The influence of crystal settling on the compositional zoning of a thin lamprophyre sill: A multi-method approach

We have studied a visibly zoned, thin (< 0.5 m) lamprophyre sill that crops out in the Catalonian Coastal Ranges (NE Spain). The sill is a camptonite composed of large abundant crystals, mainly of clinopyroxene and amphibole, set in a fine-grained groundmass. The mineral chemistry of the different crystal populations indicates that the large crystals are inherited antecrysts incorporated into the magma before emplacement. The major and trace element whole rock profiles are S-shaped, with the development of a marginal reversal in the lower chilled margin. These profiles cannot be explained by normal fractionation of the magma inwards. Instead, the whole rock zoning is controlled by the presence of antecrysts. This is proven: 1) quantitatively, through a trace element model which evaluates the contribution of the antecrysts to the overall composition of the rock, and 2) statistically, through a principal component analysis on the complete trace element data set. The mineral and groundmass compositions show rectilinear compositional profiles, indicating that the magma was emplaced in a single pulse. The accumulation of the antecrysts towards the bottom of the sill, together with the calculation of settling velocities for clinopyroxene and amphibole and cooling velocities for the magma, indicate that the settling of antecrysts during cooling is responsible for the varying proportions of antecrysts and therefore for the whole rock compositional zoning. This study proves that crystal settling is a significant process in triggering compositional zoning of igneous intrusions even at the cm-scale, provided that the magma carries large crystals upon emplacement.

► Crystal settling is a meaningful process even in small intrusions.
► Antecrysts control the variations in whole rock compositions.
► A geochemical model quantifies the contribution of the antecrysts to the rock.
► Principal component analysis is useful for studying geochemical data sets.