the [Mg]solid is similar to

281

experiment C1.

282 283

DISCUSSION

284 285

Effect of pH on the formation of amorphous calcium carbonate

286

The presence of Mg-rich ACC was observed in all experiments conducted at pH 8.8 (e.g.

287

experiment C1 in Figure 1B, Table 2). In contrast, in experiments conducted at pH 7.8 the in situ

288

Raman spectra showed that Mg-calcite is present within 1 min after the onset of the experiments

289

and in absence of an amorphous precursor (e.g. experiment A1 in Figure 1A; Table 2). Note here

290

that the resolution of in situ Raman spectra at the initial stage of the experiments (< 5 min) is low

291

due to the low amount of solid suspended in the reactive solution. Moreover, the Raman v1 band

292

of ACC is significantly broadened compared to that of crystalline CaCO3 owing to its poorly

293

ordered structure.46 Thus, the presence of ACC within this time frame cannot be completely

294

excluded. In order to evaluate the metastability conditions of ACC in the present experiments, the

295

saturation index of the reactive solutions with respect to ACC (SIACC) was calculated according

296

to Eq. (2). Reactive solutions exhibit SIACC slightly above 0 within at least 25 min of reaction

297

time in the experiments conducted at pH 8.8 (Figure 4). In contrast, in experiments conducted at

298

pH 7.8 reactive solutions exhibit SIACC < 0 during the titration of the (Ca,Mg)Cl2 solution (Figure

299

4). The concentrations and activities of the two dissolved inorganic carbon species (e.g. HCO3-

300

and CO32-) depend on the pH of the reactive solution. For example, after 1 min of reaction time in

301

experiments conducted at pH 8.8, the ratio of CO32- to HCO3- activities in the reactive solutions, 13 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 32

302

aCO32-/aHCO3-, was ~3·10-2, whereas in experiments conducted at pH 7.8 the aCO32-/ aHCO3- was

303

significant lower at ~3·10-3 (Table S5). In the latter case, Mg-ACC was likely not formed due to

304

the limited concentration and activity of CO32- in the reactive solutions at lower pH (Table S5).

305

The direct formation of Mg-calcite in experiments conducted at pH 7.8 could be followed

306

by SEM images displaying Mg-calcite aggregates of aligned rhombohedral crystals, which are

307

significantly larger in size (Figure 2A-C) compared to the Mg-calcite nanoparticles formed in

308

experiments, where Mg-ACC transformation was detected (Figure 2H-I). These polycrystalline

309

Mg-calcite aggregates (Figure 2A-C) were likely formed via a spherulitic growth mechanism,

310

where the spherical particles grow from a central nucleation site or seed by incorporation of ions

311

from the solution48. This occurred in experiments A1 to A3 as a consequence of high superstation

312

levels of the reactive solution with respect to calcite (SIcalcite = 1.9 ±0.2 after 1 min of reaction

313

time). In experiment A2 and A3, XRD results indicated the presence of vaterite and aragonite

314

beside Mg-calcite (Table 3). Between 25 min and 1 day of reaction time the proportion of vaterite

315

decreased while the proportion of aragonite increased (Table 3). In most cases, metastable

316

vaterite transforms to calcite via a dissolution-reprecipitation mechanism.32,33,35 In the present

317

study, the ongoing formation of aragonite rather than of Mg-calcite is at first documented at 25°C

318

and is suggested to be caused by the high molar Mg/Ca ratio in the reactive solution after 25 min

319

of reaction time (> 46), inhibiting calcite formation.49-51

320 321

Effect of pH and aqueous Mg2+/Ca2+ ratio on the transformation pathways of ACC

322

The experimental results revealed two pathways of Mg-ACC transformation in the

323

reactive solutions (Figure 5A). Mg-ACC transformed into (i) Mg-calcite in experiments

324

conducted at pH 8.3 with the Mg/Ca ratios of 1/6 to 1/4 (exp. MgCa4, MgCa5 and MgCa6

325

conducted by Purgstaller et al.40) as well as in experiments conducted at pH 8.8 with the Mg/Ca 14 ACS Paragon Plus Environment

Page 15 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

326

ratios 1/6 and 1/8 (exp. C2 and C3), and (ii) monohydrocalcite in experiments conducted at pH

327

8.3 and 8.8 with the Mg/Ca ratios 1/3 (exp. MgCa3) and 1/4 (exp. C1), respectively. Note that in

328

experiment MgCa3 small amounts of Mg-calcite (13 wt. %) were detected beside

329

monohydrocalcite (Table 3).

330

It has been recently suggested that the crystalline phase forming from ACC is defined by

331

the Mg/Ca ratio of the initial solution: ACC with 10 mol% Mg precipitated from a solution with

332

an initial Mg/Ca ratio of 1/9 crystallized to Mg-calcite,32 whereas ACC with 30 mol% Mg

333

precipitated from a solution with an initial Mg/Ca ratio of 3/7 crystallized to

334

monohydrocalcite.38A different pattern is documented in experiments of the present study

335

conducted with the initial Mg/Ca ratio of 1/4 (exp. MgCa4 and C1). Although in both

336

experiments formed ACC has a similar Mg content (10.1 ±0.2 mol% Mg), the in situ Raman

337

spectroscopy revealed the formation of Mg-calcite in experiment MgCa4 and of

338

monohydrocalcite in experiment C1. The formation of monohydrocalcite versus Mg-calcite from

339

Mg-ACC could be attributed to the higher pH of the reactive solution in experiment C1 (pH 8.8)

340

compared to that in experiment MgCa4 (pH 8.3, Figure 5A). However, considering that

341

monohydrocalcite was formed at pH 8.8 only in the experiment conducted with the highest initial

342

Mg/Ca ratio (exp. C1, Figure 5A), a coupled effect of pH and aqueous Mg concentration on the

343

transformation pathways of Mg-ACC is indicated.

344

In the experiments conducted in this study at the end of the (Ca,Mg)Cl2 solution titration

345

and before the transformation of the amorphous to the crystalline phase took place, the ratio of

346

the Ca2+ to CO32- activities in the reactive solution, aCa2+/aCO32-, is lower (0.2 ±0.1) in experiments

347

conducted at pH 8.8 compared to that in experiments conducted at pH 8.3 (1.7 ±0.4, Figure 5B).

348

In the latter case, the aCO32- is lower, owing to the predominance of aqueous HCO3- versus CO32-

349

ions. In order to maintain a solution that is saturated with respect to ACC, the aCa2+ is thus higher 15 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 32

350

at pH 8.3 than at pH 8.8 (Figure 5B). This affects the prevailing ratio of Mg2+ to Ca2+ activities in

351

the reactive solution after Mg-ACC was synthesized (aMg2+/aCa2+, Figure 5C). For example,

352

although the initial Mg/Ca ratio was similar in both experiments, the aMg2+/aCa2+ is significantly

353

higher in experiment C1 compared to that in experiment MgCa4 after 25 min of reaction time

354

(Figure 5C), due to the lower aCa2+ at high versus low pH (Figure 5B). Overall, the results

355

obtained revealed that in experiments where Mg-ACC transformed to Mg-calcite (exp. MgCa4,

356

C2 and C3), the aMg2+/aCa2+ is significant lower compared to that in experiment C1, where

357

transformation of Mg-ACC to monohydrocalcite was observed (Figure 5C). In experiment

358

MgCa3, where Mg-ACC transformed to monohydrocalcite and Mg-calcite (13 wt.%), the

359

aMg2+/aCa2+ is equal to 8 and this value likely defines the boundary conditions for the

360

transformation of Mg-ACC to either Mg-calcite or monohydrocalcite (Figure 5C). These

361

observations suggest that the formation of the distinct crystalline phase is controlled by the

362

prevailing aMg2+/aCa2+ of the reactive solution after Mg-ACC was formed: Mg-ACC transformed

363

to (i) Mg-calcite at 5 ≤ aMg2+/aCa2+ ≤ 8 and to (ii) monohydrocalcite at 8 ≤ aMg2+/aCa2+ ≤ 12.

364 365

Mechanism of Mg-calcite and monohydrocalcite formation via metastable ACC

366

The experiments conducted in this study document that the reactive solutions are slightly

367

supersaturated with respect to ACC up to at least 25 min of reaction time (Figure 4). During this

368

timeframe, the reactive solutions are highly supersaturated with respect to calcite (SI ≈ 2.3) and

369

monohydrocalcite (SI ≈ 1.0), which is provoking formation of these phases in the reactive

370

solutions. In experiment C1, the formation of monohydrocalcite instead of thermodynamically

371

more stable calcite can be explained by the inhibition of the latter

372

prevailing aMg2+/aCa2+ (Figure 5C). The inhibiting effect of aqueous Mg2+ is already indicated in 16 ACS Paragon Plus Environment

49-51

due to the elevated

Page 17 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

373

experiment MgCa3, where only small amounts of Mg-calcite were formed besides

374

monohydrocalcite (Table 3). As soon as precipitation of a less soluble crystalline phase (Mg-

375

calcite and/or monohydrocalcite) takes place, the solution reaches undersaturation with respect to

376

ACC (Figure 4). Thus, it is likely that the transformation of the amorphous to the crystalline

377

phase occurred via a dissolution- and re-precipitation mechanism. This mechanism is further

378

supported by the observation that Mg ions are released into the reactive solution throughout Mg-

379

ACC transformation to monohydrocalcite (exp. C1 and C1_2, between 60 and 180 min in Figure

380

3B). In this context, the Mg content of the solid decreased significantly from 10 to 4 mol%,

381

which strongly suggests a partial dissolution/re-precipitation of Mg-rich ACC (Figure 3 D). The

382

prevailing aMg2+/aCa2+ ratio of the reactive solutions and thus the formation of Mg-calcite or

383

monohydrocalcite are suggested to determine the period of Mg-ACC metastability in the reactive

384

solutions (tACC). As it can be seen in Figure 5D, tACC is longer in experiments where

385

monohydrocalcite formation was observed (at higher aMg2+/aCa2+) compared to experiments with

386

Mg-calcite as the reaction product (at lower aMg2+/aCa2+). In that sense, it has been documented

387

that elevated Mg contents in ACC increase the metastability of the amorphous phase in a

388

solution.26,32,40 In contrast, results of the present study showed that the tACC is much longer in

389

experiment C1 (~58 min) compared to that in experiment MgCa4 (~34 min) (Figure 5D),

390

although the Mg content of ACC is about the same in both experiments (10.1 ±0.2 mol%). It is

391

concluded that the metastability of the amorphous phase in the reactive solution is not determined

392

by its Mg content, but is related to the prevailing aMg2+/aCa2+ ratio of the reactive solution and

393

formation behavior of the less soluble crystalline phases, initiating the dissolution of Mg-ACC.

394

Previous experimental studies on calcite formation documented that the nucleation/growth

395

of calcite is retarded at higher aqueous Mg/Ca molar ratios.53-54 Indeed, in the present study,

396

transformation of Mg-ACC to Mg-calcite was observed earlier in reactive solutions at low 17 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 32

397

compared to high aMg2+/aCa2+ ratios (Figure 5D). Moreover, the in situ Raman results obtained in

398

our previous study40 showed that the time of Mg-ACC transformation to Mg-calcite is delayed

399

when higher Mg/Ca ratios in the (Ca,Mg)Cl2 solution were used: at initial Mg/Ca ratios of 1/6,

400

1/5, and 1/4 the transformation of Mg-ACC to Mg-calcite was observed after about 15, 25 and 36

401

min of reaction time, respectively. We attributed the prolonged presence of Mg-ACC in the

402

reactive solution to its higher Mg content, but it could also be explained by retarded formation

403

kinetics of Mg-calcite at higher Mg concentrations in the reactive solution. In experiment C1, the

404

presence of Mg-ACC is prolonged due to (i) the inhibition of calcite formation by high

405

aMg2+/aCa2+ values (Figure 5D) and (ii) the slow formation kinetics of monohydrocalcite. Distinct

406

precipitation rates can be followed by in situ Raman spectroscopy, where the formation of

407

monohydrocalcite required ~ 25 min in experiment C1 (Figure 1B). In contrast, the formation of

408

Mg-calcite is significantly faster (~ 8 min, exp. C3 in Figure S3; cf. Purgstaller et al.40).

409

Evidence for the control of the fluid composition (i.e. aqueous Mg/Ca ratio) on the

410

formation of Mg-calcite or monohydrocalcite from Mg-ACC comes also from X-ray diffraction

411

results. Our findings showed that Mg-ACC separated rapidly from reactive solutions of

412

experiments C1 and MgCa3 by membrane filtration transformed to Mg-calcite under air exposure

413

(Table 3), despite the fact that the same Mg-ACC transformed to monohydrocalcite when it

414

remained in the reactive solutions (e.g. exp. C1 in Figure 1B). Under ambient atmospheric

415

conditions H2O molecules are adsorbed onto the ACC phase, forming a water layer, which

416

promotes its dissolution and re-precipitation52,55. The only source of cations in this water layer is

417

that released during the dissolution of Mg-ACC. In the present case, the [Mg]aq/[Ca]aq ratio of the

418

water layer is most likely close to that of the Mg-ACC; in experiments C1 and MgCa3 the Mg/Ca

419

of Mg-ACC is 1/9 and 1/8, respectively. These values are much lower compared to those in the

420

reactive solutions of experiments C1 ([Mg]aq/[Ca]aq = 9/1) and MgCa3 ([Mg]aq/[Ca]aq = 6/1), 18 ACS Paragon Plus Environment

Page 19 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

421

inducing nucleation and growth of Mg-calcite rather than monohydrocalcite. The different

422

reaction products observed during the transformation of the same Mg-ACC precursor either in

423

contact with a reactive fluid or in air indicate that the crystallization of Mg-ACC in our

424

experiments is significantly affected by the physicochemical conditions of the transformation

425

environment (prevailing [Mg]aq/[Ca]aq ratio of the fluid, ions released during Mg-ACC

426

dissolution etc.) and to a lesser extent by a potential proto-crystalline pre-structuring34 of the

427

primary Mg-ACC.

428 429

SUMMARY AND CONCLUSIONS

430

Despite the fact that Mg-rich amorphous calcium carbonate was precipitated at

431

different pH values (8.3 to 8.8) and initial Mg/Ca ratios (1/3 to 1/8) our findings revealed

432

two pathways of Mg-ACC transformation in aqueous solution, which strictly depend on the

433

prevailing aMg2+/aCa2+ of the reactive solution after Mg-ACC was synthesized: (i) at 5 ≤

434

aMg2+/aCa2+ ≤ 8 Mg-calcite is formed via intermediate Mg-ACC. (ii) Conversely, at 8 ≤

435

aMg2+/aCa2+ ≤ 12, Mg-ACC transformed to monohydrocalcite.

436

Independent of the Mg content in ACC, the in situ Raman monitoring indicated a

437

significantly extended metastability of Mg-ACC in the reactive solution in experiments

438

where Mg-ACC transformed to monohydrocalcite at higher aMg2+/aCa2+ (~ 58 min) compared

439

to experiments where Mg-ACC transformed to Mg-calcite at lower aMg2+/aCa2+ (~ 36 min).

440

Moreover, the formation of monohydrocalcite is shown to be a slow process (~ 25 min),

441

whereas the formation of Mg-calcite was comparable fast (~ 8 min). These observations

442

suggest that the formation time of the less soluble crystalline phase formed under the

443

prevailing physicochemical condition of the reactive solution is the rate-limiting process in

19 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

444

the dissolution/transformation of Mg-ACC. At higher aMg2+/aCa2+ the presence of Mg-ACC in

445

the reactive solution is prolonged due to the inhibition of calcite formation and the slow

446

precipitation kinetics of monohydrocalcite. Our experimental results documented a different

447

transformation product when Mg-ACC was separated from the Mg-rich reactive solution. In

448

this case, the formation of the crystalline phase is determined by the ions released through

449

Mg-ACC dissolution in present adsorbed water layer under atmospheric conditions. These

450

observations clearly highlight the importance of the physicochemical conditions of the

451

transformation environment on the formation of distinct crystalline carbonate phases via

452

metastable Mg-ACC.

Page 20 of 32

453 454

ACKNOWLEDGEMENTS

455

We acknowledge Andrea Wolf, Judith Jernej and Andre Baldermann for their support with the

456

analytical work. Chemical analyses were conducted at NAWI Graz Central Lab for Water,

457

Minerals and Rocks. We would like to thank Gerald Auer and Patrick Grunert for their support in

458

conducting SEM images. This study benefited from the enlightening comments of Jacques Schott

459

and Pascale Bénézeth. Finally, the authors thank Stephan Sacher (Research Center

460

Pharmaceutical Engineering) for providing access to in situ Raman spectroscopy. This work has

461

been financially supported by DFG collaborative research initiative 1644 CHARON (IM44/10-1)

462

and by NAWI Graz.

463 464

SUPPORTING INFORMATION DESCRIPTION

465

Table S1 to S4: Chemical composition of the reactive solutions and solids at certain reaction

466

times from all experiments; Figure S1: The calculated [Mg]solid values obtained from Eq. 1

467

plotted against the Mg contents of the bulk digested solids; Figure S2: additional SEM 20 ACS Paragon Plus Environment

Page 21 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

468

images from experiment A3 and C1; Figure S3: in situ Raman spectra of experiment C3;

469

Figure S4: Mg concentration of the reactive solution and Mg content of the precipitated solid

470

versus the experimental run time of experiment MgCa3.

471 472

REFERENCES

473

(1) Lowenstam, H. A. Science 1981, 211, 1126-1131.

474

(2) Ito, T. J. Mineral. Petrol. Econ. Geol. 1993, 88, 485–491.

475

(3) Buchardt, B.; Israelson, C.; Seaman, P; Stockmann, G. Journal of Sedimentary Research

476

2001, 71, 176-189.

477

(4) Lu, Z.; Rickaby, R. E. M.; Kennedy, H.; Kennedy, P.; Pancost, R. D.; Shaw, S.; Lennie, A.;

478

Wellner, J.; Anderson, J. B. Earth and Planetary Science Letters 2012, 325-326, 108-115.

479

(5) Morse, J. W; Mackenzie, F. T. Develop. Sed. 1990, 48, 707.

480

(6) Mackenzie, F. T.; Bischoff, W. B.; Bishop, F. C.; Loijens, M.; Schoonmaker, J.; Wollast, R;

481

In Carbonates, Mineralogy and Chemistry; Reeder R. J.; Mineral. Soc. America: Chelsea,

482

MI, 1983, Vol 11, pp. 97–144.

483

(7) Long, X.; Ma, Y; Qi, L. (2011) Cryst. Growth Des. 11, 2866-2873.

484

(8) Krukenberg, C. F. W. Vergleichende physiologische Studien 1982, 2, 92-93.

485

(9) Minchin, E. A. Quart. J. micr. Sci. 1898, 40, 469–588.

486

(10) Beniash, E.; Aizenberg, J.; Addadi, L.; Weiner, S. Proc. R. Soc. Lond. B 1997, 264, 461-465.

487

(11) Politi, Y.; Arad, T.; Klein, E.; Weiner, S.; Addadi, L. Science 2004, 306, 1161-1164

488

(12) Hasse, B.; Ehrenberg, H.; Marxen, J. C.; Becker, W.; Epple, M; Chem. Eur. J. 2000, 6, 3679.

489

(13) Khairoun, I.; Magne, D.; Gauthier, O.; Bouler, J. M.; Aguado, E.; Daculsi, G.; Weiss, P.; J.

490 491 492 493 494

Biomed. Mat. Res. 2002, 60, 633-642 (14) Jacob, D. E.; Soldati, A. L.; Wirth, R.; Huth, J.; Wehrmeister, U; Hofmeister, W. Geochimica et Cosmochimica acta 2008, 72, 5401-5415 (15) Griesshaber, E.; Kelm, K.; Sehrbrock, A.; Mader, W.; Mutterlose, J.; Brand, U.; Schmahl, W.W. European Journal of Mineralogy 2009, 21, 715-723.

495

(16) Aizenberg, J.; Lambert, G.; Weiner, S.; Addadi, L. J Am Chem Soc. 2002, 124, 32–9.

496

(17) Sethmann, I.; Hinrichs, R.; Wörheide, G.; Putnis, A. J. Inorg. Biochem. 2006, 100, 88–96.

497

(18) Tao, J.; Zhou, D.; Zhang, Z.; Xu, X.; Tang, R. PNAS 2009, 106, 22096–22101. 21 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

498 499

Page 22 of 32

(19) Couradeau, E.; Benzerara, K.; Gérard, E.; Moreira, D.; Bernard, S.; Brown, G.E. Jr.; LópezGarcía, P. Science 2012, 336, 459-462

500

(20) Jaquet, J.-M.; Nirel, P; Martignier, A. J. Limnol. (2013) 73, 592-605

501

(21) Demény, A.; Czuppon, G.; Kern, Z.; Leél-Ossy, S.; Németh, A.; Szabó, M.; Tóth, M.; Wu,

502

C-.C.; Shen, C.-C.; Molnár, M.; Németh, T.; Németh, P.; Óvári, M. Quaternary

503

International 2016, 415, 25-32

504 505 506 507

(22) Meiron, O. E.; Bar-David, E.; Aflalo, E. D.; Shechter, A.; Stepensky, D.; Berman, A.; Sagi, A. J. Bone Miner. Res. 2011, 26 (2), 364–372. (23) Zhao, Y.; Luo, Z.; Li, M.; Qu, Q.; Ma, X.; Yu, S.-H.; Zhao, Y. Angew. Chemie Int. Ed. 2015, 54 (3), 919–922.

508

(24) Brečević, L; Nielsen, A. E. Journal of Crystal Growth 1989, 98, 504-510.

509

(25) Raz, S.; Hamilton, P. C.; Weiner, S.; Addadi, L. Adv. Funct. Mater. 2003, 13, 480-486.

510

(26) Loste, E.; Wilson, R. M.; Seshadri, R; Meldrum, F. C. Journal of Crystal Growth 2003, 254,

511 512 513

206-218. (27) Lam, R. S. K.; Charnock, J. M.; Lennie, A.; Meldrum, F. C. CrystEngComm 2007, 9, 12261236

514

(28) Han, T. Y.-J.; Aizenberg, J. Chemistry of Materials 2008, 20(3), 1064-1068.

515

(29) Politi, Y.; Batchelor, D. R.; Zaslansky, P.; Chmelka, B. F.; Weaver, J. C.; Sagi, I.; Weiner,

516 517 518 519 520 521 522

S. Addadi, L. Chem. Mater. 2010, 22, 161–166. (30) Mavromatis, V.; Schmidt, M.; Botz, R.; Comas-Bru, L.; Oelkers, E. H. Chem. Geol. 2012, 310–311, 97–105. (31) Radha, A. V.; Fernandez-Martinez, A.; Hu, Y.; Jun, Y.; Waychunas, G. A.; Navrotsky A. Geochim. Cosmochim. Acta 2012, 90, 83-95. (32) Rodriguez-Blanco, J. D.; Shaw, S.; Bots, P.; Roncal-Herrero, T., Benning, L. G. Journal of Alloys and Compounds 2012, 536S, 477-479.

523

(33) Ogino, T.; Suzuki, T.; Sawada, K. Geochim. Cosmochim. Acta 1987, 51, 2757-2767.

524

(34) Gebauer, D.; Völkel, A.; Cölfen, H. Science 2008, 322, 1819-1822

525

(35) Bots, P.; Benning, L. G.; Rodriguze-Blanco, J. D.; Roncal-Herrero, T.; Shaw, S. Crys.

526

Growth Des. 2012, 12, 3806-3814

527

(36) Kimura, T.; Koga, N. Cryst. Growth. Des. 2011, 11, 3877-3884

528

(37) Nishiyama, R.; Munemoto, T.; Fukushi, K. Geochimica et Cosmochimica Acta 2013, 100,

529

217-231 22 ACS Paragon Plus Environment

Page 23 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

530 531 532 533 534 535 536 537 538 539

Crystal Growth & Design

(38) Rodriguez-Blanco, J. D.; Shaw, S.; Bots, P.; Roncal-Herrero, T.; Benning, L. G. Geochim. Cosmochim. Acta 2014 127, 204-220 (39) Rodriguez-Navarro, C; Kudlacz, K.; Cizer, Ö.; Ruiz-Agudo, E. CrystEngComm 2015, 17, 58-72 (40) Purgstaller, B.; Mavromatis, V.; Immenhauser, A.; Dietzel, M. Geochim. Cosmochim. Acta 2016, 174, 180-195 (41) Bischoff, W. D.; Sharma, S. K.; MacKenzie, F. T. American Mineralogist 1985, 70, 581589. (42) Edwards, H. G. M.; Villar, S. E. J.; Jehlicka, J.; Munshi, T. Spectrochimica Acta 2005, Part A 6, 2273-2280.

540

(43) Böttcher, M. E.; Gehlken, P.-L.; Steele, D. F. Solid State Ionics 1997, 101-103, 1379-1385.

541

(44) Kontoyannis, C. G.; Vagenas, V. Analyst 2000, 125, 251-255

542

(45) Wang, D.; Hamm, L. M.; Bodnar, R. J.; Dove, P. M.; J. of Raman Spectrosc. 2012, 43, 543-

543 544 545

548. (46) Coleyshaw, E. E.; Crump, G.; Griffith, W. P. Spectrochimica Acta 2003, Part A 59, 22312239

546

(47) Neumann, M.; Epple M. Eur. J. Inorg. Chem. 2007, 1953-1957

547

(48) Andreassen, J.-P.; Flaten, E. M.; Beck, R; Lewis, A. E. Chemical Engineering Research and

548

Design 2010, 88, 1163-1168

549

(49) Lippmann F. (1973) Sedimentary carbonate minerals. Springer, Berlin.

550

(50) Falini, G., Gazzano, M., Ripamonti, A., 1994. Crystallization of calcium carbonate in

551 552 553

presence of magnesium and polyelectrolytes. Journal of Crystal Growth 137, 577–584. (51) Fernández-Díaz, L., Putnis, A., Prieto, M., Putnis, C.V., 1996. Journal of Sedimentary Research 66, 482–491

554

(52) Xu, X.; Han, J. T.; Kim, D. H.; Cho, K. J. Phys. Chem. B 2006, 110, 2764-2770

555

(53) Fernández-Díaz, L.; Putnis, A.; Prieto, M.; Putnis, C.V. Journal of Sedimentary Research,

556

1996, 66, 482–491

557

(54) Niedermayr, A.; Köhler, S. J; Dietzel M. Chem. Geol. 2013, 340, 105–120

558 559

(55) Konrad, F.; Gallien, F.; Gerard D. E.; Dietzel M. Crystal Growth & Design, 2016, 63106317

23 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

560

Page 24 of 32

For Table of Contents Use Only

561

Control of Mg2+/Ca2+ activity ratio on the formation of crystalline carbonate

562

minerals via an amorphous precursor

563

Bettina Purgstaller1, Florian Konrad1, Martin Dietzel1, Adrian Immenhauser2, Vasileios Mavromatis1,3

564

1

Institute of Applied Geosciences, Graz University of Technology, Rechbauerstrasse 12, 8010 Graz, Austria

565

2

Institute of Geology, Mineralogy and Geophysics, Bochum, Universitätsstraße 150, 44801 Bochum, Germany

566

3

Géosciences Environnement Toulouse, CNRS, UMR5563, Avenue Edouard Belin 14, 31400 Toulouse, France

567 568

Mg-rich amorphous calcium carbonate that was sampled during the experimental

569

runs did not remain stable after sample filtration and altered to Mg-calcite under air

570

exposure. A different pattern was documented when Mg-ACC transformed in the

571

reactive solutions. Although precipitating Mg-ACC with a similar Mg content of 10

572

mol%, the experimental results revealed two pathways of Mg-ACC depending on

573

the prevailing ratio of Mg2+ to Ca2+ activities, aMg2+/aCa2+, in the reactive solution.

574 575 24 ACS Paragon Plus Environment

Page 25 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

576

Table 1

577

Initial compositions of the experimental solutions for ACC (trans)formation. Mg/Ca denotes

578

molar ratio referred to total concentration (Mg + Ca) of 0.6 M. Experiment A1 A2 A3 C1 C1_2 C2 C3 MgCa3

pH 7.8 7.8 7.8 8.8 8.8 8.8 8.8 8.3

Mg/Ca 1/4 1/6 1/8 1/4 1/4 1/6 1/8 1/3

579

25 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 32

580

Table 2

581

Frequencies (cm-1) of observed carbonate vibration bands from in situ Raman and ATR-FTIR

582

spectra and identified mineralogy of the precipitates at certain reaction times. Note that in

583

experiment MgCa3 in situ Raman spectra of reactive solution were not obtained after 60 min

584

(n.a.). Mg-Cc: Mg-calcite, Arg: aragonite, Vtr: vaterite, Mhc: monohydrocalcite, Nesq:

585

nesquehonite. Experiment

time

Raman v1

A1

A2

A3

C1

C2

C3

MgCa3

13 min 25 min 60 min 180 min 1d 13 min 25 min 60 min 180 min 1d 13 min 25 min 60 min 180 min 1d 13 min 25 min 60 min 180 min 1d 13 min 25 min 60 min 180 min 1d 13 min 25 min 60 min 180 min 1d 13 min 25 min 60 min 180 min 1d

1088 1088 1088 1088 1088 1087 1087 1087 1086 1086 1087 1087 1087 1086 1086 1083 1083 1083/1068 1068 1100/1068 1081 1081 1087 1088 1088 1082 1082 1088 1088 1088 1082 1083 1083/1069 n.a. n.a

Mineralogy libration mode 282 283 283 283 283 283 283 283 283/205 283/205 282 282 282 282/205 282/205 282 283 283 282 282 282 n.a. n.a.

Mg-Cc Mg-Cc Mg-Cc Mg-Cc Mg-Cc Mg-Cc Mg-Cc Mg-Cc Mg-Cc, Arg Mg-Cc, Arg Mg-Cc Mg-Cc Mg-Cc Mg-Cc, Arg Mg-Cc, Arg Mg-ACC Mg-ACC Mg-ACC, Mhc Mhc Mhc, Nesq Mg-ACC Mg-ACC Mg-Cc Mg-Cc Mg-Cc Mg-ACC Mg-ACC Mg-Cc Mg-Cc Mg-Cc Mg-ACC Mg-ACC Mg-ACC, Mhc n.a n.a

v1

ATRFTIR v2

v4

1082 1083 1083 1083 1083 1082 1082 1083 1082 1083 1082 1082 1082 1082 1082 1073 1074 1073 1068 1068 1072 1074 1085 1085 1083 1073 1073 1084 1084 1084 1073 1073 1073 1068/1088 1068/1088

870 870 871 870 870 871 870 870/856 870/855 871/855 871 870/856 870/856 870/855 870/854 860 860 870 /860 870 871 860 860 871 870 870 861 860 871 871 871 860 859 871/860 870 872

713 713 712 712 712 743/712 743/712/700 712/700 711/700 712/700 743/712 743/712/699 743/712/699 743/712/699 712/699 697 753/726/697 751/726/697 711 712 711 712 712 712 697 755/727/697 761/728/699

586 587 26 ACS Paragon Plus Environment

Mineralogy

Mg-Cc Mg-Cc Mg-Cc Mg-Cc Mg-Cc Mg-Cc, Vtr Mg-Cc, Arg, Vtr Mg-Cc, Arg Mg-Cc, Arg Mg-Cc, Arg Mg-Cc, Vtr Mg-Cc, Vtr, Arg Mg-Cc, Vtr, Arg Mg-Cc, Vtr, Arg Mg-Cc, Arg Mg-ACC Mg-ACC Mg-ACC, Mhc Mhc Mhc Mg-ACC Mg-ACC Mg-Cc Mg-Cc Mg-Cc Mg-ACC Mg-ACC Mg-Cc Mg-Cc Mg-Cc Mg-ACC Mg-ACC Mg-ACC, Mhc Mhc, Mg-Cc Mhc, Mg-Cc

Page 27 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

588

Table 3

589

Mineralogical composition from XRD-patterns of precipitates sampled at certain reaction times.

590

The compositions refer to precipitates, which are completely transformed to crystalline solids

591

after sampling. Quantifications were realized by Rietveld refinement. Mg-Cc: Mg-calcite, Arg:

592

aragonite, Vtr: vaterite; Mhc: monohydrocalcite; Nesq: nesquehonite. Experiment A1

A2

A3

C1

C2

C3

MgCa3

time min/d 25 min 180 min 1d 25 min 180 min 1d 25 min 180 min 1d 25 min 180 min 1d 25 min 180 min 1d 25 min 180 min 1d 25 min 180 min 1d

Mg-Cc wt. % 97 96 95 74 72 75 61 61 61 100 100 100 100 100 100 100 100 13 8

Arg wt. % 3 4 5 13 18 25 18 24 33 -

Vtr wt. % 14 9 21 15 6 -

593

594

27 ACS Paragon Plus Environment

Mhc wt. % 100 86 87 79

Nesq wt. % 14 13

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 32

595

596 597

Figure 1. Waterfall plot of in situ Raman spectra showing the vibration band of aqueous HCO3-

598

and CO32- as well as of v1 band of CO32- related to Mg-ACC, Mg-calcite (Mg-Cc) and

599

monohydrocalcite (Mhc) and ATR-FTIR spectra of precipitates sampled during certain reaction

600

times of the experiment A1 (A) and C1 (B).

28 ACS Paragon Plus Environment

Page 29 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

601 602

Figure 2. SEM images of precipitates sampled after 1 day of reaction time from experiments of

603

the present study and from experiments MgCa4, MgCa6 and MgCa8 of Purgstaller et al.40

604

showing Mg-calcite (Mg-Cc), aragonite (Arg) and monohydrocalcite (Mhc) crystals.

605 606 607 608 609

29 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 32

610 611

Figure 3. (Left) Concentration of dissolved Mg ions of the reactive solution, [Mg]aq, versus

612

experimental run time of experiments conducted at pH 7.8 (A) and pH 8.8 (B). (Right) Mg

613

content of the precipitated solid, [Mg]solid, calculated according to Eq. 1 versus experimental run

614

time of experiments conducted at pH 7.8 (C) and pH 8.8 (D). Titration of the (Ca,Mg)Cl2

615

solutions into the NaHCO3 solutions was stopped at 25 min of experimental time (dashed line).

616

When not visible, error bars are smaller than the symbols.

617 618

30 ACS Paragon Plus Environment

Page 31 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

619 620

Figure 4. Evolution of the saturation index with respect to ACC (SIACC) over the experimental

621

time of the reactive solutions of experiments conducted at pH 7.8 (exp. A1 to A3), and 8.8 (exp.

622

C1 to C3).

623

31 ACS Paragon Plus Environment

Crystal Growth & Design

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 32

624 625

Figure 5. (A) Pathways of Mg-ACC transformation in the reactive solutions depending on pH

626

and initial Mg/Ca ratio. (B) Activities of Ca2+ and CO32- in reactive solutions after the titration of

627

the (Ca,Mg)Cl2 solution was stopped and before transformation of the amorphous to the

628

crystalline phase took place. (C) Initial Mg/Ca ratio versus the prevailing ratio of Mg2+ to Ca2+

629

activities of reactive solutions, aMg2+/aCa2+, after the titration of the (Ca,Mg)Cl2 solution was

630

stopped and before the transformation of the amorphous to the crystalline phase took place (D)

631

Reaction time at which Mg-ACC transformation was indicated by in situ Raman spectroscopy

632

(tACC) versus aMg2+/aCa2+. Note that, except for experiment MgCa4, the aMg2+/aCa2+ values obtained

633

in experiments MgCa5 and MgCa6 conducted at pH 8.3 by Purgstaller et al.40 could not be

634

displayed in Figure 5C-D because the transformation of crystalline carbonate minerals took place

635

before the titration of the (Mg,Ca)Cl2 solution was stopped.

636

32 ACS Paragon Plus Environment