First Results of Tracing Recycled Trench Sediment with 10Be in the Central Trans-Mexican Volcanic Belt

Studies of the cosmogenic 10Be isotope in arc magmas resulted in the general acceptance of sediment recycling in subduction zones (Morris et al. 1990, Nature). Subsequently, however, the use of the 10Be tracer became rare due to the analytical challenges to measure minute amounts of 10Be in volcanic rocks. Recently, the method has been revived in the LDEO Surface Exposure Dating Laboratory at Lamont/USA, where 10Be in arc volcanic rocks can now be analyzed with a detection limit of a few 10,000 atoms/g. We obtained new 10Be data from high-Mg# (mostly >60) arc magmas from the central Trans-Mexican Volcanic Belt (TMVB) which was previously considered as barren. All magmas have a late Pleistocene age and erupt within 100 km along-strike the arc front, which should largely neutralize 10Be variations due to differing trench composition and transit time. Previous studies (major and trace elements, Sr-Nd-Pb-Hf-O-He isotope ratios) denote three different compositional series. Next to the dominant calc-alkaline arc magmas with the arc-typical high ratios large-ion-lithophile (LILE) to high-field-strength (HFSE) elements, there are subgroups of high-K magmas that are strongly enriched in light rare earth elements and have garnet signatures, and OIB-type arc magmas with LILE/HFSE ratios resembling those of intraplate magmas. Fourteen out of 20 TMVB volcanic rocks contain 10Be abundances above the detection limit. Consistent with the long arc-trench transit time of 5-6 Myr, the average TMVB 10Be is low (11470 k atoms/gram) and lower than 10Be in the Andes Southern Volcanic Zone (1,778604 k atoms/gram; Cooper et al., 2012, AGU) and the N Honshu arc (635611 atoms/gram, Shimaoka et al. 2016, Chemical Geology) where the transit times are shorter. Remarkably, 10Be is present at similar levels in all three TMVB compositional groups, without any apparent correlation with the existing element and isotope systematics. Thus, while the new data are generally consistent with a model of sediment recycling in arcs, there is no correlation with other sediment tracers. One possible explanation is that recycled 10Be is released from slab in a separate agent (fluid?) that is decoupled from the principal agents that transfer sediment components (e.g. slab melts, diapirs) from slab to wedge.