1582 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
(intensity of approximately 80% and 40%, respectively).
Thus, using these two elements, it is possible to effectively
differentiate muscovite, lepidolite, and potassium feldspar.
Moreover, potassium is one of the first elements that can
be reliably quantified by XRF spectroscopy tools, including
portable and micro XRF (Rose et al., 1963).
The use of elemental mappings of K and Rb to define
these three minerals is clearly highlighted in the mineral-
ogical classification map of the thin section (Figure 4).
This classification is achieved using chemical proxies: the
quartz mask is defined using Si, the apatite mask using Ca
and P, the potassium feldspar, muscovite, and lepidolite
masks using K and Rb as described earlier. Finally, as Na
is a noisy element, the albite mask is constructed by negat-
ing all defined chemical boundaries. Using this approach,
all mineral phases are successfully identified and classi-
fied (Figure 4). This thin section does not show any resin.
However, it should be noted that when resin was present on
thin sections, it was realistically defined.
The quantifications of mineral phases by the MARCIA
mineralogical classification code and by modal mineral-
ogy calculated by matrix calculation (EMC) are given in
Table 2. A very good agreement between the two methods
is obtained. Only the quantifications of albite and quartz
vary (Table 2). However, the proportion of these two min-
erals can vary locally depending on the weathering degree
of the granite. Furthermore one analysis is carried out on a
local hard sample (thin section), whereas the other analysis
is carried out on a volume that is representative of the drill
core considered.
Automated Mineralogy on Polished Section Using
SEM, µXRF and µLIBS
Sample
The analyzed sample is a flotation concentrate with a grain
size class of 10–500µm. This sample was obtained during a
flotation test for a study on the treatment of Beauvoir gran-
ite. The purpose of this study was to define the liberation
mesh of the granite, as well as the technical and scientific
limits of the treatment scheme defined for the valorization
of this deposit. The chemistry of the sample, obtained by
ED-XRF, is provided in Table 3.
The lithium content in this concentrate is approxi-
mately 3% Li2O, which is a concentration typically
obtained by flotation (Table 3). Furthermore, the tin con-
tent in the concentrate is quite high and will be discussed
in this case study.
Using this sample a polished section was produced
following the procedure described in the Materials and
Methods part. A photograph of the polished section is dis-
played in Figure 5.
Figure 3. (a) Thin section observed in cross polarized light along its (b) K map and (c) Rb map. Qz
=quartz, Ab =albite, Kfd =K-feldpar, Lpd =lepidolite et Msc =muscovite
(intensity of approximately 80% and 40%, respectively).
Thus, using these two elements, it is possible to effectively
differentiate muscovite, lepidolite, and potassium feldspar.
Moreover, potassium is one of the first elements that can
be reliably quantified by XRF spectroscopy tools, including
portable and micro XRF (Rose et al., 1963).
The use of elemental mappings of K and Rb to define
these three minerals is clearly highlighted in the mineral-
ogical classification map of the thin section (Figure 4).
This classification is achieved using chemical proxies: the
quartz mask is defined using Si, the apatite mask using Ca
and P, the potassium feldspar, muscovite, and lepidolite
masks using K and Rb as described earlier. Finally, as Na
is a noisy element, the albite mask is constructed by negat-
ing all defined chemical boundaries. Using this approach,
all mineral phases are successfully identified and classi-
fied (Figure 4). This thin section does not show any resin.
However, it should be noted that when resin was present on
thin sections, it was realistically defined.
The quantifications of mineral phases by the MARCIA
mineralogical classification code and by modal mineral-
ogy calculated by matrix calculation (EMC) are given in
Table 2. A very good agreement between the two methods
is obtained. Only the quantifications of albite and quartz
vary (Table 2). However, the proportion of these two min-
erals can vary locally depending on the weathering degree
of the granite. Furthermore one analysis is carried out on a
local hard sample (thin section), whereas the other analysis
is carried out on a volume that is representative of the drill
core considered.
Automated Mineralogy on Polished Section Using
SEM, µXRF and µLIBS
Sample
The analyzed sample is a flotation concentrate with a grain
size class of 10–500µm. This sample was obtained during a
flotation test for a study on the treatment of Beauvoir gran-
ite. The purpose of this study was to define the liberation
mesh of the granite, as well as the technical and scientific
limits of the treatment scheme defined for the valorization
of this deposit. The chemistry of the sample, obtained by
ED-XRF, is provided in Table 3.
The lithium content in this concentrate is approxi-
mately 3% Li2O, which is a concentration typically
obtained by flotation (Table 3). Furthermore, the tin con-
tent in the concentrate is quite high and will be discussed
in this case study.
Using this sample a polished section was produced
following the procedure described in the Materials and
Methods part. A photograph of the polished section is dis-
played in Figure 5.
Figure 3. (a) Thin section observed in cross polarized light along its (b) K map and (c) Rb map. Qz
=quartz, Ab =albite, Kfd =K-feldpar, Lpd =lepidolite et Msc =muscovite
