Catching a collapsing solidification front through thermal gradient experiments

Large explosive eruptions commonly emplace differentiated (i.e. rhyolite, phonolite), crystal-poor juveniles. This is a well-known paradox in volcanology, considering that magmatic differentiation implies crystallization and that crystal-melt separation processes (e.g. crystal settling) are more efficient in high-temperature, primitive magmas. Conversely, differentiated, crystal-poor juveniles are usually associated to shallow, thermally-zoned feeding systems. Here, the generation of differentiated, crystal-poor magmas may be explained throughout the development of a “solidification front” at the roof of the chamber. Although natural evidences and theoretical models support the solidification front concept, its capability to originate differentiated, crystal- poor magmas remains unconstrained. By experimentally investigating the formation of a solidification front in a thermally zoned environment we demonstrate its capability to originate glassy belts and pockets phonolitic in composition. We recognize in the instability and collapse of rigid crystal frame the driving mechanism producing segregation and upward accumulation of crystal-poor melts and suggest this model may apply to thermally zoned magma chambers