Chemical heterogeneities in the mantle: Progress towards a general quantitative description

Abstract

Chemical equilibration between two different assemblages (peridotite type and gabbro-eclogite type) has been determined using basic thermodynamic principles and certain constraints and assumptions regarding mass and reaction exchange. When the whole system (defined by the sum of the two subsystems) is in chemical equilibrium the two assemblages will not be homogenized, but they will preserve distinctive chemical and mineralogical differences. Furthermore, the mass transfer between the two subsystems defines two petrological assemblages that separately are also in local thermodynamic equilibrium. In addition, when two assemblages previously equilibrated as a whole in a certain initial mass ratio are held together assuming a different proportion, no mass transfer occurs and the two subsystems remain unmodified. By modeling the chemical equilibration results of several systems of variable initial size and different initial composition it is possible to provide a quantitative framework to determine the chemical and petrological evolution of two assemblages from an initial state, in which the two are separately in chemical equilibrium, to a state of equilibration of the whole system. Assuming that the local Gibbs energy variation follows a simple transport model with an energy source at the interface, a complete petrological description of the two systems can be determined over time and space. Since there are no data to constrain the kinetics of the processes involved, the temporal and spatial scale is arbitrary. The evolution model should be considered only a semiempirical tool that shows how the initial assemblages evolve while preserving distinct chemical and petrological features. Nevertheless, despite the necessary simplification, a 1-D model illustrates how chemical equilibration is controlled by the size of the two subsystems. By increasing the initial size of the first assemblage (peridotite like), the compositional differences between the initial and the final equilibrated stage become smaller, while on the eclogite-type side the differences tend to be larger. A simplified 2-D dynamic model in which one of the two subsystems is allowed to move with a prescribed velocity shows that after an initial transient state, the moving subsystem tends to preserve its original composition defined at the influx side. The composition of the static subsystem instead progressively diverges from the composition defining the starting assemblage. The observation appears to be consistent for various initial proportions of the two assemblages, which somehow simplify the development of potential tools for predicting the chemical equilibration process from real data and geodynamic applications. Four animation files and the data files of three 1-D and two 2-D numerical models are available following the instructions in the Supplement.