Fluid-Absent Melting Behavior of an F-Rich Tonalitic Gneiss at Mid-Crustal Pressures: Implications for the Generation of Anorogenic Granites

Abstract

Fluid-absent melting experiments on a biotite (20 wt.%) and hornblende (2 wt.%) bearing tonalitic gneiss were conducted at 6 kbar (900–975°C), 10 kbar (875–1075°C), and 14 kbar (950–975°C) to study melt productivity from weakly peraluminous quartzofeldspathic metamorphic rocks. At 6 kbar, biotite dehydration–melting is completed at 975°C via incongruent melting reactions that produce orthopyroxene, two oxides, and ˜25 wt.% granitic melt. At 6 kbar, hornblende disappears at 900°C, probably in reaction with biotite. At 10 kbar, biotite dehydration–melting produces <10 wt.% melt up to 950°C via incongruent melting reactions that produce orthopyroxene, garnet, and granitic melt. Hornblende disappears in the satne temperature interval either by resorption or by reaction with biotite. Widespread biotite dehydration–melting occurs between 950 and 975°C and produces orthopyroxene, two oxides, and ˜20 wt.% fluorine-rich (up to 0⋅31 wt.%) granitic melt. At 14 kbar only a trace of melt is present at 950°C, and the amounts of hornblende and biotite are virtually the same as in the starting material. At 975°C, hornblende is gone and ˜10 wt.% granitic melt is produced by melting of both biotite and hornblende. Our results show that hornblende-bearing assemblages cannot go through dehydration–melting on their own (although they can in combination with biotite) if the Ca content in the source rock is too low to stabilize clinopyroxene. In such rocks, hornblende will either resorb or melt by reaction with biotite. Under fluid-absent conditions, intrusion of hot, mantle-derived magmas into the lower crust is necessary to initiate widespread dehydration–melting in rocks with compositions similar to those discussed here. We argue that the high thermal stability of biotite in our starting material is caused mainly by the incorporation of fluorine. The relatively high F content in biotite in the starting material (0⋅47 wt.%) suggests that the rock has experienced dehydroxylation in its past. F enrichment by a previous fluid-absent partial melting event is excluded because of the lack of phases such as orthopyroxene and garnet which would have been produced. Our experiments show that the dehydration–melting of such F-enriched biotite produces F-rich granitic liquids, with compositions within the range of A-types granites, and leaves behind a granulitic residue dominated by orthopyroxene, quartz, and plagioclase. This study therefore supports the notion that A-type granites can be generated by H2O-undersaturated melting of rocks of tonalitic composition (Creaser et al., 1991), but does not require that these source rocks should be residual after a previous melting event.