Resumen de A coupled petrological and petrophysical study of high pressure dehydration reactions in subduction settings: insights from the Betic cordillera and the Kohistan paleo-arc
Our understanding of subduction zones relies on our ability to integrate observations at diverse spatial and temporal scales in these settings. Volatiles are perhaps one of the most distinctive features of subduction zones, which are, in turn, the flagships of Earth's tectonics. A key process in subduction zones is the cycling of volatiles from the hydrated incoming slab and their recycling back into the Earth's surface through arc volcanism. The cycling of volatiles in these settings depends crucially upon metamorphic reactions involving volatile-bearing phases. Our ability to understand the transfer of volatiles in subduction settings therefore relies on our understanding of devolatilization reactions in these settings. Unfortunately, the paucity in the geological record of fossilized examples of arrested devolatilization reactions hampers a better understanding of devolatilization processes deep in the subduction factory.
In this Ph. D. thesis I have selected two specific dehydration reactions that are fundamental in the volatile cycling of subduction zones: the breakdown of amphibole in the lower arc crust, and the breakdown of antigorite deeper in the mantle wedge-slab. The originality of this work relies on the study of two unique outcrops worldwide where these breakdown reactions occur in arrested, dehydration fronts: (i) the dehydration-melting of amphibole as recorded in the Jijal Complex (Kohistan paleo-arc Complex, NW Pakistan); and (ii) the dehydration of antigorite as recorded in the Cerro del Almirez ultramafic massif (Nevado-Filábride Complex, Betic Cordilleras, SE Spain). The study of such fronts provides a unique opportunity to investigate in detail how these reactions operate in nature and what are the allied petrological, physical and chemical modifications induced by these reactions. Field mapping and sampling of reactants and products across the arrested reaction fronts make it possible to compare textural and chemical changes during the progression of the reactions and to unravel the mechanism of devolatilization. The methodology followed in the present research consists of detailed field mapping (in the case of the Cerro del Almirez ultramafic massif), combined with the application of basic petrological tools and state-of-the-art instrumental techniques (Transmission Electron Microscopy, TEM; Electron Backscattered Diffraction, EBSD, High-Pressure experiments, etc.). The study of mineral phase relations, as well as phase diagrams derived from computational thermodynamics, has been also an important part of this Ph.D. thesis. This methodology was used to estimate the pressure and temperature conditions at which both reactions took place and to constrain the possible mechanisms controlling mineral growth and textural development.
The first case study is the high-pressure antigorite dehydration reaction, based on the detailed mapping of the Cerro del Almirez ultramafic massif (Nevado-Filábride Complex, Betic Cordillera, SE Spain). In this massif, the antigorite breakdown took place during the Alpine collision when serpentinite (antigorite + magnetite + olivine + tremolite/diopside) was subducted to depths of up to 70 km, and transformed into Chl-harzburgite (olivine + orthopyroxene + magnetite + tremolite). TEM observations of the antigorite microstructure show that, prior to the transformation, antigorite is exceptionally ordered and consists of the polysome m = 17 with no polysomatic defects. Close to the dehydration front, however, limited disorder features occur, mainly as (001) twins, reaction rims and reduction of m down to 14-15. Moreover, the observed, locally incomplete, breakdown of titanclinohumite in olivine-rich veins supports that serpentinite was annealed at high pressure conditions and very close to its maximum thermal stability. Furthermore, the peak pressure and temperature have been experimentally constrained at 680-710ºC and 1.6-1.9 GPa. This experimental work also put important constrains on the phase relation of silica-rich serpentinite, demonstrating that in such lithology dehydration starts at lower temperatures than the terminal antigorite reaction. Detailed mapping of the Atg-serpentinite to Chl-harzburgite arrested, dehydration front in Cerro del Almirez reveals the presence of a layer of a transitional lithology made up of antigorite-chlorite-orthopyroxene-olivine. This transitional lithology, which delineates the devolatilization front, shows a sequence of mineral assemblages that, together with their mineral composition variation, demonstrates an increase of the metamorphic grade from the Atg-serpentinite to the Chl-harzburgite. An important finding of this Ph.D. thesis is the discovery of metric-to-decametric alternations of contrasting texture of the prograde Chl-harzburgite assemblage and the preservation of serpentinite lenses downstream the devolatilization front. Three textural varieties of Chl-harzburgite have been recognized: granofels, spinifex-like, and recrystallized textures. The former textural variety is crystallized near the atg-breakdown equilibrium, whereas the spinifex texture grew under faster growth rates at a high affinity of the atg-breakdown equilibrium. Recrystallization of these two textures is interpreted, on the basis of field and detailed microstructural observations, as due to brittle deformation, most probably due to hydrofracturing at high pressure conditions. The spatial alternation of granofelsic and spinifex-like textures is interpreted as reflecting the cyclic evolution of the fluid pore pressure excess during the development and advancement of the atg-dehydration front. It witnesses for different fluid expulsion mechanisms operating in subduction environments.
The second case study is the arrested transformation of hornblende gabbronorite to opx-free, garnet granulite, involving the coeval breakdown of amphibole and orthopyroxene, and the formation of garnet and quartz in the lower crust of a Paleo-Island arc (Jijal Complex, Kohistan paleo-arc Complex, NW. Pakistan). Close to the reaction front, clinopyroxene from the granulite displays a strong Ca-tschermark zoning with lower Al-contents in the rims. REE zoning of clinopyroxene and pseudosection diagrams indicate that only clinopyroxene rims reflect chemical equilibrium with garnet in the reaction front metamorphic conditions (P = 1.1 ± 0.1 GPa, T = 800 ± 50ºC). Clinopyroxene cores retained high-Al contents inherited from precursor gabbronorite clinopyroxene and remained in chemical disequilibrium. Further beyond the reaction front, the garnet granulite mineral assemblage is, nevertheless, completely re-equilibrated. Pseudosections calculations indicate that hornblende gabbronorite assemblages are highly metastable at lower arc crust depths. The transformation to garnet granulite was therefore substantially overstepped in terms of pressure and temperature. A key observation is the preservation of the original magmatic foliation in the garnet granulite. This suggests that the gabbronorite-granulite transformation took place at static conditions. In fact, product garnet displays a strong lattice preferred orientation, attesting for a topotactic replacement of amphibole and orthopyroxene from the igneous protholith. Thus, in the absence of deformation, the orientation of mafic precursor phases conditioned the nucleation site and the crystallographic orientation of garnet. Topotaxial transformation reactions and homoepitaxial growth of garnet occurred during the formation of high-pressure, mafic garnet-granulite after low-pressure mafic protoliths. This result shows that metastable low-pressure lithologies can be buried down the root of the island arc crust, where amphibole dehydration melting reactions triggers their transformation into dense mafic garnet granulites. This process attests for an intra-crustal recycling of volatiles and provides a mechanism for the development of dense roots in island arc. Delamination of such dense roots has important petrological and geochemical consequences for the evolution of mature island arcs.
