Rare-metal pegmatites begin with the source

01

Why this matters

Lithium-caesium-tantalum (LCT) pegmatites are commonly treated as the final, highly fractionated products of granitic magmas or low-degree partial melts of fertile protoliths. These models explain how incompatible elements can become concentrated during melt evolution, but not why major rare-metal pegmatite provinces form in particular regions or tectonic intervals. A highly evolved melt can only concentrate the metals available to it; source fertility and the processes that mobilise that metal budget must therefore be considered before fractionation begins.

Tang et al. (2023) address that larger-scale problem using Triassic pegmatites and associated granites across East Asia. Their question is not simply whether a pegmatite fractionated from a granite, but how the crust first became enriched in metals, how that material was transferred to depth, and what tectonic process subsequently generated fertile magmas. Their model links chemical weathering, sediment accumulation, crustal recycling, anatexis and Late Triassic extension during the assembly of Pangea.

02

What the study found

The authors combine new zircon U-Pb ages, zircon trace-element and Hf-isotope data, apatite Nd isotopes, and whole-rock geochemistry from Bailongshan in western Kunlun with published datasets from the Altai, Tianshan, Qaidam and Songpan-Ganze belts. Most major rare-metal pegmatites considered formed between approximately 220 and 200 Ma. At Bailongshan, S-type granites and barren and mineralised pegmatites overlap at about 212–208 Ma, whereas nearby I-type granitoids are older, at approximately 218–216 Ma.

The isotopic evidence links the pegmatites to the S-type granites. S-type granite, barren pegmatite and mineralised pegmatite have comparable enriched Sr-Nd-Hf isotope compositions and are distinct from the I-type granitoids. Their inherited-zircon populations resemble those of the regional Triassic sedimentary succession. Across East Asia, Late Triassic S-type granites and barren pegmatites contain roughly an order of magnitude more Li and Cs than older granitoids, indicating that rare metals were not progressively accumulated through repeated crustal melting alone.

Tang et al. instead identify sedimentary rocks as the fertile reservoir. Lithium released by continental weathering can be retained in clay minerals, while mica may further concentrate Li within sediments. In western Kunlun, granulite-facies metasedimentary rocks contain only 4.8 ppm Li and 0.8 ppm Cs—mean decreases of approximately 95% and 91% relative to sedimentary precursors. The authors attribute these losses to mobilisation during muscovite dehydration melting. Their mass balance indicates that 1 km³ of sedimentary rock could supply approximately 243,000 t Li and 23,000 t Cs during granulite-facies metamorphism.

03

Our interpretation

The paper’s most useful contribution is its separation of source preparation from source activation. Long-term weathering and sedimentation create a potentially fertile reservoir; subduction and collision incorporate it into the middle to lower crust; and later melting releases its Li-Cs budget into S-type magmas. Other metal-enrichment processes, including fractional crystallisation and low-degree partial melting, remain important, but they act on a crustal source whose fertility was established much earlier.

The proposed trigger is Late Triassic extension associated with slab retreat around Pangea’s eastern margin. Closure of the Paleo-Asian, Proto-Tethys and Paleo-Tethys oceans created arc and back-arc basins, recycled sediments into the deep crust and assembled East Asia by about 220 Ma. Subsequent slab retreat promoted intracontinental extension and melting. This provides a coherent explanation for the broadly synchronous 220–200 Ma rare-metal pegmatite event across several belts.

The framework is persuasive but incomplete at deposit scale. Temporal coincidence with Pangea assembly does not independently prove that slab retreat was the decisive trigger, and the study resolves source fertility more convincingly than the transition from barren to spodumene-bearing pegmatite. Melt fraction, fractionation, volatile behaviour, immiscibility, structural focusing, emplacement level and post-magmatic modification may all determine whether a fertile magma produces an economic deposit. Individual terranes therefore require local geochemical and geological studies.