LIQUID IMMISCIBILITY IN THE JOIN NAALSI3O8-CACO3 TO 2-5 GPA AND THE ORIGIN OF CALCIOCARBONATITE MAGMAS

Abstract

Field evidence from intrusive and effusive carbonatites supports the existence of calciocarbonatite magmas. Published experimental evidence in the model system Na2O−CaO−Al2O3−SiO2−CO2 indicated the formation of nearly pure (99%) CaCO3 immiscible liquids from a carbonated silicate liquid. This evidence has been used to support interpretations of extremely CaCO3 -rich calciocarbonatite magmas, and immiscible liquids with compositions of almost pure CaCO3 in metasomatized mantle peridotite and eclogite. Detailed phase relationships are constructed in the model system, based on phase fields intersected by the join NaAlSi3O8 −CaCO3 (Ab-CC) at 1.0, 1.5, and 2.5 GPa between 1100 and 1500°C, and analyzed immiscible liquids. The miscibility gap between silicate-rich liquid and carbonate-rich liquid intersected by the join Ab-CC contracts considerably with decreasing pressure: 2.5 GPa, between Ab10 CC90 (by wt%) and Ab65 CC35 above 1310°C; 1.5 GPa, between Ab23 CC77 and Ab43 CC57 above 1285°C; 1.0 GPa, not intersected. The liquidus piercing point between calcite and silicates becomes enriched in CaCO3 with decreasing pressure, from Ab80 CC20 at 2.5 GPa to Ab47 CC53 at 1.0 GPa. No immiscible liquid contains more than ∼80% dissolved CaCO3 , and all contain at least 5% Na2CO3 . A round CaCO3 phase exhibiting morphology similar to that displayed by immiscible liquid globules is determined to be crystalline calcite under experimental conditions. The topology of the phase fields and field boundaries illustrates the kinds of processes and controls existing in magmatic systems. Calciocarbonatite magmas cannot be produced by equilibrium immiscibility process in the mantle. Carbonated silicate magmas in the crust yield residual calciocarbonatite magmas by fractionation along the silicate-calcite field boundary, reached either directly from the silicate liquidus or more commonly via the miscibility gap. Immiscible carbonaterich magmas when freed from the silicate parent cool down a sleep silicate liquidus until they reach a silicate-carbonate field boundary. There is no experimental evidence for immiscible calciocarbonatite magmas with > 80% CaCO3 , and calcite lapilli cannot be formed from 99% CaCO3 magmas. Sovites are surely cumulates.