The crystallisation history of normal mid-ocean ridge basalts from the Eastern Pacific Ocean and implications for the composition of primary mid-ocean ridge magmas : evidence from mineralogy, pillow-rim glasses and melt inclusion studies

McNeill, AW 1998 , 'The crystallisation history of normal mid-ocean ridge basalts from the Eastern Pacific Ocean and implications for the composition of primary mid-ocean ridge magmas : evidence from mineralogy, pillow-rim glasses and melt inclusion studies', PhD thesis, University of Tasmania.

 Preview
PDF (Whole thesis (published material removed))

| Preview
PDF (Whole thesis)
Full text restricted
Available under University of Tasmania Standard License.

Abstract

An understanding of MORB petrogenesis requires compositional data on primary melts, as these provide direct information on the PT-conditions and mechanisms of mantle melting and melt segregation beneath mid-ocean ridges. The reconstruction of primary melts from the compositions of erupted liquids (pillow-rim glasses) requires an understanding of the early crystallisation history of the primary melts. As recent models of MORB petrogenesis imply that not only early crystallisation, but, also extensive mixing of primary melts occurs, information on the early stages of evolution of the primary melts is even more crucial. Such information may be preserved by early formed phenocrysts and can be obtained by studying these phenocrysts and their melt inclusions. However, the ability of melt inclusions to preserve the compositions of trapped liquids needs to be examined (e.g., homogenised melt inclusions in cotectic phenocrysts should have the same compositions, and those trapped during a period of fractionation, as recorded by the pillow-rim glasses, should have compositions similar to the glasses).
The three MORB suites described in this thesis are from ODP/DSDP Holes 896A and 504B, Costa Rica Rift, and Sample D9-1 from the Gorda Ridge, in the east Pacific Ocean. Compositional and textural variations of phenocrysts (olivine, plagioclase and spine!) have been used to interpret the crystallisation history for each suite and, together with pillow-rim glass compositions, form the basis for interpreting the compositional variations of melt inclusions.
Samples from the Costa Rica Rift and the Gorda Ridge have phenocryst assemblages dominated by calcic plagioclase (>An$$_{88}$$) with lesser olivine (<Fo$$_{91.6}$$) and Cr-Al spinel. The magmatic histories of samples from both ridge segments are dominated by low pressure (<2 kb) fractionation (both crystal and in-situ), magma mixing and minor crustal assimilation. The high pressure fractionation of phases other than olivine is not an important process in the history of these suites.
The Hole 896A and 504B samples are strongly LREE-depleted, and record cotectic olivine+plagioclase crystallisation at <9.5 wt.% MgO. However, melt inclusions and phenocrysts record an initial period of olivine-only crystallisation from 1340°C, ~15 wt.% MgO and Fo$$_{91.6}$$ to ~1215°C, 9.5 wt.% MgO and Fo$$_{87}$$ followed by cotectic olivine (<Fo$$_{87}$$)+plagioclase (<An$$_{94}$$) crystallisation as recorded by the pillowrim glasses.
In contrast, the Gorda Ridge sample is less LREE-depleted, and melt inclusions and phenocrysts recorded cotectic olivine (<Fo$$_{90}$$) and plagioclase(<An$$_{94}$$) crystallisation from ~1230°C, 10.5 wt.% MgO. The melt inclusions and phenocrysts are related to the host pillow-rim glasses (Mg0<8.5 wt.%) by mixing and in-situ fractionation processes.
Melt inclusions in primitive phenocrysts do not preserve evidence for diverse melt fractions produced in a polybaric melting column. Crystallisation, therefore, commenced after aggregation of these liquids, or the trapped liquids are the result of fractionation from isobaric batch melts.
Olivine-addition calculations for Costa Rica Rift (Hole 896A) and Gorda Ridge samples indicate primary batch melts were in equilibrium with a mantle source (MORB pyrolite-90) at 18-20 Kb and this is confirmed by preliminary results of basaltperidotite sandwich experiments on a potential Hole 896A primary liquid (T. J. Falloon pers. comm., 1997).
Melting parameters (P$$_o$$, P$$_f$$, F$$_{max}$$ ) were estimated using published polybaric melting models. The major element (FeO* and Na$$_2$$O) compositions of possible aggregate liquids for the studied suites are consistent with both equilibrium and fractional melting, with P$$_o$$=20-27 kb, P$$_f$$ =5-18 kb and F$$_{max}$$=12-24%. For the Costa Rica Rift and Gorda Ridge samples currently available models cannot differentiate between these alternative origins as aggregates of equilibrium or fractional melts, or by fractionation from an isobaric batch melt.
The results of heating stage experiments indicate that if the magmatic liquids were fluid-saturated and the kinetics of melting during reheating are taken into consideration, then melt inclusions in plagioclase and olivine can be successfully homogenised and yield trapping, or crystallisation temperatures. However, these homogenisation temperatures are up to 50°C lower than trapping temperatures inferred from published vertical furnace experiments (e.g., Nielsen et al., 1995). The melt inclusions in vertical furnace experiments are here interpreted to have been overheated (and possibly poorly quenched), and consequently, their compositions are not representative of the liquids trapped to form the melt inclusions. As such, they cannot be used to infer aspects of the petrogenesis of a given suite of samples.
A comparison of the compositions of melt inclusions in cotectic (plagioclase, olivine and spinel) phenocrysts, and in phenocrysts interpreted to have crystallised from liquids similar to the pillow-rim glasses, has identified compositional variations, particularly for TiO$$_2$$ and FeO* in plagioclase-hosted inclusions, that are not related to trapped liquid compositions. These compositional variations, found in all phenocryst phases, are thought to be caused by post-trapping re-equilibration. However, the melt inclusions do partially recover trapped liquid compositions and can, with care, be used to interpret the magmatic history of a sample, or suite of samples.

Item Type: Thesis - PhD McNeill, AW Basalt, Magmatism, Submarine geology Copyright 1997 the author - The University is continuing to endeavour to trace the copyright owner(s) and in the meantime this item has been reproduced here in good faith. We would be pleased to hear from the copyright owner(s). Thesis (PhD)--University of Tasmania, 1998. Includes bibliographical references View statistics for this item