CRYSTAL SETTLING AND INSITU CRYSTALLIZATION IN AQUEOUS-SOLUTIONS AND MAGMA CHAMBERS

Laboratory experiments are described in which aqueous solutions of potassium nitrate are cooled from above. Depending on the viscosity of the solution and the rate at which it is cooled, the mechanism of crystallization varies. In the least viscous, most slowly cooled experiment, after an initial period, crystals nucleate and grow on pre-existing crystals on the floor of the tank driving vigorous compositional convection. In more viscous or more rapidly cooled experiments, in addition to crystallization on the floor, crystals nucleate and grow at the top of the fluid. These crystals settle through the solution adding to the crystals on the floor. The crystallization mechanism adopted by the laboratory experiments depends on the degree to which the solutions become supersaturated. At low cooling rates and viscosities the critical supersaturation for homogeneous nucleation is not exceeded: crystals do not form within the fluid and crystallization occurs only on the floor. At higher cooling rates and viscosities the critical supersaturation is exceeded. In this case crystals are able to form at the top of the fluid and crystal settling operates in addition to crystallization on the floor. Applying a similar argument to magma chambers, it is suggested that in situ crystallization may occur on its own in the most slowly cooled, least viscous mafic magma chambers. In other chambers crystal settling is expected to accompany in situ crystallization. A quantitative assessment of the crystallization of an olivine-saturated mafic magma is presented. For in situ crystallization to occur alone (i.e. without any superimposed crystal settling) in a chamber which is cooled by heat conduction through the country rock, supercoolings of between 1° and 20°C must be maintained in the chamber without homogeneous nucleation occurring. If crystal settling does occur, then the crystals may not form in the thermal boundary layer because, although it is the coldest part of the magma it is not necessarily the most supersaturated part because of the effect of pressure on the liquidus. © 1990.