From melt to pluton: Magmatic emplacement, differentiation and duration of the Beauvoir rare-metal granite
Abstract:
Constraining the construction and evolution of granitic intrusions together with the kinematics associated with their magmatic differentiation remain a major objective in igneous petrology. To progress on these issues, this PhD work is based on the 900 m long borehole crosscutting the Beauvoir granite (recovery: ~99%), a small and shallow rare-metal granite intrusion from the French Massif Central (France), allowing a high-resolution sampling of a fully recovered plutonic body.
Based on rock descriptions and high-resolution major and minor element compositions of the mineral phases, mainly lepidolite (in-situ measurements and chemical maps), it appears that the Beauvoir intrusion was formed through the stacking of at least eighteen deca- to hectometric compositionally different crystal-poor sills. Furthermore, the systematic compositional changes within these sills, together with the detailed study of sill boundaries, provide a dynamic record of the pluton assembly: although globally constructed by over-accretion (i.e., from bottom to top), sills can also be locally under-accreted. The systematic FeO/MnO decrease observed in both lepidolite (Li-mica) and columbo-tantalite (Nb-Ta oxide) suggests that these sill parental melts were progressively extracted during a precursory differentiation process. The absence of magma mixing between these different sills indicates that the host reservoir crystal fraction was >50% of fully solidified before the intrusion of a new sill. Nevertheless, the ubiquitous presence of dissolution textures as well as oscillatory and reverse zoning patterns in minerals strongly suggests that each sill progressively inflated by successive and rapid magma replenishments.
Once intruded, early-crystallised quartz and topaz were progressively fractionated to form cumulates at the base of the corresponding sills, while residual melts are progressively enriched in rare-metals (e.g., Li and Be) and in volatiles (e.g., H2O and F). Textural and mineralogical evidences suggest an efficient extraction of these residual melts from the quartz-rich mush, with crystallised products segregated under the form of albite, lepidolite and amblygonite (Li-phosphate) veins and ovoids. During the magmatic-hydrothermal transition, fluids strongly modified the original texture and composition of some primary minerals, which in zircon is recorded by the progressive replacement of their Si4+ and Zr4+ by secondary “non-formula” elements: P5+, Al3+, Ca2+, U4+, Mn2+, Fe2+, F and H2O. Combined with the high α-decay damage that experienced these zircon grains, their amorphous nature severely limits their use to accurately date rare-metal granite emplacement. Emphasis is thus placed on another alternative: apatite, as its high-resolution U-Pb dating yielded a Beauvoir emplacement age of 314.6 ± 1.1 Ma (2σ), placing the Beauvoir magmatic activity during the collapse of the Variscan orogenic belt.
The lifespan of the Beauvoir intrusion has been assessed using numerical thermal simulations, and is estimated to be ~10 kyr. The solidification time for each sill ranges from tens to thousands of years; implying that the magma cooling rate and associated undercooling, in particular that of the margins of each sill were high (e.g., >0.1 °C.yr-1). Such rapid cooling results in a disequilibrium geometry at three-grain junctions involving two grains of plagioclase and one of lepidolite. Indeed, these corresponding dihedral angles have median and standard deviation values slightly lower than would be expected from an impingement texture. This evidence for early and rapid crystal growth under diffusion-limited conditions following a sill injection is supported by the presence of skeletal cores in lepidolite as well as plagioclase hopper-like morphologies. This study unpicks the use of dihedral angle in felsic rocks, and their application in larger/hotter plutons to constrain their magma solidification kinetics seems promising.
Membre du jury :
Directeurs de thèse:
Lydéric France – Maître de conférence HDR – CRPG – Université de Lorraine (France)
Pierre Bouilhol – Maître de conférence – CRPG – Université de Lorraine (France)
Rapporteurs:
Olivier Bachmann – Professeur – Département des Sciences de la Terre, ETH Zürich (Suisse)
Madeleine Humphreys – Professeure – Département des Sciences de la Terre, Durham University (United Kingdom)
Examinateurs:
Emilie Bruand – Chercheuse CNRS – Laboratoir Géo-Océan – Université de Bretagne Occidental (France)
Marieke Van Lichtervelde – Chargée de recherche – IRD – Université de Toulouse (France)
Michel de Saint-Blanquat – Directeur de recherche – CNRS – Université de Toulouse (France)
Membres invités:
Michel Cuney – Professur émérite – Georessources – Université de Lorraine (France)