Has the Classical Thermodynamic Concept of Solubility Lost Its Meaning for Carbonate Minerals in Complex Solutions

For well over half a century the application of classical chemical thermodynamic concepts to mineral solution-interactions has served well the advancement of aquatic geochemistry, not only for abiotic processes, but also in advancing our understanding of biomineralization. Probably no group of minerals has received more attention in this regard than carbonate minerals and in particular, calcite and aragonite. As a result, serious questions have arisen about the limitations of equilibrium thermodynamics as a useful predictive theoretical framework to describe carbonate mineral-solution interactions that have a direct bearing on biomineralization processes. One of the central questions is whether or not it is useful to apply equilibrium thermodynamics to very non-equilibrium mineral-solution interactions. The classical example of this problem is that pure calcite can't be in equilibrium with a multicomponent solution such as seawater; attempts to "solve" this problem through introduction of stoichiometric solubility constants arbitrarily fix compositional relations and offer neither insight nor solution. Simply put, if dissolution and precipitation reactions aren't the same you can't have thermodynamic equilibrium. This problem further manifests itself when solubility is predicted by extrapolating reaction rates to some apparent "kinetic" solubility that differs from that predicted thermodynamically and is a function of reaction inhibitor concentration. It thus becomes rather meaningless to formulate reaction kinetics in terms of thermodynamic free energy "distance" from equilibrium. Many other macroscopic examples could be given including problems with rate dependence of distribution coefficients, heterogeneity, nucleation and growth of a metastable phase on a stable phase, for example aragonite on calcite and in structural ordering involving double carbonates. However, the most recent and disturbing problems have arisen from the many rapidly evolving techniques for observing processes occurring in the near mineral-solution interfacial region. These observations reveal complex processes, often of a highly heterogeneous nature, on a nanoscale size range. The great challenge is how to integrate these into observations into a useful macroscale predictive formulation. It is our opinion that a potential path for doing so has been shown in chemistry where statistical mechanics have been used to provide molecular-based mathematical models capable of predicting macroscale processes in integrated form. This approach has particular relevance for advancing understanding of biomineralization where organisms employ strategies involving molecular level manipulation to "escape" what would be expected from equilibrium thermodynamics.