Experimental simulations of anatexis and assimilation involving metapelite and granitic melt

Abstract

Assimilation of foreign material into a granitic magma body entails mixing between two end-member components: the initially solid assimilant, and the original magma, whose composition is mostly represented by a granitic liquid. In this study, we assess the interactions between a haplogranite liquid made to the composition of the minimum at 200MPa H2O (Ab38.23Or28.72Qtz33.04Crn00.01) with a schist composed of muscovite and quartz with minor biotite and garnet and traces of ilmenite, plagioclase, apatite, and graphite, from 700° to 800°C at 200MPa±aqueous-carbonic fluid. The primary focus of this work is to assess the changes in compositions of granitic liquids that form in each reservoir as a result of the assimilation of schist by granitic magma. Before doing so, we evaluate the experimental database for the compositions of granitic liquids in equilibrium with peraluminous mineral assemblages. Most experiments that entail partial melting of peraluminous mineral assemblages begin with finely milled powders. This method yields melt pools (glass) too small for accurate analysis by electron beam methods. As a result, most published analyses of glasses from such experiments are more aluminous and less sodic than their actual compositions. The analytical errors are large, up to 50% relative to the mean value for some elements. In addition, the experimental melts are commonly more hydrous than reported, and vapor-saturated, even in experiments that are cited as “vapor” or “fluid-absent.” Experiments that combine sand-sized mineral grains and powdered granitic glass yield the best approach to chemical equilibrium and facilitate accurate analysis of the glass. Glasses from such experiments define the compositions of liquids that may be derived either from the anatexis of aluminous metasediments, or by the assimilation of aluminous rocks into a granitic magma whose composition is approximately represented by the minimum composition in the haplogranite system Ab–Or–Qtz–H2O.New experiments designed to simulate the assimilation of metapelites by granitic melt juxtaposed a solid core of anhydrous granitic glass with a solid core of mica schist, both as described above. At comparable conditions, melting within the mica schist alone is far less extensive than when coupled with the granitic liquid. This is because field diffusion, the long-range and highly organized migration of H, Na, and K throughout the entire volume of interconnected melt, quickly brings the composition of silicate liquid to chemical equilibrium across the metapelite–granite interface with respect to these components and in regard to the ASI, Aluminum Saturation Index and the H2O content of the melt. The most significant change entails the diffusion of Na from granitic liquid into the metapelite, which shifts the bulk composition of the metapelite toward the minimum in the granite system, and, in turn, promotes a greater proportion of melting within the metapelite. A weaker counter-flow of K back into the haplogranite raises its K* (molar K/[Na+K]) and lessens the increase in ASI (molar Al/[Na+K]) due to Na loss. The solubility of H2O in granitic liquid is positively correlated with the ASI value. Hence, changes in the ASI of melt that follow from field diffusion of H, Na, and K ions also strongly affect the aH2O within the coupled metapelite–granite system.