An investigation of crustal contamination through petrology and geochemisty

Abstract

The contamination of mantle-derived magmas by the continental crust is an important process during petrogenesis of volcanic rocks at active continental margins e.g. The Andes. Investigating the evolution of continental arc magmas is, however, hampered by our limited knowledge of, and poor constraints on, the nature of the underlying crustal basement and the mechanisms of crustal anatexis. This thesis reports results from:

1) a whole rock geochemical and in-situ geochronological investigation of a suite of crustal xenoliths from the Bolvian Altiplano, Central Andes;
2) a whole rock geochemical study of the xenoiths’ host lavas and;
3) detailed in-situ geochemical studies of crustal partial melts (quenched to glasses) trapped within their crustal progenitors from Bolivia, NE China and SE Spain.

Sampled crustal xenoliths from the Bolivan Altiplano provide a rare insight into the nature of the Central Andean continental basement and reveal lithological and geochemical heterogeneity exists at depth with 87 Sr/86 Sr values extending to 0.7368 which is more radiogenic than any Srisotopic signature exhibited by the recent (< 60 Ma) volcanic record. In-situ U-Pb dating of zircon separates reveal predominant age peaks at 1.7-1.9 Ga, 1.0-1.2 Ga and 495-380 Ma which correspond to periods of supercontinent formation and break-up e.g. construction of Rodinia. Lavas erupted from monogenetic centres on the eastern Bolivian Altiplano show petrographic and geochemical evidence for crustal contamination. The geochemical heterogeneity exhibited by the lavas is, however, difficult to reconcile through simple two component crust-magma interaction models (bulk mixing, AFC and EC-AFC). Instead, contamination is inferred to have involved numerous crustal components. The geochemical signatures observed in lavas from monogenetic centres towards the active Andean arc (between ~18-21o S) are distinct (e.g. lower 87 Sr/86 Sr, higher Sr/Y, higher Ba/Nb at higher Zr/Nb) and may indicate a lower degree of crust-magma interaction, an increase in the contribution from slab-derived fluids and thinner crust arc-wards, the latter which has previously been inferred from geophysical studies. In-situ analysis of anatectic melts reveals that Sr-isotopic disequilibrium between a crustal melt and its source can exist on the sub-millimetre scale. This is understood to reflect the melting of aged minerals with different Rb/Sr (and therefore 87 Sr/86 Sr) more quickly than the isotopic composition can diffusively equilibrate between melt and minerals. Results suggest therefore that crustal anatexis can produce melts which are geochemically heterogeneous both spatially and temporally. This highlights the need for detailed microscopic investigations coupled with petrogenetic modelling in order to develop a more robust characterisation and well-constrained quantification of crustal contamination in open magmatic systems.