Chlorite stability in the subduction zone: Implications for water transport to the deep mantle, slab diapirs and mantle melting

Abstract

Hydrous minerals subducted at convergent margins provide the bulk of water to arc magmas. At temperatures beyond serpentinite stability, chlorite becomes the main water-bearing contributor in the subduction zone. The maximum stability field of chlorite is not well constrained, and the experimental studies which have been conducted have involved synthesis experiments using chlorite of clinochlore composition which may not adequately represent the variety of behaviour of naturally-occurring chlorite in a subduction setting. This thesis provides new data which successfully determined the upper stability field of chlorite in a range of subduction lithologies. Piston cylinder experiments were conducted over a range of pressures (1.0 GPa-6.3 GPa) and temperatures (500°C-1150°C) which were analysed using field emission scanning electron microscopy (FE-SEM), Raman and X-Ray Diffraction (XRD). Only natural mineral samples were used. Results confirmed those of previous research below ~4.0 GPa, when adjusted for compositional differences, but at higher P,T conditions, some unexpected findings were revealed. Series 1 experiments (Chapter 3) examined chlorite stability in ultramafic chlorite schists (Mg#=0.94) examining two terminal chlorite reactions Reaction 1: chlorite = orthopyroxene + olivine + spinel + H2O and Reaction 2: chlorite = garnet + olivine + spinel + H2O. Series 2 experiments (Chapter 4) explored chlorite stability in a fertile chlorite peridotite (Mg#=0.94) exploring two reactions reported in previous studies: Reaction 3: chlorite + clinopyroxene = garnet + olivine + H2O ± spinel, and Reaction 4: chlorite + orthopyroxene = garnet + olivine + H2O ± spinel. Series 3 experiments (Chapter 5) were completed in two parts. The first examined two samples of mafic chlorite schist (Mg#=0.50 and 0.68) to test the feasibility of chlorite schist forming mélange diapirs (Marschall and Schumacher, 2012) to transport fluid through the mantle wedge. The second part used chlorite schist (Mg#=0.94) to locate the wet peridotite solidus, the subject of much debate. Results of Series 1 and 2 experiments showed chlorite possessed a maximum pressure stability of 6.2 GPa in both chlorite schist and chlorite peridotite with the former lithology also possessing ~40°C higher thermal stability. This represented enhanced stability some 40 kms deeper than determined by previous research and places stable chlorite within the upper mantle. The high-pressure breakdown of chlorite produced some unexpected phases. Chlorite schist reacted to garnet, olivine, water and 11.5Å-phase (12.1 wt% water) whilst the transformation of chlorite peridotite formed olivine, water and Mg-sursassite (7.2 wt% water). Both hydrous minerals are stable to high P,T conditions and so in a cooler subduction settings, these phases could transport significant quantities of water to the deeper mantle. Results of Series 3 experiments showed mafic chlorite schist attained a maximum stability at 3.0 GPa of 780°C and 765°C respectively. Upon chlorite breakdown, mafic schist transformed to a garnet peridotite, a denser rock than average peridotite, which disproved the feasibility of the diapir model. The location of the wet peridotite solidus was determined to be >1100°C which disputed the existence of chlorite melting in hydrous peridotites.