Serpentinites néoprotérozoïques : une fenêtre sur la lithosphère océanique associée à la dislocation de Rodinia

This Doctoral thesis focus on the end of the Precambrian, the Neoproterozoic, a pivotal period in the Earth history, by the study of ophiolitic units dated between 800 and 700 Ma. This work essentially deals with the petrography, geochemistry and magnetism of serpentinites from these ophiolites. They are associated with Panafrican orogenic belts along the West African and the Amazonian cratons and witnesses the subductions and the oceanic lithosphere associated to the break-up of the supercontinent Rodinia. The study of these neoproterozoic ophiolites allowed us to propose the formation settings for these sections of oceanic lithosphere. Another question behind this work is: Does the study of serpentinization and associated hydrothermalism allows one to provide constraints on the oceanic paleoenvironment? Work on the serpentinites and associated mafic rocks from the Araguaia belt ophiolitic units (757 ± 49 Ma1, Brazil) constitute the first geochemical study realized on these units. The Cr-spinel chemistry coupled with geochemical modeling, concerning essentially the REE concentrations, attest to a highly refractory protolith that endured between 14 and 24% of partial melting. This work also indicate an abyssal affinity (i.e. MOR rather than SSZ) of these serpentinites, which is confirmed by the N-MORB and E-MORB geochemical signature of the associated basalts. It appears that the serpentinization of these units probably took place in two steps, first in oceanic domain and then by interaction with fluids derived from the surrounding sediments after obduction, leading to strong LILE, B and Li enrichments in the serpentinites. Finally, we propose two potential settings concerning the origin of the sections of oceanic lithosphere presently found in the Araguaia belt: (1) an ocean-continent transition or (2) slices of a more mature lithosphere trapped into an accretionary wedge. Fieldwork, petrography and geochemical analysis on serpentinites from the Khzama (762± 2 Ma²) and Aït Ahmane ophiolitic sections (Anti-Atlas, Morocco) firstly allowed to confirm the supposed link between these two ophiolites. Cr-spinels in serpentinites from these units are characterized by very high Cr# and low Mg# testifying of a protolith having endured partial melting greater the 25%, confirming their affinity with the anti-atlasic subduction zone. This is corroborated by very low incompatible element concentrations such as Al, Ti, HREE and HFSE. Thus, the most likely setting for the genesis of these sections of lithosphere is a same intra-oceanic back-arc spreading center at the margin of the Rodinia supercontinent. The comparative study of all studied serpentinites from Neoproterozoic ophiolites in different settings also demonstrates the ability of magnetic measurements in monitoring the Cr-spinel alteration degree, a method which can be extended to other types of rocks. In particular, we show that Cr-magnetite and ferritchromite can be identified using thermomagnetic curves. Finally, the association of magnetic, petrographic and geochemical methods allows us to propose a model concerning the formation of the massive magnetite veins of the North Aït Ahmane unit. We show that the triptych consisting in: altered serpentinites, spinels and centimetric magnetite veins, results from a fossil, black smoker type, abyssal hydrothermal system. In this context, the oxygen isotopic compositions of the magnetite veins provide a 18O value of the deep Neoproterozoic oceans of -1.33 ± 0.98 (VSMOW), a value that is very similar to the current one (between -1 and +1 VSMOW). It also settles the debate on the evolution of the 18O of the oceans during geological time, attesting of its consistency, at least since the Neoproterozoic.