Regionally variable timing and duration of celadonite formation in the Troodos lavas (Cyprus) from Rb Sr age distributions

Abstract Celadonite (K[Mg2+,Fe2+][Fe3+,Al3+](Si4O10)(OH2)) is a common void-filling mineral in ocean floor and ophiolite lavas that forms during low-temperature, off-axis, hydrothermal alteration. Its occurrence provides important information about the sink of K from seawater into the upper oceanic crust. New age determinations for celadonitic samples from a ~ 20 km section of the lavas in the Troodos ophiolite, along with published ages, show that most celadonite formed within the first ~20 Myrs after crust accretion. However, significant regional differences in the duration of celadonite formation are observed. In comparison to a region of flat paleo-seafloor, in a paleo-seafloor topographic low celadonite started forming later and continued forming for ~20 Myrs longer. The paleo-seafloor low is associated with abundant hydrothermal sediments which are interpreted to have played a key role in controlling the chemical conditions required for celadonite precipitation. It is hypothesised that celadonite formed for longer largely because these hydrothermal sediments acted as a source for labile Fe for a longer time than in the region where the Fe required for celadonite formation was mainly leached from silicate phases. In the region of flat paleo-seafloor, sediment did not begin accumulating for ~20 Myrs after crustal accretion allowing prolonged oxidative alteration of the upper crust. This led to Fe released by silicate phase breakdown being incorporated in stable Fe3+ containing phases, such as Fe-oxyhydroxides, rather than being transported as aqueous Fe2+ to sites where celadonite could form. If this model is correct, the uptake of K from seawater into the oceanic crust may be highly spatially heterogeneous and intimately linked to the subsurface redox conditions and the distribution of hydrothermal sediments.