EVIDENCE FOR LIQUID IMMISCIBILITY IN LADDER DIKES AND LAYERED GRANODIORITES, SIERRA NEVADA, CALIFORNIA

Ladder dikes (LDs) and laterally continuous modal layers (MLs) are common features of granodiorites worldwide. Although a variety of explanations have been offered for these phenomena, all invoke standard processes of crystal-liquid separation such as gravitational settling and shear flow. Here we show, based on field work, whole-rock geochemistry, and microprobe analysis, that the geochemistry of LDs and modal layering is inconsistent with crystal-liquid separation but bears a striking resemblance to mafic liquids produced by liquid immiscibility.

LDs are roughly tabular bodies, <1m thick, composed of nested, concave-up, mafic-felsic troughs with an overall tabular, dike-like form (see Bartley et al., this volume). MLs are decimeter-scale couplets composed of a sharp-sided mafic layer that grades into a felsic layer that is in turn in sharp contact with another mafic layer. Both LDs and MLs are greatly enriched in mafic minerals, including biotite, hornblende, magnetite, titanite, apatite, and zircon, leading to extreme enrichments in many elements including REE, Fe, Zr, P, and Ti; Fe₂O₃^t in some LD mafic layers approaches 35 wt%, and La approaches 1000x chondrite. Mafic layers in LDs and MLs also have low Mg numbers (typically <0.4). In contrast, crystallizing assemblages in granodioritic magmas should be low in these elements and have high Mg numbers owing to preferential incorporation of Mg over Fe in the crystallizing assemblage. Biotite and hornblende crystals are typically more Fe- and apatite-rich than those of the host plutons. All of these features argue that the minerals in the layers were not derived from the surrounding pluton by crystal-liquid separation.

Fe-enrichment trends are dramatically different from predictions of crystal-liquid equilibria but match observed segregation of Fe, REE, P, etc. into Fe-rich immiscible liquids in both experiments and in volcanic rocks on the earth and moon. Although immiscibility is difficult to detect in plutonic rocks owing to crystallization of the immiscible liquids, it has been proposed for several layered mafic complexes, and LDs and MLs carry the same geochemical fingerprint. The immiscibility hypothesis cannot explain all the peculiar field relations of LDs and MLs, but neither can other explanations advanced to date.