The granitoids of the Paleoproterozoic domain have been the most studied in the West African craton. They appear in very varied aspects and have been described in turn as “magmatic”, “migmatitic or metasomatic”. Some authors thought that the granitoids reflected the composition of the rocks crossed during their emplacement [1] - [3] . The structuring (deformation) of these granitoids is more or less intense. The classifications proposed for these granitoids [4] take into account their petrography and their structural context. In Ivory Coast, based on petrographic, geochemical and geochronological criteria, [5] [6] and [7] mainly distinguish: 1) granites with calc-alkaline affinity which show an Archean TTG type character and are dated around 2123 Ma; 2) late metaluminous to peraluminous granites dated around 2097 Ma. These are large batholiths of leucogranites (e.g. Ferkessédougou type) accompanied by small subcircular massifs of trondhjemites, pyroxene granites and alkaline granites (syenites and monzosyenites). The work of [8] in Guinea also shows that alkaline granites were emplaced at the end of the Paleoproterozoic. Geochronological data on the entire West African craton from [5] [7] [9] - [13] indicate that the emplacement of the Birimian granitoids was spread out during the Paleoproterozoic.

[14] showed that the coexistence of muscovite with biotite is a common mineralogical indicator of a highly peraluminous composition of plutonic rocks. Thus, the identification of primary muscovite is important because it is usually a good indicator of both magma composition and crystallization depth. With this group of common and relatively abundant minerals in the granites and granitic pegmatites of Issia [14] , we can thus, with these different micas, provide the characteristics of the different stages of the evolution of the granitic complex of Issia.

In this article, we present a geochemical study of micas minerals from outcrops of G1, G2 and G3 granites [15] , intragranitic pegmatites and pegmatite clusters observed in the Issia region, a granite complex linked to the presence of rare metals in the alluvial, colluvial and eluluvial placers. Our objective in this work is to assess the degree of evolution of the complex, with an emphasis on mica minerals.

The Issia granite complex represents the southern part of the Ferkessedougou batholith and is located in west-central Ivory Coast. It is a large, elongated, multi-story plutonic batholith that extends 500 km from the Burkina Faso border in the northeast to the Sassandra-Cavally domain (SASCA) in the southwest of the Ivory Coast ( Figure 1 ). The Ferké batholith is composed of a two-mica granite with aluminum-potassium chemistry. It is oriented NE direction, forming a linear structure 5 to 50 km wide ( Figure 1(A) ). The deformations affecting this batholith have been described by [16] and [15] . According to [15] , the Issia granite complex can be subdivided into three major group: the G1, G2 and G3 granites. According to [15] , G1 are two-mica granites with dominant biotite ( Figure 1(B) ), igneous or I type, calc-alkaline to weakly peraluminous with low phosphorus contents (0.07% - 0.16% wt% P₂O₅). They contain oxides as ilmenite, rutile and titanomagnetite. G2 and G3 are two-mica leucogranites with dominant muscovite ( Figure 1(B) ), type S therefore of metasedimentary origin, strongly peraluminous (ASI > 1.19) with intermediate phosphorus contents (0.2% - 0.5% P₂O₅) and accessory minerals such as Zn-rich ilmenite, rutile and apatite.

The G3 granites also have an abundance of tourmaline and are provided by pegmatite veins of varying thickness. The ilmenite of these granites is rich in Nb and Ta, and is sometimes associated with Nb-Ta oxides. These three groups of granites are also distinguished from each other by their rare earth spectra (REE) as well as their degree of differentiation. This is determined from classic differentiation parameters such as the Ba-Sr-Rb-Cs contents or the K/Rb and Nb/Ta ratios.

Thin sections of different types of granites and pegmatites were studied under an optical microscope using transmitted and reflected light in order to characterize their textures and the different relationships between the mineral phases. The micas minerals were the subjected to in-situ geochemical analyzes for the major elements (microprobe) and trace elements (LA-ICPMS). These analyses provided insights into the evolution of the magmatic system and helped characterize the nature and origin of events, whether primary (i.e. magmatism) or secondary (i.e. hydrothermalism).

Major element analyzes were focused on several mineral species of micas including biotite and muscovite. The concentrations of major elements were obtained using a CAMECA SX Five electronic microprobe (EPMA) at the Raimond Castaing Microcharacterization Center of the Paul Sabatier University of Toulouse. The analysis was subjected with an accelerating voltage of 15 keV and an intensity of 10 nA, following the standard program Micas (MVL-Mica-Rb-Cs)-Silicates F-Cl-Ba dedicated to silicate minerals. The standards used were Albite (Na Ka), Rb Glass (Rb La), Al₂O₃ (Al Ka), Wollastonite (Si Ka, Ca Ka), Sanidine (K Ka), Cs Glass (Cs La), MnTiO₃ (Ti Ka, Mn), Fe₂O₃ (Fe Ka), Topaz-TR (F Ka), MgO (Mg), BaSO₄ (Ba La), Cr₂O₃ (Cr Ka), Graf (P Ka), SnO₂ (Sn La), Ta (Ta Ma), Tugtupite (Cl Ka), Cr₂O₃ (Cr Ka), Ni-G5 (Ni Ka). Since analyzes of mineral phases using a microprobe do not allow us to have the contents of trace elements such as Li, we used other techniques. The Li contents of the Issia (IS24, IS25), Liga (IS03), Bolia (IS14) massifs and the IS01 pegmatite observed in the metasediments were obtained using LA-ICP-MS (laser-ablation inductively coupled plasma-mass spectrometry) from the GET laboratory. For the other formations were calculated using the Tischendorf regression equations based on the ratios between Li₂O, F and Rb₂O with Li₂O = 0.3935F^1.326 (R² = 0.843, n = 199) [di 1] et Li₂O = 1.579Rb₂O^1.45 (R² = 0.71, n = 209) [di 2].

The different geochemical variations observed in the geochemical analyzes of the micas show that the granites of the Issia region could come from different sources. We can see that the FeO, MgO and TiO₂ contents of the muscovites in the granites decrease from G1 granites, G2 granites to G3 granites, without direct correlation between the different groups.

The projection of the analysis points of dioctahedral micas (white micas) on the binary diagram R3 + as a function of Fe + Mg + Ti + (Si-3) of [18] ( Figure 2(A) & Figure 2(B) ) allowed us to observe a Tschermak substitution or phengitic substitution (Al + Al ↔ Fe, Mg + Si) of the muscovites of the Issia granitoids. This Celadonite or Tschermakitic substitution can be defined as R³⁺ + Al^lV = Si + R²⁺ (R³⁺ = Al, Fe³⁺ and R²⁺ = Mg, Fe²⁺, Mn).

This substitution depends on temperature and pressure. Relatively high pressure and low temperature have been shown to promote an increase in the phengitic component in muscovite [27] . [28] propose, based on thermodynamic considerations, that celadonitic muscovite is more stable than pure muscovite at a higher temperature. These diagrams also show that the most phengitic muscovites are those of the G1 and G2 granites. We can also see that the muscovites which tend more towards the pure muscovite pole are those of the G3 granites and pegmatites ( Figure 2(C) ), which would therefore mean that these muscovites were formed at lower temperatures than those of the G1 and G2 granites.

The progressive enrichment in Na of muscovites from granites G3 to pegmatite IS01 (pegmatite outside granite) passing through the muscovites of intragranite pegmatites, suggests an evolution of granitic fluid towards a pegmatitic fluid. This evolution would be put in place by a fractional crystallization of the source magma, with pegmatite IS01 representing the most evolved stage. The transition of the micas of granites to pegmatites, as observed in the binary diagram of [22] ( Figure 3(B) ), from the muscovite pole to the zinnwaldite pole via Li-muscovite and Li-phengite, shows an enrichment of the source magma in Li and other incompatible elements.

As shown by the work of [15] using whole rock geochemistry, the biotites of the G3 granites display characteristics of peraluminous magma of metasedimentary origin (S-type granite). In contrast, the biotites of the G2 granites exhibit characteristics of calc-alkaline magmas. The biotites of the G1 granites are distributed within the fields of both calcalkaline and peraluminous magmas, indicating a mixed origin (I- and S-type), as demonstrated in the geochemistry section by [15] .

Chemical analyzes on the mineral phases were carried out on dioctahedral (muscovite) and trioctahedral micas. For dioctahedral micas, it should be noted that the muscovites in the G1 and G2 granites have the particularity of being richer in Fe but poorer in Na than the G3 granites. Different substitution diagrams showed that most of the muscovites have undergone phengitic substitution and that the muscovites which approach the pure muscovite pole are those in G3 granites and pegmatites. All analyzed muscovites show characteristics of being either magmatic (primary muscovites) or late magmatic (secondary muscovites). With the exception of the granodiorite biotites (DAL06), which showed phlogopite characteristics, all the granite biotites are located between the siderophyllite-annite poles and are rich in Fe (Fe-biotites) and Al. However, the biotites G1 granites are less ferrous than those of G2 and G3 granites. The study of biotites has shown that the G3 granites are peraluminous and of type S, whereas the G2 and G1 granites display the characteristics of granites originating from magma of mixed origin (calcalkaline and peraluminous), thus classified as both type S and I. Although the majority of the biotites studied present the primary characteristics, we also observe some re-equilibrated primary biotites and secondary biotites.

In perspective to this, since rare metal granites and pegmatites are very important rocks economically because they can be extremely enriched in some rare elements such as Li, Ta, Sn, Nb, Be and Cs, which are essential for the development of high-tech industries; it would be interesting to study these elements which can be enriched in micas.

This work was supported by the project T2GEM (Technologies Géophysiques et Géochemiques pour l’Exploration Minière). We thank the French Institute for Research and Development (IRD) for supporting D. Baratoux’s and M. Van Lichtervelde’s visits to UFHB between 2015 and 2022. We also thank Philippe De Parseval and Thierry Aigouy for for their assistance with the SEM and microprobe data acquisition at Géosciences Environnement Toulouse (GET) and Castaing Center.