White micas in subducted crust: their role in the subduction zone nitrogen cycle and as a tracer of fluid processes

Subduction is the primary mechanism for transferring volatiles from the Earth’s surface to the mantle. White micas (muscovite, phengite, paragonite) can be significant mineral hosts for a number of volatiles, including water, nitrogen, boron and halogens, in a range of subducted crustal lithologies. This thesis explores the role of white micas in transporting nitrogen and other fluid-mobile elements in subducted crust, and, more generally, their utility as a recorder of fluid-rock interaction, through the study of formerly subducted and subsequently exhumed metamorphic rocks. I present the first in situ measurements of N concentrations in minerals from subduction-related rocks, using secondary ion mass spectrometry. Mineral N concentrations in a suite of rocks representing formerly subducted oceanic crust confirm that white micas contain the highest N concentrations of any common high pressure mineral. However, when the concentrations and modal abundances of each mineral are accounted for, omphacite, amphiboles and chlorite also host a significant fraction of the N budget in some samples. The role of paragonite as a host for N, and other volatile elements, may be underappreciated compared to phengite, which is the more common white mica phase. Comparison of bulk rock N concentrations with mineral N concentrations reveals that non-mineral hosts for N such as fluid inclusions can also be significant. By comparing the behaviour of N to other fluid mobile elements in samples that record high pressure fluid-rock interaction, I am able to place constraints on the fluid-mica partition coefficient of N at high pressure. These are the first natural constraints on this parameter and they suggest that the assumption made by some previous studies, that N behaves similarly to Rb, is incorrect, and N is in fact more fluid mobile. The stability of micas during fluid-rock interaction has a major effect on the retention or loss of N, B and other mica-hosted trace elements during subduction. I also present the first inter-mineral partition coefficients amongst white micas, omphacite, Na-Ca amphiboles and chlorite, showing that N partitions strongly into white micas over a range of subduction-related pressure-temperature conditions. Through several case studies of rocks that record high pressure fluid-rock interaction, I demonstrate that the analysis of in situ N concentrations, B concentrations and B isotopes, and other fluid-mobile element concentrations in white micas can be an effective tracer of fluid sources in high pressure rocks, particularly for sediment-derived fluids. These tracers are then applied to a suite of formerly subducted oceanic crustal rocks and show that fluid-mediated N transfer from sediments into oceanic crust during subduction appears to be a relatively common process, but some samples can also retain N that was acquired during seafloor alteration. Different styles of seafloor alteration may increase or decrease the ability of oceanic crust to retain N during subduction, depending on their effect on the high pressure mineralogy.