Recovery of Be-bearing phosphates as by product from rare-metal granite deposit: a multi-scale approach

This study demonstrates the feasibility of an unconventional processing route based on direct anionic flotation under highly acidic conditions (pH 2) for the selective recovery of Berrylium bearing phosphate minerals from a low-grade rare-metal granite containing approximately 1% by weight P 2 O 5 . Silica gangue interference and clay-induced “slime-coating” were successfully mitigated through a targeted 20 µm desliming stage prior to flotation. Among the evaluated collectors, the ethoxylated formulation AERO 855 exhibited the best metallurgical performances in bench-scale batch trials. At an optimum dosage of 1500 g/t, a concentrate grading approximately 10% by weight P 2 O 5 was produced at 61% P 2 O 5 recovery, corresponding to a tenfold enrichment ratio relative to the feed. Single-mineral pure minerals flotation identified a remarkable intrinsic selectivity window at 100 mg/L, where fluorapatite recovery reached 84% while quartz and albite co-recoveries remained below 10%. Adsorption isotherms determined by Total Organic Carbon (TOC) revealed a substantially higher adsorption density of AERO 855 on phosphate minerals surface than on silicate minerals. A specific adsorption density of 6.8 mg/g was measured at optimum conditions and a saturation plateau was observed at approximately of 70 mg/g suggesting a closely packed surfactant saturated adsorption layer. Comparison between pure minerals flotation and bench-scale flotation results revealed that the fivefold increase in collector demand may be driven by ore complex mineralogy and textural complexity. High-resolution automated mineralogy (SEM-MLA) further identified the ultimate structural bottleneck for the upgrading: fewer than 30% of the fluorapatite and 20% of the beryllium-bearing phosphates exhibit a liberation degree greater than 70%, with beryllonite occurring predominantly within the secondary phosphate aggregates. These findings prove that while surface chemistry is highly efficient, maximizing concentrate grade requires a targeted rougher concentrate regrinding strategy to bypass textural limitations.