LAUSR

Abstract The source(s) of lode gold deposits formed in Precambrian cratons related to accretion/collision and cratonic reactivation, formerly attributed to either supercrustal volcano-sedimentary sequences or deep crustal/sub-crustal origin, remain controversial largely because gold deposits are spatially related to metamorphosed rocks, but geochemical data somewhat indicate a poorly understood deep source. Reconciling such conflicts is important to better understand the main factor controlling the formation of ore deposits and their genetic link with specific tectonic settings. Giant Late Mesozoic lode gold provinces in the North China Craton (NCC) were formed ca. 1.7 Ga later than cratonization metamorphism, and contemporaneous with intensively felsic to mafic magmatism related to cratonic reactivation. In this study, we conduct a comprehensive Pb isotope study on major gold deposits from the eastern Yanshan belt, northern margin of the NCC. In order to constrain the source(s) of gold, we attempt to map regional Pb isotope variations of lower continental crust (LCC), and develop a two stage quantitative model (punctuated by three prominent geological events at 2.80 Ga, 1.85 Ga and 0.16 Ga) to reproduce time-integrated Pb isotopic signatures of deep-seated lithospheric reservoirs to make a comparison with Pb isotopic signatures of the gold mineralization. Gold-bearing pyrites within different types of host rocks have relatively uniform Pb isotopic ratios, which are significantly different to high-grade metamorphosed host rocks, but similar to those of spatially associated Late Mesozoic granitic rocks. Our data show that the Pb isotopic signatures of gold deposits vary consistently with presumed regional Pb isotopes of the LCC during the Late Mesozoic. Lead isotopic heterogeneity of the LCC was likely caused by underplating of mafic magmas derived from mantle sources. During underplating, highly chalcophile elements (e.g. Au, Ag and Cu) were concentrated at the base of the LCC due to sulfide saturation from mafic magmas. Integrating petrological, geochemical, geochronological, and considering chalcophile element solubility, we propose a new genetic model to describe the formation of Late Mesozoic gold deposits in the NCC: (1) early formation of sulfide-bearing cumulates with high Au/Cu ratios during magma differentiation at the base of the LCC; (2) subsequent fluid-fluxed remelting of these cumulates at the onset of lithospheric extension and release and ascent of ore-forming fluids to the site of precipitation. Considering the metallogenic characteristics of other Au ore-forming systems, we suggest that metal fertilization in deep-seated reservoirs and subsequent tectonic decompression are important factors controlling the development of Au-rich ore deposits worldwide. This study demonstrates how Pb isotope can be employed to trace source(s) of gold deposits.