Residence and partitioning of REE and selected trace elements in amphibolite-facies metabasites : an example from the St. Anthony complex, northern Newfoundland

Mulrooney, Daniel Joseph (2004) Residence and partitioning of REE and selected trace elements in amphibolite-facies metabasites : an example from the St. Anthony complex, northern Newfoundland. Masters thesis, Memorial University of Newfoundland.

[English] PDF (Migrated (PDF/A Conversion) from original format: (application/pdf)) - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (53MB) [English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. (Original Version)

Abstract

Metabasites from the metamorphic sole of the Hare Bay Allochthon, northern Newfoundland, exhibit a narrow metamorphic field gradient from greenschist- through epidote-amphibolite to amphibolite-facies. The metamorphic facies are defined from the distribution of the following mineral assemblages (all with quartz and titanite): chlorite-epidote-plagioclase-Ca amphibole (greenschist), Ca amphibole-plagioclase-epidote (epidote amphibolite), and Ca amphibole-plagioclase (plagioclase amphibolite). The assemblage Ca amphibole-plagioclase-garnet-biotite-apatite (biotite amphibolite) equilibrated in a K₂O-enriched shear-zone. Representative samples from each metamorphic facies were were analyzed for rare earth elements (REE) and a suite of trace elements (Sc, V, Cr, Ni, Cu, Zn, Sr, Rb, Y, Zr, Nb) in bulk rocks and the major minerals Ca amphibole, plagioclase, epidote, titanite and garnet, by solution ICP-MS and laser-ablation microprobe (LAM)-ICP-MS respectively. Major-element concentrations of bulk rocks and minerals were determined by XRF and electron probe microanalysis (EPMA) respectively. -- On the basis of bulk-rock, major- and trace-element data and chondrite-normalized REE data, two suites of metabasites were distinguished: (i) gamet-absent greenschist and amphibolite characterized by relatively low SiO₂ and A1₂O₃ and Y+MREE-enrichment; and (ii) garnet-bearing amphibolite with relatively low FeOtotal + MgO, high P₂O₅ and Zr+LREE enrichment. The bulk P₂O₅ and LREE enrichment in garnet-bearing rocks is inferred to be sequestered in unanalysed fine-grained apatite. -- Ca amphibole, plagioclase, epidote, garnet and titanite exhibit systematic distributions of major- and most trace-elements, including REE in epidote amphibolite, plagioclase amphibolite and biotite amphibolite assemblages, indicating a close approach to equilibrium. However, variations in the distributions of a few trace elements (e.g., Rb and Nb between certain phases) suggest that equilibrium may not have been achieved for all elements at a thin-section scale. -- In garnet-absent amphibolites, LREE are partitioned in the sequence titanite>epidote>>Ca amphibole≥plagioclase, whereas HREE are partitioned in the sequence titanite>>CA amphibole>epidote>>plagioclase. For biotite amphibolites, LREE and HREE are partitioned in the sequence titanite>>Ca amphibole ≈ plagioclase>garnet and titanite≥garnet>>Ca amphibole>>plagioclase respectively. -- Controls on trace-element incorporation in individual phases, and hence on partitioning behaviour, were qualitatively evaluated in this study. Crystal structure exerts a major control on trace-element partitioning so that the distribution of elements that occupy non-analogous sites in the two minerals (e.g., Sr in the A site in plagioclase and the X site in garnet) is characterized by a quasi-'parabolic' curve in Onuma-type diagrams in which distribution coefficients are plotted against ionic radius. Such diagrams are useful for the qualitative extraction of site parameters such as site radius and elasticity in the two minerals. In the case of mineral pairs that partition a trace element on analogous sites in the two minerals (e.g., Ni in MVI sites in Ca amphibole-epidote pairs), information on site size and elasticity is masked by interference between the similar competing sites and the qualitative extraction of site parameters is more problematic. In addition to crystallographic controls, several major-element compositional controls on trace-element incorporation in Ca amphibole, plagioclase, epidote and garnet were recognized and evaluated. For example, REE and Y partitioning into Ca amphibole, garnet and plagioclase are influenced by the major-element occupancy of the M4, X and A sites of these three phases respectively; and abundances of several divalent first row transition metals (Zn, Cu, Ni, Co) are positively correlated with the content of octahedral A1 in M3 sites in epidote. -- Integration of mineral modes with trace-element mineral concentration data was carried out to determine mass balances for garnet-absent samples of similar bulk composition and the results were cross-checked against bulk-rock trace-element data. The results yield very good matches between the measured and reconstructed trace-element abundances in all cases, indicating that all the hosts for these elements had been analysed. They also illustrate the change in the relative importance of the four minerals, Ca amphibole, plagioclase, epidote and titanite, as carriers of the REE and analysed trace elements across the epidote-out isograd. In particular, they show that in most epidote-amphibolite, the trace-element enriched phases titanite and epidote, especially the former, are the dominant carriers of Y, Nb and all REE, accounting for approximately 90% and 65% of the LREE and HREE budgets respectively despite their low modal abundances, and that the ΣREE carried in plagioclase is negligible. However, where modal titanite and epidote are relatively less abundant (or where epidote is absent in plagioclase amphibolite), the relative capacities of these phases as trace-element or REE reservoirs are significantly diminished, Ca amphibole becomes the primary REE reservoir and ΣREE in plagioclase is increased. -- Major-element mineral compositions from representative epidote amphibolite and plagioclase amphibolite that straddle the epidote-out isograd were examined using the algebraic technique of singular value decomposition to quantitatively model the epidote-out reaction isograd. The following balanced reaction provides a realistic representation of the isograd, textures, and modal variations observed in natural samples throughout the study area: 1.00Ca amp₄₆ + 1.70PI₄₆ + 0.34Ep₄₆ + 1.77Ttn₄₆ = 1.07Ca amp₄₉ + 1.92Pl₄₉ + 1.80Ttn₄₉ where the superscripts 46 and 49 refer to compositions of minerals in samples below and above the epidote-out isograd respectively. -- The results of evaluating the redistribution of selected trace elements and REE across the epidote-out isograd using the model reaction are compatible with the uptake of LREE by Ca amphibole, and to a lesser extent plagioclase, due to an reduction in LREE abundance in titanite and the breakdown of epidote. These linked changes indicate that there are both crystal-chemical and mineral assemblage controls on REE and trace-element partitioning in these rocks. This study illustrates the significance of the minerals epidote and titanite as the principal carriers of REE in epidote-amphibolite-facies metabasites and the quantitative estimation of partition coefficients shows that Ca amphibole acts as an important carrier for REE above the epidote-out isograd.

Actions (login required)

View Item