An Evidence-Constrained Four-Domain Exploration Framework for Rare Earth Elements in The Bangka Belitung Tin-Granite System

Abstract

Rare earth element (REE) enrichment in the Bangka Belitung tin province occurs in genetically and technologically distinct materials that require different exploration and processing tests. This study develops an evidence-constrained framework that integrates fresh granite, weathered granite, placer/alluvium, and tin-processing residues without merging non-comparable geochemical values. The analysis separates primary accessory-mineral-hosted, residual mineral-grain, ion-exchangeable clay-associated, mechanically concentrated, and anthropogenically concentrated REEs. Site-specific datasets were evaluated by sample medium, analytical basis, mineral host, spatial support, process diagnosis, recovery evidence, and U-Th/NORM risk. Weathered profiles show systematic enrichment toward lower horizons: reported total REE concentrations increase from 156-188 to 196-268 ppm at Toboali and from 230-330 to 342-450 ppm at Air Gegas. Mineralogical studies identify monazite as the principal Ce-rich light-REE host and xenotime as a Y-heavy-REE host across primary, placer, concentrate, and tailings materials. Diagnostic leaching from South Bangka demonstrates that part of the REE inventory is extractable, but strong-acid recoveries cannot be generalized as evidence of low-intensity ion-exchange recovery. The ordinal screening assigns weathered granite the highest geological prospectivity but only moderate evidence confidence; tailings and placers remain attractive near-surface domains with greater heterogeneity and radiological uncertainty. Remote sensing is retained as a proxy for regolith, drainage, and disturbed ground, not as direct REE detection. The framework provides a reproducible basis for prioritizing field verification while preventing conceptual prospectivity from being misreported as a resource or economic valuation.