The role of Mg in the crystallization of monohydrocalcite
Abstract
Monohydrocalcite is a member of the carbonate family which forms in Mg-rich environments at a wide range of Mg/Ca ratios Mg2+aq/Ca2+aq≥0.17<65. Although found in modern sedimentary deposits and as a product of biomineralization, there is a lack of information about its formation mechanisms and about the role of Mg during its crystallization. In this work we have quantitatively assessed the mechanism of crystallization of monohydrocalcite through in situ synchrotron-based small and wide angle X-ray scattering (SAXS/WAXS) and off-line spectroscopic, microscopic and wet chemical analyses. Monohydrocalcite crystallizes via a 4-stage process beginning with highly supersaturated solutions from which a Mg-bearing, amorphous calcium carbonate (ACC) precursor precipitates. This precursor crystallizes to monohydrocalcite via a nucleation-controlled reaction in stage two, while in stage three it is further aged through Ostwald-ripening at a rate of 1.8 ± 0.1 nm/h1/2. In stage four, a secondary Ostwald ripening process (66.3 ± 4.3 nm/h1/2) coincides with the release of Mg from the monohydrocalcite structure and the concomitant formation of minor hydromagnesite. Our data reveal that monohydrocalcite can accommodate significant amounts of Mg in its structure (χMgCO3 = 0.26) and that its Mg content and dehydration temperature are directly proportional to the saturation index for monohydrocalcite (SIMHC) immediately after mixing the stock solutions. However, its crystallite and particle size are inversely proportional to these parameters. At high supersaturations (SIMHC = 3.89) nanometer-sized single crystals of monohydrocalcite form, while at low values (SIMHC = 2.43) the process leads to low-angle branching spherulites. Many carbonates produced during biomineralization form at similar conditions to most synthetic monohydrocalcites, and thus we hypothesize that some calcite or aragonite deposits found in the geologic record that have formed at high Mg/Ca ratios could be secondary in origin and may have originally formed via a metastable monohydrocalcite intermediate.
High-Mg monohydrocalcite (χMgCO3 > 0.06) consists of individual nanometer-sized crystals (<35 nm) (Fig. 9c) with a significant part (6-25%) of its structural H2O being associated with the Mg ion, therefore displaying a progressive dehydration during heating to >500 °C (Fig. 9b). Such high-Mg monohydrocalcites are uncommon in nature, but can be synthesized in the laboratory at high initial supersaturation levels (SI > 3.25). Low-Mg monohydrocalcite (χMgCO3 < 0.06) which forms a “type 2” spherulite morphology. Less than 6% of the structural water in the low-Mg monohydrocalcite is bonded to Mg, so it fully dehydrates at low temperatures (150-200 °C). They have the same composition as natural monohydrocalcites reported in the literature, and can be synthesized in the laboratory at lower supersaturation levels (SI < 3.25). These observations indicate that despite their different morphologies (single nanometer sized crystals and low-angle branching spherulites, respectively; Fig. 9c and d) and levels of supersaturations at which they form, high- and low-Mg monohydrocalcite both crystallize via a nucleation-dominated growth process. The difference in particle size and morphology is likely controlled by the aqueous Mg concentration. At high concentrations, Mg poisons the formation of spherulites but still allows direct nucleation in solution, producing the non-aggregated, individual high-Mg monohydrocalcite crystals. At low supersaturations, the Mg concentration is low and monohydrocalcite forms via growth-front nucleation permitting the development of the low angle branching “type 2” spherulites.Combining the mechanistic results described above with chemical data from our on-line experiment, and data from other studies (Fig. S3) reveals interesting relationships. Firstly, our on-line experiment shows an increase in nanocrystal sizes during the secondary crystallization of monohydrocalcite, which is coupled with a significant decrease in χMgCO3 (from ∼0.26 to ∼0.065). This corresponds to the transition from high- to low-Mg monohydrocalcite, suggesting that the former would be metastable and rapidly transforming to the latter, possibly triggered by the removal of Mg from aqueous solution. Secondly, Davis et al. (2000) determined that the solubility of Mg-calcite (Ca1-xMgxCO3; x = 0-0.20) varies by approximately half an order of magnitude depending on the Mg content of the solid (Ksp = 10-8.0-10-8.5). A similar behavior should be expected for monohydrocalcite. The saturation indexes calculated for monohydrocalcite using the available solubility products from Hull and Turnbull (1973) and Kralj and Brečević (1995) are negative and show a difference of ∼0.55 (Fig. 8). We suggest that may be due to difference in the Mg contents of the monohydrocalcites used in their respective studies. This hypothesis is supported by the recent findings of Nishiyama et al. (2013), who reported that the solubility of synthetic monohydrocalcite increases with higher Mg/Ca ratios in the solid. They showed that the Ksp of monohydrocalcite can reach maximum values of Ksp = 10-6.77, which is almost one order of magnitude higher than the value of Hull and Turnbull (1973). Furthermore, Nishiyama et al. (2013) also reported an decrease in monohydrocalcite crystallite size (broadening of Bragg peaks in PXRD) with increasing solid Mg/Ca ratios, which again support our interpretation of the transition from high- to low-Mg monohydrocalcite.Although an in-depth study of the structural changes in monohydrocalcite as a function of Mg content are outside of the scope of this study, the changes in unit cell parameters during crystallization (Fig. 6) may be better understood when the χMgCO3 of the monohydrocalcite is taken into account. Regardless of Mg content, the unit cell volume remained virtually constant during stages II-III and only decreased slightly during stage IV (ΔV ≈ -0.2 Å3), in parallel with the decrease in monohydrocalcite χMgCO3. This small decrease in volume is a consequence of a mirrored change in a-axis and c-axis dimensions, which may be explained by the change in the monohydrocalcite internal structure during the loss of Mg and the transition from the high to low-Mg type (stage III to IV). The structure of monohydrocalcite is less dense and more open than calcite or aragonite (Effenberger, 1981; Neumann and Epple, 2007; Swainson, 2008), therefore it may more easily adapt to a changing Mg content.- Publication:
-
Geochimica et Cosmochimica Acta
- Pub Date:
- February 2014
- DOI:
- 10.1016/j.gca.2013.11.034
- Bibcode:
- 2014GeCoA.127..204R