Anatectic granites in their source region: A comparison between experiments, thermodynamic modelling and nanogranitoids

Anatectic granites represent the products of crustal maturation by partial melting, therefore, studies on their origin are crucial to understanding how continental crust forms and evolves. This contribution compares composition of melts obtained from three tools allowing the investigation of anatectic granites in their source region: experiments, phase equilibria modelling and nanogranitoids. Benchmark experimental runs are selected along with nanogranitoids from the Ivrea–Verbano Zone (IVZ; NW Italy) and compositions of melt are then calculated using the same bulk rock compositions considered in experimental and natural case studies. The possible effects of melt loss, diverse P–T paths and variable equilibration volumes on melt composition are modeled for IVZ rocks. Results are then compared with those of previous comparative studies to verify the presence of systematic compositional divergences and similarities. Collectively, all data reflect pressure and temperature conditions (≈700–950 °C and 0.3–1.5 GPa) typical of the suprasolidus orogenic crust, with source regions ranging from metapelite, metagreywacke to metagranite. Contents of FeO, MgO, K2O and, to a lesser extent, CaO in the melt show the largest divergences. FeO, MgO and CaO tend to be enriched in experimental melts and nanogranitoids, whereas K2O is mostly overestimated in calculated melts. After a consideration of the common pitfalls for each approach we discuss the potential causes for the main geochemical discrepancies. The most extreme deviations of FeO and MgO, often observed in low-T (≤ 750 °C) melts, seem to be related to the selected melt models, which have been mostly calibrated on the basis of high-T (≥800 °C) experiments. Some experimental melts tend to be strongly depleted in K2O due to the metastable persistence of micas at much higher temperatures in laboratory runs, whereas the compositions of nanogranitoids in some investigated rocks may be partially controlled by diffusion in the melt. Disequilibrium melting of plagioclase may explain the CaO enrichment observed in some experimental and natural melts. When local equilibration volumes are considered in the calculations, a better match is observed for some oxides (e.g. FeO and MgO) between nanogranitoids and their calculated counterparts, even though other significant divergences remain unaffected. The combination of diverse petrologic tools, with proper recognition of their shortcomings, represent the best approach to shed light into the complexities of melting processes in nature which control the composition of anatectic granites at the source.