The role of thermal migration and low-temperature melt in granitoid formation: Can granite form without rhyolitic melt?

The bulk composition of silicic igneous rocks hovers directly over the minimum melt located on the NaAlSi₃O₈-KAlSi₃O₈-SiO₂ ternary. It is universally accepted that these rocks are igneous, reflecting the thermodynamic equilibrium between minerals and melt. However, there is a contradiction between the use of this phase diagram and current models of differentiation: fractional crystallization or partial melting (or any mechanical separation process) implies granites were melts extracted from a quartz-bearing residue.Here, I propose that a thermodynamically controlled process, wet thermal migration, provides a more consistent explanation for producing silicic rocks. This process, when coupled with slow incremental emplacement of sills, allows magmatic differentiation to take place without mechanical separation and produces minimum-melt compositions from input magmas not saturated in quartz. Examination of phase equilibria in the SiO₂-Al₂O₃-Na₂O-K₂O-H₂O system, as well as assessment of thermodynamic models, provides a template for understanding the wet thermal migration process (diffusion-based differentiation of crystal-mush in a temperature gradient). Phase equilibria in SiO₂-Na₂O-K₂O-Al₂O₃ indicate a cotectic surface between quartz and alkali feldspar slopes down-temperature as melt peralkalinity increases. Experiments show quartz and two feldspars coexist with a single water-rich (>40 wt.% H₂O) melt at 400°C and 0.1 GPa. Modelling suggests that development of water-rich melt at the hot end of a temperature gradient drives the process. Given the propensity of alkalis to rapidly diffuse down-temperature gradients and form these melts, small amounts (<5%) of interconnected melt form, leading to the differentiation of partially molten materials by wet thermal migration.Because the quartz-feldspar cotectic ranges from 650°C to 330°C, granites can reflect formation by a process that never involves the existence of rhyolitic melt. If this is correct, the process has implications for understanding the formation of plutons and batholiths and, therefore, the continental crust.