Since discovery of the mineral dolomite [CaMg(CO3)2] more than 200 years ago (de Dolomieu, 1791), an immense volume of scientific research has been generated on the various genetic origins of dolomite and dolostones (refer to reviews by Van Tuyl, 1916; Murray & Pray, 1965; Friedman & Sanders, 1967; Land, 1985; Given & Wilkinson, 1987; Hardie, 1987; McKenzie, 1991; Mazzullo, 1992; Kupecz et al., 1993; Sun, 1994; Budd, 1997; Warren, 2000; Machel, 2004; McKenzie & Vasconcelos, 2009). The ongoing discussion concerning the origin of dolomite is due, in large part, to the fact that dolomites commonly yield genetically ambiguous geochemical characteristics with regard to the temperatures and chemical compositions of the fluids in which they formed. Furthermore, understanding the petrogenesis of natural dolomite requires determination of how the process or processes of recrystallization may have altered initial precipitates. Coarse-grained, well-ordered and near stoichiometric dolomites with relatively negative oxygen isotopes in rocks with features consistent with deposition in peritidal environments have been inferred to have recrystallized from finer crystalline, poorly ordered, calcium-rich and isotopically heavier dolomite crystals (e.g. Land, 1980, 1985; Montanez & Read, 1992; Reinhold, 1998; Al-Aasam, 2000). Many of the inferred instances of recrystallization are predicated on reasonable assumptions about characteristics of solids that were recrystallized. For example, Frisia (1994) compared textures, stable carbon and oxygen isotopes, trace elements and transmission electron microscope microstructures of peritidal Jurassic dolomites with modern dolomites from a sedimentologically similar setting in Abu Dhabi. On the basis of sedimentology, not specific textural or chemical characteristics of the dolomite, Frisia (1994) inferred that the initial Jurassic dolomites were replaced by later dolomites. The only direct observational evidence of recrystallization in natural, low-temperature dolomites is the cathodoluminescence microscopy that demonstrates replacement of one generation of dolomite by a later generation (e.g. Montanez & Read, 1992). However, unquestionable evidence of one generation of dolomite replacing another is rare.
The most common argument for recrystallization of ancient dolomite is that if the physical and/or chemical characteristics, such as cation ordering, stoichiometry, trace element composition, C and O stable isotopes and/or microstructures, of a particular ancient dolomite believed to have formed in a certain depositional or diagenetic setting (for example, arid tidal flat), are different from characteristics of modern dolomite found in that setting, then the ancient dolomite has recrystallized. Land (1980) and many others since (e.g. Banner et al., 1988; Reinhold, 1998; Montanez & Read, 1992; Smith & Dorobek, 1993; Al-Aasam, 2000; Kirmaci & Akdag, 2008; Ronchi et al., 2011) have used this argument to explain differences between cation ordering, stoichiometry, trace element composition and stable isotopes of ancient and modern dolomites. Given the available evidence from the rock record, the inferences of recrystallization cited above may be the best argument by abduction, but these arguments ultimately rest on the untestable assumption that the original dolomite characteristics were similar to modern analogues. This assumption may be of limited value because it has been demonstrated that modern dolomites have a wide range of chemical and physical characteristics, which may be related to different solution compositions, temperatures and crystal growth rates (e.g. Rosen et al., 1989). This paper seeks to improve understanding of dolomite recrystallization, and to evaluate the various lines of geochemical and petrographic evidence commonly used to infer recrystallization in ancient dolomites. The specific objectives of this work are to:
- Provide an explanation for why some high-temperature dolomitization experiments show a change in stoichiometry with reaction progress and other experiments do not.
- Test whether cation ordering increases in the absence of an ambient solution. If dry samples do not recrystallize, then it is most likely that all dolomite recrystallization reactions are dissolution/reprecipitation reactions.
- Use a significantly larger data set than Kaczmarek & Sibley (2011) to test the hypothesis that the rate of increase of cation ordering in dolomite is the same prior to and after complete replacement of calcite reactants.
- Use the same larger data set to more accurately determine the relative timing of stoichiometry, cation ordering and nanotopography changes that occur during the dolomitization reaction.
- Define a model of dolomitization that incorporates both replacement of a CaCO3 precursor and recrystallization of the initial replacement dolomites.