Tungsten (W), a hard, heat-resistant, and corrosion-resistant metal, is indispensable to modern high-tech industries—from aerospace and defense to computing. While its global distribution is uneven, most tungsten deposits share defining geological traits: close ties to highly evolved, volatile-rich granites; formation from melted sedimentary rocks (anatexis) in tungsten-rich granitoids; and occurrence in back-arc or intraplate zones rather than convergent tectonic margins. These features long supported theories of a purely crustal origin for tungsten mineralization.

However, a new study challenges this view, highlighting significant mantle contributions to ore-forming fluids. Published in Communications Earth & Environment, the research—led by Prof. YANG Jiehua from the Institute of Geochemistry of the Chinese Academy of Sciences (IGCAS), with guidance from CAS member Prof. HU Ruizhong, and Prof. ZHOU Meifu—reveals that mantle activity plays a critical, previously underappreciated role in tungsten metallogeny.

The team addressed a key mystery: why helium (He) and argon (Ar) isotopes in global tungsten deposits indicate mantle involvement in ore-forming fluids, even as associated granites show crustal signatures. This "isotopic decoupling" had never been systematically explored, leaving the mantle's role in tungsten formation unclear.

To investigate, the researchers analyzed mercury (Hg), He, and Ar isotopes in ore minerals and granitoids from representative South China tungsten deposits. They also compiled global datasets on Hg, He, Ar, strontium (Sr), and neodymium (Nd) isotopes from major tungsten provinces, combining statistical analysis, geochemical modeling, and machine learning to decode mantle contributions to global tungsten cycles.

Global He-Ar isotope data show significant mantle input in ore-forming fluids: ~10% in South China's tungsten deposits and over 40% elsewhere. In South China, positive Δ¹⁹⁹Hg anomalies in tungsten-rich granitoids and long-term Sr-Nd isotope trends in mafic rocks suggest that Paleo-Pacific Plate slab roll-back altered the region's ancient lithospheric mantle. This triggered intense crust-mantle interaction—critical for extensive tungsten mineralization.

Machine learning identified highly evolved geochemical traits as hallmarks of tungsten-rich (rather than barren) granitoids. Integrating whole-rock chemistry, isotopic data, and global deposit distributions, the team proposed a generalized model: extension linked to oceanic subduction is more conducive to large-scale tungsten mineralization than processes like continental collision or intracontinental rifting. In this framework, the mantle provides heat to drive slab "devolatilization," releasing He, Ar, Hg, and fluorine (F) into the crust—fuelling magma differentiation and tungsten mineralization.

Compiled Nd model ages (T_DM2) from global tungsten-rich granitoids and scheelites peak at 1.8–1.2 billion years, coinciding with the assembly and breakup of the Nuna supercontinent. As a moderately "siderophile" (metal-loving) element, tungsten from Earth's core likely reached the crust via mantle plumes during this period—supported by tungsten concentrations and ruthenium-tungsten isotopes in ocean island basalt (OIB)-like rocks. Notably, South China's Cathaysia Block, once part of Nuna's interior, saw weathering of ancient tungsten-rich rocks create a pre-enriched crustal basement, priming it for Mesozoic tungsten deposits.

Comparative analysis of Nd isotopes in global tungsten provinces revealed that accretionary orogens (formed by tectonic plate "stitching") host larger deposits than collisional orogens, despite the latter having more tungsten-rich crust. South China's exceptional tungsten endowment stems from a unique combination: a high proportion of pre-enriched crust and intense crust-mantle interaction driven by post-subduction extension.

The study identifies granitoids formed 1.8–1.2 billion years ago (T_DM2) in post-subduction extensional zones as high-potential targets. Researchers propose a data-driven exploration strategy combining Nd isotope mapping, geochemical anomalies (tungsten, tin, Zr/Hf, Nb/Ta ratios), and machine learning. Promising regions include the Circum-Pacific (Russian Far East, Alaska, northwestern Canada), Central Asian Orogenic Belt, and eastern Tethyan Belt.

This work was financially supported by the National Natural Science Foundation of China, and the National Key Research and Development Plan, among other sources.

Generalized Tungsten Mineralization Dynamics Model. (Image by Prof. YANG Jiehua's group)