Evolution of spreading rate and H2 production by serpentinization at mid-ocean ridges from 200 Ma to Present

Mid-oceanic ridge (MOR) spreading rate spatially varies today from 20 to 200 mm/yr and geological records attest of important temporal variations, at least during the past 200 My. The spreading rate has a direct impact on the mechanisms accomodating extension (magmatic vs tectonic), hence on the nature of the rocks forming the oceanic lithosphere. The latter is composed of variable amount of magmatic and mantle rocks, that dominate at fast and (ultra-) slow spreading ridges, respectively. Serpentinization of mantle rocks contributes to global fluxes and notably to those of hydrogen and carbon by providing a pathways for dihydrogen (H2) production, carbone storage by mineralization, and carbon reduction to CH4 and possibly complex organic compounds. Quantification of the global chemical impact of serpentinization through geological time requires a coupling of geochemical parameters with plate-tectonic reconstructions. Here we quantify serpentinization extent and concurrent H2 production at MOR from the Jurassic (200 Ma) to present day (0 Ma). We coupled mean values of relevant petro-chemical parameters such as the proportion of mantle rocks, initial iron in olivine, iron redox state in serpentinites, % of serpentinization to high-resolution models of plate motion within the GPlates infrastructure to estimate the lengths in 1 Myr intervals for the global MOR plate boundary (spreading and transform components), and spreading ridges as a function of their rate. The model sensitivity to selected parameters has been tested. The results show that fragmentation of Pangea resulted in elevated H2 rates (>1012 to 1013 mol/yr) starting at 160 Ma compared to Late Mesozoic (<160 Ma) rates (<1011-1012 mol/yr). From 160 Ma to present, the coupled opening of the Atlantic and Indian oceans as well as the variation in spreading rates maintained H2 generation in the 1012 mol/yr level, but with significant excursions mainly related to the length of ultra-slow spreading segments. For the first time, this model offers a framework toward flux modeling at MOR through time. The model can be further implemented by adding supplementary geochemical parameters or serve other geochemical issues.