From granite waste to strategic value: unlocking critical raw materials through magnetic separation process innovation

This study explores the potential recovery of critical raw materials (CRMs) from extractive waste generated during the exploitation of ornamental granite in the Montorfano and Baveno plutons (Verbania, Piedmont, Italy). Current industrial flowsheets are primarily designed to maximise the purity of quartz and feldspar concentrates for ceramic and glass applications. Consequently, separation strategies favour the rejection of middlings, consisting of quartz-feldspar particles partially intergrown with Fe-bearing minerals that would compromise final product specifications, but show enrichments in CRMs. These middlings still contain significant amounts of recoverable non-magnetic silicate minerals and could be valorised through targeted process optimisation. Granite offcuts are reprocessed to extract non-magnetic fractions rich in feldspar and quartz, classified by grain size and chemical purity, while the magnetic fraction - dominated by micas - is typically sold as construction filler. However, this magnetic by-product hosts REE-bearing accessory phases and moderate concentrations of strategic elements - REE, Y, Sc, Li, Ti, Mg, and Mn - representing an underexploited CRM source. Feldspar itself is classified as a CRM under the EU Critical Raw Materials Act (CRMA), further reinforcing the strategic relevance of improving recovery efficiency from granite waste streams. Moreover, the mica-bearing magnetic fraction offers opportunities for market diversification: mica can be further valorised for high-value applications, including cosmetic formulations that exploit its lamellar structure and optical properties. The research addresses mineralogical and processing constraints limiting selective recovery, including low-grade and heterogeneous assemblages, complex mineral intergrowths, and incomplete liberation. A key challenge lies in balancing comminution to achieve sufficient liberation prior to magnetic separation while minimising the generation of ultrafine particles, which  -  despite facilitating physical separation  - negatively affect downstream processing. Additional complexity arises from overlapping magnetic susceptibilities between target phases and Fe-bearing gangue silicates, frequently resulting in mixed products with suboptimal enrichment. Beyond REE minerals such as monazite, allanite, and xenotime - commonly locked within biotite - several elements of interest are plausibly incorporated within the crystal lattice of micas, underscoring the importance of prioritising mica concentration as a pre-treatment step, followed by targeted chemical extraction routes. Process optimisation through refined magnetic separation could simultaneously increase the recovery of non-magnetic material for further processing and alternative quartz and feldspar applications, and improve the grade of the magnetic concentrate, enabling more efficient CRMs recovery. Although CRM concentrations remain modest compared to primary deposits, valorising these materials offers significant advantages, including reduced waste volumes, alignment with EU CRMA objectives, and leveraging existing infrastructure to minimise costs and permitting requirements. From a circular economy perspective, this approach supports near-zero-waste operations while mitigating the environmental impacts associated with primary mining. The study focused on comprehensive mineralogical and geochemical characterisation of feed granites and magnetic separation products using ICP-MS, XRF, XRPD, and SEM-EDS, combined with laboratory-scale processing trials. Both dry permanent magnet and electromagnetic separation techniques were evaluated across different granulometric classes to identify optimal operational parameters. The ultimate goal was to define a scalable processing protocol suitable for pilot-scale validation and integration into the current industrial flowsheet.