Structural and other contributions to the third-law entropies of silicates

The true third-law entropies of many minerals are frequently quite different from those values derived from calorimetric measurements. The discrepancy can usually be attributed to neglected residual or unextracted entropies related either to site-mixing and molecular disorders or to the lack of significant magnetic ordering at those temperatures reached by the heat capacity measurements. A literature review indicates that many silicates present site-mixing and vacancies in one or several of their crystallographic sites. The effect on entropy is well known in feldspars, but residual entropies of similar or greater magnitudes are also present in many amphiboles, micas, chlorites, zeolites, scapolites, feldspathoids, and other silicates. Less conspicuously, disorder in water molecules or hydrogen bonding may be responsible for yet another frequently overlooked entropy contribution. Unextracted entropy results from limited heat capacity measurements so that magnetic ordering effects to be expected in minerals with transition metals are either not registered or only incompletely recorded. In some cases, significant magnetic ordering probably only takes place at temperatures well below 15 K.The discussion in this paper centers on the causes resulting in discrepancies between calorimetric and third-law entropies. A set of tables reproduces the crystallographic information for most important rock-forming silicates and indicates the entropy contribution arising from site-mixing and vacancies, and possible magnetic ordering in those substances with transition metals. In addition, most elements appear in several isotopic forms, and it is this effect that gives rise to isotopic site-mixing and thus to another configurational entropy. It can on the whole be neglected. A discussion centered on the system fayalite-iron oxides (wüstite, hematite, magnetite) indicates the uncertainties involved in deriving third-law entropies from either equilibrium data or calorimetric investigations, which is especially relevant when dealing with substances presenting vacancies (wüstite), transition metals, and the possibility of magnetic ordering. The published entropies of many minerals are probably only approximations to true third-law values and should be checked against structural and magnetic information.