The effect of experiment geometry on the mechanism and rate of dissolution of quartz in basanite at 0.5 GPa and 1350 °C

Abstract

Mineral dissolution is an important factor in many magmatic processes such as melting, assimilation and magma mixing. Since it is not possible to determine dissolution rates or mechanisms from natural samples, experimental measurements are very useful. However, the geometry of the crystal–melt system can have a large effect on the measured rate, depending on whether the contaminated melt formed during dissolution is gravitationally stable or unstable. This study examines the effects of the crystal–melt geometry on the dissolution rate and mechanism. The experiments were performed using basanite melt and cylinders and spheres prepared from a single crystal of natural quartz. All of the experiments were performed in the piston cylinder apparatus at 0.5 GPa and 1350 °C. Four crystal–melt geometries were used: (1) quartz cylinders on top of a column of melt; (2) quartz cylinders beneath a column of basanite melt; (3) quartz cylinders in the middle of column of melt; (4) quartz spheres on top of a column of basanite melt. These geometries allow an examination of non-convective, convective and mixed non-convective/convective dissolution. Sphere experiments were included, as this has been the most commonly used geometry in previous experimental studies. In all of the experiments quartz dissolves directly into the basanite without formation of cristobalite or tridymite. Quartz on top of a column of melt dissolves at a rate almost proportional to the square root of time and forms a silica-rich compositional boundary layer that is gravitationally stable. All of the samples show well-defined compositional gradients in the boundary layer; however, the melt at the interface varies in composition with time and plots of concentration as a function of distance normalized to time show that the diffusion rate of SiO2 increases with time. These data suggest that the rate-controlling step during quartz dissolution is interface reaction rather than cation diffusion. Quartz on the bottom of a column of basanite dissolves much more quickly than in the quartz-on-top experiments and the dissolution rate is linear, due to the periodic gravitational instability and resultant convection of the boundary layer. Even though interface kinetics are the rate-controlling step in quartz dissolution, convection causes an increase in dissolution rate because it replenishes the boundary layer with new, silica-undersaturated melt, which dissolves the quartz more quickly than the contaminated melt. These data suggest that the interface reaction rate is controlled by the degree of undersaturation of the solvent melt in the dissolving component. Both quartz-in-middle and quartz sphere experiments dissolve at a rate intermediate between the two extremes and both show a power law rate. Both dissolve by a combination of convective and non-convective dissolution but the sphere experiments are affected by an additional factor. During the experiment the sphere can sink through the capsule causing forced convection which adds another complication to the interpretation of the dissolution rate data. The results of this study indicate that the choice of experiment geometry plays a major role in determining the observed dissolution rate. Mineral spheres, which have been widely used in the past, are not ideal for dissolution studies. Instead, dissolution rates and mechanisms are best determined in the absence of convection. These experiments have an additional advantage in that for diffusion-controlled dissolution, they allow determination of cation diffusivity.