A mineralogical study of Gonçalo lithium-bearing mica-rich pegmatite ore (Portugal) indicated that lepidolite occurs in coarse-grained textures, which allows an appreciable liberation of gangue minerals (quartz, k-feldspar, and albite) from lepidolite. However, the intergrowth of these gangue minerals results in uncomplicated liberation (i.e., inclusions). Taking advantage of this coarse gangue liberation, optical ore sorting through image analysis was attempted in order to predict the grades of different-sized fractions using a random comminution algorithm. The ore-sorting process allowed the production of a marketable Li pre-concentrate product for metallurgy. Moreover, this method also highlighted the possible valuation of the reject as low-Li-content quartz–feldspar mixtures for the ceramic industry (reduction in the temperature of porosity closing). Furthermore, a scaled approach of grinding and sieving allowed the formation of a lepidolite-rich fraction (>210 μm), which was processed using an electrostatic separator by varying key process parameters. The lepidolite and muscovite were separated to obtain a Li pre-concentrate assaying 3.5 % Li₂O from a feed grade containing 1.8 % Li₂O. Nevertheless, according to the zeta-potential measurements, the flotation test performed with the finer-sized fraction (–210 + 63 μm) showed that lepidolite flotation was optimised between pH 3 and 5. In this pH range, concentrates from the rougher stage assayed 4.2–4.5 % Li₂O, corresponding to 87–95 % Li recovery. At pH > 5, the selectivity decreases, and SiO₂ analysis suggests the flotation of quartz and other silicates rather than lepidolite. Feldspar/quartz flotation was also tested using lepidolite flotation rejects to promote the separation of feldspars from quartz and obtain products for ceramic applications.

对 Gonçalo 含锂富含云母的伟晶岩矿石（葡萄牙）进行的矿物学研究表明，锂云母存在于粗粒结构中，这使得脉石矿物（石英、钾长石和钠长石）可以从锂云母中明显释放出来。然而，这些脉石矿物的共生导致了简单的释放（即夹杂物）。利用这种粗矸石释放的优势，尝试通过图像分析进行光学矿石分选，以便使用随机粉碎算法预测不同尺寸部分的品位。矿石分选过程允许生产可销售的用于冶金的锂预精矿产品。而且，该方法还强调了将废品作为陶瓷工业中低锂含量石英-长石混合物的可能估值（降低孔隙闭合温度）。此外，研磨和筛分的规模化方法允许形成富含锂云母的部分 (>210 μm)，使用静电分离器通过改变关键工艺参数对其进行处理。将锂云母和白云母分离以获得锂预浓缩物，含量为 3.5%₂ O 来自含有 1.8 % Li ₂ O 的进料级。然而，根据 zeta 电位测量，用更细粒度的部分（–210 + 63 μm）进行的浮选测试表明，锂云母浮选在 pH 3 和5. 在这个 pH 值范围内，来自较粗阶段的精矿测定为 4.2-4.5% Li ₂ O，对应于 87-95% 的锂回收率。在 pH > 5 时，选择性降低，SiO ₂分析表明浮选石英和其他硅酸盐而不是锂云母。长石/石英浮选也使用锂云母浮选废料进行了测试，以促进长石与石英的分离并获得用于陶瓷应用的产品。