Modelización geoquímica de los procesos de fusión parcial

Abstract

During melting processes both stable isotopes and trace elements fractionate, whereas radiogenic isotopes do not change. The distribution of the former between the phases that participate, follows diluted solutions laws in such a way that it is possible to establish relatively simple equations to model these processes. Additionally, the radiogenic isotopes behaviour implies that the magmas retain the source signature thus allowing its identification. In the case of equilibrium melting, the highly incompatible elements abundance is very high in the liquid at the beginning of the process and decreases progressively as the melting degree increases. On the contrary, the concentration in compatible elements grows very slowly during the first steps to increase sharply for the highest F values. During fractional melting, the first liquid generated removes almost all the incompatible elements thus producing a relative depletion in those elements in the successive liquids. In the case of incongruent melting, the magmas are enriched in the trace elements with high distribution coefficients for the phases that melt and low for the newly generated phases, and are impoverished in the elements that constitute the new phases. If melting is produced in the presence of a fluid phase, the liquid will be depleted in those elements with fluid/liquid distribution coefficients close to 1, relative to the same liquid generated without a fluid phase. Finally, during disequilibrium or nondiffusive melting, the first liquid fraction has a concentration in incompatible dements higher and in compatible elements lower than that in the source, so the solid-liquid interface is depleted and enriched, respectively. However, at the end of the process the concentration of elements in the liquid is equated to the abundance in the solid that melted. To model equilibrium me1ting two different approaches can be followed, depending on the availability of the mineral-liquid distribution coefficients and the percentages in which the mineral phases have participated. When these parameters are known, it is possible to duplicate the concentrations observed in the primary liquids by selecting reasonable constants. On the contrary, when these parameters are unknown the approach to follow will depend on the knowledge of the composition of the liquids or that of the residue. In the first case, the element concentrations of the rocks are used to obtain a simultaneous best-fit solution of a system constituted by tile equations that describe the process, either for a number of elements, or individually for each parameter and element. If the composition of the residue is known, it is necessary to guess the composition of the protolith. Then, from the most residual element the two remaining parameters (the residue- melt bulk distribution coefficient and the degree of melting of each restite) are defined, either assuming the degree of melting, the elements bulk distribution coefficient, or their concentration