1584 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
This observation is consistent with microscopic observa-
tions. Hence, here the flotation seems to be selective on
lepidolite grains rather than muscovite.
The modal composition of the sample is given in
Table 4. While good agreements are obtained for the quan-
tification of quartz and potassium feldspar, limitations are
observable for the quantification of lepidolite and albite
(Table 4). This may be related to the semi-quantitative
determination of sodium content by ED-XRF, which can
lead to erroneous quantification of the albite content of the
sample. Furthermore, a large amount of minerals is attrib-
uted to topaz by matrix calculation (Table 4). This may be
related to the reduced chemistry of this mineral (i.e., Al, Si,
F) and the low amount of sodium determined by ED-XRF.
Finally, several challenges related to optimizing lepido-
lite recovery can be identified: (i) mixed particles contain-
ing lepidolite, suggesting that for this granular fraction,
liberation is not achieved (ii) albite and potassium feldspar
Figure 6. Lithium intensity map from the 670.7 nm lithium emission line analysed
through µXRF (60,000 to 0 cps), along with LA-ICP-MS assays spots (n =42)
Figure 7. False coloured mineralogical classified map. The red rectangle represents the part analysed through µLIBS method.
Light blue, blue, pink, oranget and green masks represent quartz, muscovite, lepidolite, K-feldspar and albite, respectively
Table 4. Modal mineralogy calculated through MARCIA (automated mineralogy) and EMC (ED-
XRF) methods
Mineral Quartz Lepidolite Muscovite K-feldspar Phosphate Albite Topaz
%-MARCIA 2.4 69.4 1.1 — 26.3 0.7
%-EMC 5.0 57.0 0.6 — 8.4 22.57
This observation is consistent with microscopic observa-
tions. Hence, here the flotation seems to be selective on
lepidolite grains rather than muscovite.
The modal composition of the sample is given in
Table 4. While good agreements are obtained for the quan-
tification of quartz and potassium feldspar, limitations are
observable for the quantification of lepidolite and albite
(Table 4). This may be related to the semi-quantitative
determination of sodium content by ED-XRF, which can
lead to erroneous quantification of the albite content of the
sample. Furthermore, a large amount of minerals is attrib-
uted to topaz by matrix calculation (Table 4). This may be
related to the reduced chemistry of this mineral (i.e., Al, Si,
F) and the low amount of sodium determined by ED-XRF.
Finally, several challenges related to optimizing lepido-
lite recovery can be identified: (i) mixed particles contain-
ing lepidolite, suggesting that for this granular fraction,
liberation is not achieved (ii) albite and potassium feldspar
Figure 6. Lithium intensity map from the 670.7 nm lithium emission line analysed
through µXRF (60,000 to 0 cps), along with LA-ICP-MS assays spots (n =42)
Figure 7. False coloured mineralogical classified map. The red rectangle represents the part analysed through µLIBS method.
Light blue, blue, pink, oranget and green masks represent quartz, muscovite, lepidolite, K-feldspar and albite, respectively
Table 4. Modal mineralogy calculated through MARCIA (automated mineralogy) and EMC (ED-
XRF) methods
Mineral Quartz Lepidolite Muscovite K-feldspar Phosphate Albite Topaz
%-MARCIA 2.4 69.4 1.1 — 26.3 0.7
%-EMC 5.0 57.0 0.6 — 8.4 22.57