Geochemistry of dikes and lavas from Hess Deep: Implications for crustal construction processes beneath mid-ocean ridges, and, the stable-chlorine isotope geochemistry of mid-ocean ridge basalt glasses.

摘要

The Hess Deep rift in the eastern equatorial Pacific exposes a cross-section through the upper oceanic crust that parallels a flowline from the East Pacific Rise. As such, the exposures at Hess Deep represent one of only four tectonic windows into the fast- to intermediate-spread oceanic crust including Deitz Deep, Pito Deep and the Blanco Fracture Zone. Observations from digital images obtained using Argo II as well as observations and sampling during manned submersible dives aboard Alvin along a 25-km exposure of the northern wall of the Hess Deep rift basin record ∼400-ky of spreading history along the flowline (135 mm/yr full spreading rate). The detailed observation and spatial control on sample location and field relationships presented in Part I of this dissertation are unique among investigations of fast-spread oceanic crust, and thus, the results discussed here can be placed into a four-dimensional context with a level of confidence hitherto impossible in studies of fast-spread oceanic crust.; Part I constitutes a study of the petrogenesis and spatial systematics of basaltic dikes and lavas recovered from Hess Deep with implications for the magmatic accretion of the upper oceanic crust. The spatial systematics of dike compositions across the 25 km flowline, and of adjacent dikes in particular, reveal a non-systematic distribution of dike compositions that suggests a four-dimensional model of crustal construction whereby dikes emanate from chemically distinct portions of the axial melt lens and are injected into the upper crust for 10s of kilometers along axis. The majority of lavas recovered are chemically distinct form the dikes and this difference is consistent with accumulation of plagioclase in the magmas which decreases magma density and aids their eruption as lavas. The distinction between dikes and lava compositions, implies that the lavas alone do not accurately record the true range of magma compositions. Furthermore, the along axis transport of magma in dikes further implies that lavas may not directly overly the magma chamber from which they originate. Thus, studies of magmatic processes focusing on lava chemistry lack spatial control on the origins of the lavas, and do not sample the true range of compositional variability.; In Part II the analytical method for the determination of stable-Cl isotope composition in low-Cl abundance samples is evaluated and results from mid-ocean ridge basalts are presented. We demonstrate improvements in the sample preparation procedures for stable-Cl isotope analysis using experimental results (Chapter 5). Here we demonstrate that sample purification can be improved by simple modifications that result in removal of >95% of Fluorine and Sulfate from the sample solutions leaving a nearly pure aqueous solution of Chlorine for analysis by thermal ionization mass spectrometry. Correlations between stable-Cl isotope composition and mantle enrichment signatures in mid-ocean ridge basalts are documented in Chapter 6 and a distinction is made in the Cl isotope composition between mid-ocean ridge basalt with recycled crust signature and mid-ocean ridge basalt with relatively un-degassed more primordial mantle signature. Thus, stable-Cl isotope determinations in MORB represent a powerful tool not only for distinguishing distinct mantle enrichment components but also providing mechanisms for creating the enrichment in the mantle (e.g. subduction derived fluids, residual slabs, primordial un-degassed mantle).