Opto-magnetic sorting via water-based wetting and magnetic lifting for coarse REE ore pre-concentration
Hongli Su (TU Delft - Civil Engineering & Geosciences)
Alim Yolalmaz (TU Delft - Civil Engineering & Geosciences)
Rajeev Bheemireddy (TU Delft - EMSD EEMCS Project engineers E)
Yongli Wu (TU Delft - Civil Engineering & Geosciences)
Peter Rem (TU Delft - Civil Engineering & Geosciences)
Francesco Di Maio (TU Delft - Civil Engineering & Geosciences)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Mineral grinding often represents a major fraction of total energy costs and coarse pre-concentration can significantly decrease unnecessary processing of barren material. Compressed-air ejection is effective at industrial scale, but suffers from low accuracy at millimeter scale. An opto-magnetic sorting process for coarse pre-concentration of REE-bearing particles before grinding was developed and assessed at labscale. The process combines image-based optical thresholding, water-based wetting of selected particles, magnetite adhesion to wetted surfaces, and magnetic lifting. This process thus couples selective magnetite coating (enabled by localized wetting) and magnetic lifting for particle sorting. The process was run in a reject-oriented mode to facilitate early mass rejection before subsequent comminution. Lab-scale experiments on rauhaugite revealed increasing pre-concentration with decreasing particle size, resulting in a low-grade fraction of 30.4 wt% of the 2–4 mm feed for possible early rejection. The high-grade fraction (57% of the 2–4 mm feed) achieved a TREO concentration of 2.32%, reflecting an enrichment factor of approximately 1.35 compared to the feed (1.71%), consistent with a partial realization of the intrinsic upgrading potential of the ore at this mass yield, as inferred from the TREO distribution of RGB-classified particles. The lab system processed 84 kg/h, corresponding to approximately 1 tonne of feed processed within 12 h. Based on an instantaneous power demand of ∼ 0.8 kW, this corresponds to an energy consumption of ∼ 9.6 kWh/tonne under steady-state conditions. The process also exhibited low water usage (∼5.7 L/tonne feed) and > 99% magnetite recyclability (after 3 runs). Beyond REE beneficiation, the proposed approach shows potential for selective pre-concentration of heterogeneous particulate streams requiring localized actuation.