Abstract

Major and accessory minerals from the Ibity granite, Tsarasaotra monzonitic and granite dykes, and Antsahakely granite of the Itremo domain in the Precambrian basement of Madagascar were characterized by using microscopic observations and chemical analyses with the aim of understanding their chemical characteristics and estimating the crystallization pressure and oxygen fugacity of their host rocks. Plagioclases in these rocks are albite and oligoclase, while alkali feldspars are orthoclase. For the phlogopite-micas, Fe-biotite and Li-phengite are common for the Ibity and Antsahakely granites, Mg-biotite is common for the Ibity granite and the Tsarasaotra monzonitic and granite dykes, and siderophyllite and Zinnwaldite are specific to the Ibity granite. Phlogopite-micas in the studied rocks are mainly primary, accessorily re-equilibrated, and rarely secondary. Calcic amphiboles distributed in the Magnesio-and Ferro-hornblende are identified in the Tsarasaotra monzonitic, whereas amphibole is rare and absent in the other rocks. Igneous titanite is observed in the Ibity granite and in the Tsarasaotra monzonitic rocks, which have similar compositions to some REE oxide-rich titanites. Concerning the Fe-Ti oxide phases, the rhombohedral and spinel/trifer tetroxide phases are found in both the Tsarasaotra monzonitic and the Tsarasaotra granite dyke, the trifer tetroxide and spinel + wüstite phases are found only in the Ibity granite, and the pseudobrookite + rhombohedral phase is found only in the Tsarasaotra granite dyke. The epidote mineral, rarely found in the Antsahakely granite, could be an indicator of metamorphism or hydrothermal activity involved during the emplacement of this rock. Aluminum in hornblende geobarometer gave pressure ranges of around 5 kbar for the Tsarasaotra monzonitic rocks. The Titanite geobarometer gave pressures of 2.5 - 3.2 kbar for the Ibity granite, 2.9 kbar for the Tsarasaotra monzonitic, and 7.1 kbar for the Antsahakely granite. Both amphibole and Fe-Ti oxide-base oxygen fugacity reveal high oxygen fugacity conditions for the Tsarasaotra monzonitic and granite dyke emplacements, which might have a relationship with a porphyritic environment.

Keywords

Itremo, Granitic, Monzonitic, Mineral Chemistry, Geobarometer, Fugacity

Share and Cite:

1. Introduction

Among the geodynamic domains of the Precambrian terrane of Madagascar, the Itremo domain is located in the center-south of Madagascar, surrounding the Ambatofinandrahana district and Ibity town, which are two famous places in terms of mineral potential and productivity in Madagascar (Lacroix, 1922). These mineral potentials, composed of gemstones and semi-precious stones as industrial ores, have been operated by small-scale mining operators as well as by companies for decades (e.g. Ravoniarisoa & Rakotomanana, 2002).

Several works discuss the petrology and the geochemistry of the magmatic suites in the Itremo domain, such as the Imorona-Itsindro suite (Handke, Tucker, & Ashwal, 1999; McMillan et al., 2003; Moine, Bosse, Paquette, & Ortéga, 2014; Rasoamalala et al., 2014; Tucker, Roig, Moine, Delor, & Peters, 2014; Yang et al., 2015) and the Ambalavao suite (Roig, Tucker, Peters, Delor, & Theveniaut, 2012; Tucker, Peters, Roig, Théveniaut, & Delor, 2012; Tucker, Roig, Moine, Delor, & Peters, 2014; Rasoamalala et al., 2014). Other studies are related to the metamorphic rocks in the Itremo domain (Cox, Armstrong, & Ashwal, 1998; Morteani & Ackermand, 2006; Tucker, Kusky, Buchwaldt, & Handke, 2007). The majority of magmatism-related works were based on whole rock geochemistry approaches, with a primary emphasis on the petrology and geochemistry of these magmatic rocks as isotopic and geochronological studies. A mineral chemistry-based approach such as Rasoamalala et al. (2014) was lacking for the study of these rocks, as well as the geographic distribution and relationships among the two magmatic suites in the Itremo domain. Moine, Nédélec, & Ortéga (2014) build on previous work on the geology of Madagascar’s Precambrian terrane, particularly in terms of geological cartography. Recent work by Rakotondravaly & Randrianja (2022) contributes to these works by providing more information related to the geographical distribution of these two magmatic suites in the Itremo domain, as well as the membership and possible relationships of the granitic rocks in that domain. However, mineral chemistry information regarding these rocks remains poorly documented.

This work aims at studying the mineral chemistry of the granitic rocks in the Itremo domain, and will focus on the study of the main minerals from the Tsarasaotra monzonitic and granite dyke, the Ibity and the Ambatofinandrahana granites from the Itremo domain, as well as their formation environment by using geochemical approaches. Mineral chemistry studies of these rocks will provide additional information regarding their geochemical characteristics.

2. Geological Setting of the Study Area

The Itremo domain is among the nine (09) divisions of the tectonometamorphic domains of the Madagascar Precambrian socle (Collins & Windley, 2002; Moine, Nédélec, & Ortéga, 2014), which is mainly composed of Archean rocks in the North and Proterozoic rocks in the South (Yoshida, 1998; Roig, Tucker, Peters, Delor, & Theveniaut, 2012; Tucker, Roig, Moine, Delor, & Peters, 2014). This domain consists of parametamorphic formations composed of schist, quartzite, and cipolin known as the Schisto-Quartzo-Dolomite formation or Itremo group, which is intruded by the Imorona-Itsindro suite and the Ambalavao suite (Roig, Tucker, Peters, Delor, & Theveniaut, 2012) (Figure 1).

The Imorona-Itsindro suite, composed of gabbro and diorite, and granite typical of calc-alkaline porphyritic massive granite (Moine, 1974; Daso, 1986; Tucker,

Peters, Roig, Théveniaut, & Delor, 2012), is thought to have been emplaced by: 1) crustal extension before 747 Ma, followed by 2) lithospheric subduction and arc magmatism around and after 729 - 727 Ma (Yang et al., 2015). The Ambalavao suite, composed of alkali-potassic plutonic formations (essentially granite and syenite) (Emberger, 1956; Tucker, Peters, Roig, Théveniaut, & Delor, 2012), is thought to have been emplaced by the magmatic events related to the continental collision between East and West Gondwana between 580 and 520 Ma (Tucker et al., 1999; Archibald et al., 2019).

3. Sampling and Analytical Techniques

Field research was carried out in the Itremo domain in the outcrops of Ibity (around 20.049˚S and 46.983˚E), Tsarasaotra Ambositra (20˚26'15.7"S and 47˚11'35.1"E), and Antsahakely (20˚33'14.8"S and 46˚57'01.6"E). A dozen granitic rock samples were collected for laboratory investigations at the Faculty of International Resource Sciences of Akita University, Japan.

Microscopic observations of slide glass-covered thin sections were performed by using a NIKON ECLIPSE 50i microscope. Polished thin sections and polished samples were made for both microscopic observation and qualitative and quantitative microanalysis under JEOL SUPERPROBE JXA-733. Quantitative analyses of silicates and oxides were conducted with an accelerating voltage of 15 KeV, an electron beam of 20 nA, and 5 to 10 µm in diameter. The measuring time for the standard is 10 s repeated 5 times, and that of the background is 40 s. A Lithium Fluoride (LiF) crystal was used to detect Fe Kα and Mn Kα, Pentaerythritol (PET) crystal for K Kα, Ti Kα, Ca Kα and Thallium Acid Phthalate (TAP) crystal for Na Kα, Si Kα, Mg Kα and Al Kα. As standards, SiO₂, TiO₂, Al₂O₃, Fe₂O₃, MnO, MgO, CaSiO₃, NaAlSi₃O₈, and KAlSi₃O₈ minerals were used. Bence & Albee (1968)’s procedures were used to correct the data.

4. Petrography

More detailed petrographic descriptions can be found in Rakotondravaly & Randrianja (2022).

The Ibity granitic rocks are mainly leucocratic holocrystalline pseudo-idiomorphic grainy granite classified as granite, after Le Maitre et al. (2002) classification scheme. It is essentially composed of plagioclase, alkali feldspar, and quartz, and its accessorily minerals are mafic silicate minerals, magnetite, ilmenite, apatite, and zircon in trace (Figure 2).

The Tsarasaotra principal granitic rocks are mesocratic holocrystalline hypidiomorphic medium-to-coarse grained rocks which are classified into monzonite and quartz monzonite after Le Maitre et al. (2002) classification scheme. They are mainly composed of plagioclase, alkali feldspar, and mafic silicate minerals, and their accessory minerals are quartz, magnetite, ilmenite, apatite, and zircon (Figure 2).

The Tsarasaotra granitic dykes are leucocratic holocrystalline pseudo-idiomorphic coarse-grained rocks classified as granite after Le Maitre et al. (2002)

classification scheme. They are essentially made of plagioclase, alkali feldspar, and quartz, and their accessory minerals are mafic silicate minerals, magnetite, ilmenite, apatite, and zircon in trace (Figure 2).

The Ambatofinandrahana granitic rocks are leucocratic holocrystalline hypidiomorphic medium-to-coarse grained rocks which are classified into granite after Le Maitre et al. (2002) classification scheme. These rocks are essentially composed of plagioclase, alkali feldspar, and quartz, and their accessory minerals are mafic silicate minerals, magnetite, ilmenite, apatite, and zircon in trace (Figure 2).

5. Mineral Chemistry

5.1. Feldspars

The structural formulae of plagioclase feldspars were recalculated on the basis of 8 oxygens. Representative electron microprobe data of plagioclase is shown in Table 1.

For the Ibity granite, plagioclases are albite (An 1.38 - 7.50) and oligoclase (An 10.50 - 24.09). Albites have SiO₂ concentrations of 64.07 - 69.57 wt%, Al₂O₃ concentrations of 21.28 - 23.06 wt%, CaO concentrations of 0.24 - 1.52 wt%, and Na₂O concentrations of 8.72 - 10.73 wt%. Oligoclases have concentrations of SiO₂ = 62.16 - 67.35 wt%, Al₂O₃ = 21.66 - 25.38 wt%, CaO = 2.04 - 3.07 wt% and Na₂O = 5.19 - 10.49 wt%. Alkali feldspars are orthoclase (Or 92.19 - 97.71) having concentrations in SiO₂ = 60.11 - 66.02 wt%, Al₂O₃ = 18.22 - 20.55 wt%, Na₂O = 0.27 - 0.88 wt%, and K₂O = 15.69 - 18.33 wt% (Table 1 and Figure 3).

Regarding the Tsarasaotra monzonitic rocks, plagioclases are distributed in albite (An 3.78 - 9.29) with concentrations of SiO₂ = 65.72 - 68.32 wt%, Al₂O₃ = 21.89 - 23.29 wt%, CaO = 0.73 - 1.86 wt%, and Na₂O = 9.95 - 10.32 wt%. Alkali feldspars are orthoclase (Or 83.16 - 94.93), with SiO₂ concentrations of 62.08 - 64.44 wt%, Al₂O₃ concentrations of 18.92 - 20.21 wt%, Na₂O concentrations of 0.58 - 1.61 wt%, and K₂O concentrations of 14.53 - 16.96 wt% (Table 1 and Figure 3).

Plagioclases are distributed in oligoclases (An 12.73 - 21.09) in the Tsarasaotra granitic dykes, with SiO₂ = 62.16 - 65.34 wt%, Al₂O₃ = 22.85 - 24.75 wt%, CaO = 2.58 - 4.28 wt%, and Na₂O = 8.58 - 9.70 wt%. Alkali feldspars are orthoclase (Or 89.22 - 97.83) with SiO₂ concentrations of 45.63 - 64.39 wt%, Al₂O₃ concentrations of 18.92 - 35.26 wt%, Na₂O concentrations of 0.17 - 1.26 wt%, and K₂O concentrations of 11.39 - 17.83 wt% (Table 1 and Figure 3).

For the Antsahakely granite, plagioclases are rare, so analyses were focused on alkali feldspars. Alkali feldspars are orthoclase (Or 95.06 - 97.15), having compositions of SiO₂ = 61.26 - 64.91 wt%, Al₂O₃ = 19.01 - 20.15 wt%, Na₂O = 0.34 - 0.58 wt%, and K₂O = 17.13 - 17.50 wt% (Table 1 and Figure 3).

5.2. Phlogopite-Micas

On the basis of 11 oxygens, the structural formulae of plagioclase feldspars were recalculated. Representative electron microprobe data of plagioclase and their calculated formulae are given in Lithium composition estimates and the classification diagram are after Tischendorf et al. (1997). Then, the 10 × TiO₂-MgO-FeO^T + MnO ternary system was proposed by (Nachit, Ibhi, Abia, & Ohoud, 2005) for discriminating primary, re-equilibrated primary, and secondary biotite.

For the Ibity granite, phlogopite-micas are distributed mainly within the siderophyllite, Fe-biotite, and Mg-biotite zones, accessorily within the Li-phengite zone, and rarely in the Zinnwaldite zone (Figure 4). Siderophyllites in the Ibity granite have concentrations of SiO₂ = 35.71 - 37.98 wt%, TiO₂ = 1.13 - 2.80 wt%, Al₂O₃ = 16.62 - 17.43 wt%, FeO = 19.83 - 29.28 wt%, MgO = 3.62 - 5.71 wt%, K₂O = 9.42 - 10.59 wt%, and Li₂O = 0.66 - 1.32 wt%. Fe-biotites have compositions of SiO₂ = 34.20 - 38.36 wt%, TiO₂ = 0.56 - 3.80 wt%, Al₂O₃ = 14.87 - 20.62 wt%, FeO = 14.01 - 26.78 wt%, MgO = 3.75 - 7.83 wt%, K₂O = 9.36 - 13.18 wt%, and Li₂O = 0.23 - 0.84 wt%. Then, Mg-biotites have concentrations of SiO₂ = 34.45 - 38.07 wt%, TiO₂ = 2.28 - 3.31 wt%, Al₂O₃ = 15.06 - 16.90 wt%, FeO = 17.58 - 20.88 wt%, MgO = 10.15 - 11.63 wt%, K₂O = 6.06 - 10.37 wt%, and Li₂O = 0.08 - 0.12 wt%. Regarding Li-phengites, they have compositions of SiO₂ = 41.74 - 45.08 wt%, TiO₂ = 0.03 - 0.78 wt%, Al₂O₃ = 30.82 - 34.54 wt%, FeO = 4.56 - 5.06 wt%, MgO = 0.20 - 2.20 wt%, K₂O = 10.59 - 11.20 wt%, and Li₂O = 2.41 - 3.37 wt%. Concerning the Zinnwaldites, they have concentrations of SiO₂ = 45.39 wt%, TiO₂ = 0.27 wt%, Al₂O₃ = 32.71 wt%, FeO = 6.35 wt%, MgO = 1.08 wt%, K₂O = 10.49 wt%, and Li₂O = 3.46 wt%. Biotite in the Ibity granite is distributed mainly in the domain of primary and re-equilibrated primary biotite and rarely in the secondary biotite (Table 2 and Figure 5).

For the Tsarasaotra monzonitic rocks and Tsarasaotra granitic dykes, phlogopite-micas are Mg-biotites. For the monzonitic rocks, they have compositions of SiO₂ = 36.67 - 37.51 wt%, TiO₂ = 2.74 - 4.01 wt%, Al₂O₃ = 14.18 - 15.18 wt%, FeO = 16.15 - 19.73 wt%, MgO = 10.38 - 12.46 wt%, K₂O = 9.98 - 10.60 wt%, and Li₂O = 0.06 - 0.11 wt%, while the granitic dykes have concentrations of SiO₂ = 34.93 - 37.24 wt%, TiO₂ = 2.25 - 3.11 wt%, Al₂O₃ = 14.60 - 15.15 wt%, FeO = 18.01 - 18.70 wt%, MgO = 10.39 - 11.27 wt%, K₂O = 10.27 - 11.08 wt%, and Li₂O

= 0.08 - 0.11 wt%. Except a rare re-equilibration signature identified for the granite dyke sample TS-2 (10 × TiO₂ = 22.46 wt%; FeO + MnO = 19.03 wt%; MgO = 10.89 wt%), all other biotites are primary (Table 2 and Figure 5).

For the Ambatofinandrahana granites, phlogopite-micas are distributed in the Fe-biotite, Zinnwaldite, and Li-phengite fields (Figure). SiO₂ = 34.77 wt%, TiO₂ = 1.98 wt%, Al₂O₃ = 18.76 wt%, FeO = 24.31 wt%, MgO = 4.83 wt%, K₂O = 7.69 wt%, and Li₂O = 0.39 wt% are the compositions of Fe-biotite in these rocks. Zinnwaldite has compositions of SiO₂ = 45.69 wt%, TiO₂ = 0.53 wt%, Al₂O₃ =

33.50 wt%, FeO = 5.81 wt%, MgO = 1.39 wt%, K₂O = 11.19 wt%, and Li₂O = 3.54 wt%, whereas Li-phengite has compositions of SiO₂ = 43.72 wt%, TiO₂ = 0.44 wt%, Al₂O₃ = 33.90 wt%, FeO = 5.94 wt%, MgO = 1.09 wt%, K₂O = 11.12 wt%, and Li₂O = 2.98 wt%. Except for a primary biotite observed (10 × TiO₂ = 6.01 wt%; FeO + MnO = 5.25 wt%; MgO = 1.54 wt%), other biotites were objects of re-equilibration (Table 2 and Figure 5).

5.3. Amphiboles

Structural formulae of amphiboles were recalculated on the basis of 23 oxygen (22 oxygen and 2OH) by following the procedures after Li et al. (2020) . Representative chemical compositions of amphibole determined by electron microprobe analyses and their calculated formulae are given in Table. Nomenclature and classification diagram is after Leake et al. (1997).

In the studied rocks, amphiboles were identified and analyzed in the Tsarasaotra monzonitic rocks. They are calcic amphiboles with Ca_B ≥ 1.50 and (Na + K)_A < 0.50. These amphiboles distribute in the Magnesio and Ferro-hornblende (Si = 6.70 - 6.82; Mg# = 0.49 - 0.52; (Na + K)_A = 0.34 - 0.40 et Ca_A = 0.00) according to the classification diagram (Figure). They have concentrations of SiO₂ = 42.34 - 44.86 wt%, TiO₂ = 0.86 - 1.37 wt%, FeO = 15.96 - 16.78 wt%, MgO = 7.93 - 8.84 wt%, CaO = 10.69 - 11.50 wt%, Na₂O = 1.56 - 1.66 wt%, and K₂O = 1.30 - 1.46 wt% (Table 3 and Figure 5).

5.4. Titanite

The structural formulae of titanites were recalculated on the basis of five oxygens. Representative microprobe analyses of titanites with recalculated formulae are shown in Table 4.

Among the studied rocks, the Ibity granite, the Tsarasaotra monzonitic rocks, and the Antsahakely granite all contain some titanite minerals.

For the Ibity granite, titanites have compositions of SiO₂ = 29.91 - 30.42 wt%, TiO₂ = 28.95 - 34.84 wt%, Al₂O₃ = 1.87 - 2.56 wt%, FeO = 1.90 - 2.17 wt%, MnO = 0.11 - 0.28 wt%, and CaO = 25.96 - 26.72 wt% (Table 4), which are similar to those of the black titanite in the Quoscescer pegmatite in the North-East of Harar, Ethiopia. That black titanite also contains P₂O₅ (0.06 wt%), BaO (0.04 wt%), ZrO₂ (in trace) as well as REE oxides such as Ce₂O₃ (2.98 wt%) and Y₂O₃ (1.53 wt%) (Deer, Howie, & Zussman, 2013).

Titanite in the Tsarasaotra monzonitic rocks has concentrations of SiO₂ = 29.68 wt%, TiO₂ = 34.86 wt%, Al₂O₃ = 2.32 wt%, FeO = 2.21 wt%, MnO = 0.29 wt%, and CaO = 25.61 wt% (Table 4).

Regarding the Antsahakely granite, the observed titanite has the compositions of SiO₂ = 29.16 wt%, TiO₂ = 30.11 wt%, Al₂O₃ = 6.40 wt%, FeO = 3.16 wt%, MnO = 0.13 wt%, and CaO = 25.51 wt% (Table 4).

Based on Aleinikoff et al. (2002)’s diagram, titanites in the Ibity granite and Tsarasaotra monzonitic rocks are magmatism-related (Al = 0.08 - 1.00; Fe = 0.05 - 0.06) and (Al = 0.09; Fe = 0.06), respectively, whereas those of the Antsahakely granite have a metamorphic origin (Al = 0.26; Fe = 0.09) (Figure 6).

5.5. Fe-Ti Oxides

The structural formulae were recalculated on the basis of four oxygens for the magnetite and three oxygens for the ilmenite, as well as Fe²⁺ and Fe³⁺ components

according to Carmichael (1966)’s method. The molar fractions in percent of Ulvöspinel (Usp mol%) and Ilmenite (Ilm mol%) were recalculated according to Carmichael (1966) and Lindsley & Spencer (1982)’s methods. In the studied rocks, Fe-Ti oxides were identified except in the Ambatofinandrahana granites, for which representative microprobe data are shown in Table 5.

According to Taylor (1964) phase equilibria in the ternary system FeO-TiO₂-Fe₂O₃ at 1300˚C, the stable phases of Fe-Ti oxides in the Ibity granite are spinel under the Trifer tetroxide (TiO₂ = 0.01 wt%; FeO = 35.5 wt%; Fe₂O₃ = 38.4 wt%) and Spinel + wüstite (TiO₂ = 0.01 - 0.04 wt%; FeO = 30.7 - 31.0 wt%; Fe₂O₃ = 68.0 - 68.7 wt%) (Table 5 and Figure 7).

For the Tsarasaotra monzonitic rocks, identified stable phases are spinel/trifer tetroxide (TiO₂ = 0.02 - 0.04 wt%; FeO = 31.4 - 32.3 wt%; Fe₂O₃ = 69.6 - 71.2 wt%) and rhombohedral (TiO₂ = 42.7 - 46.5 wt%; FeO = 35.7 - 38.9 wt%; Fe₂O₃ = 12.3 - 14.4 wt%) (Table 5 and Figure 7).

Concerning the Tsarasaotra granitic dykes, the spinel/trifer tetroxide phase (TiO₂ = 0.00 - 0.04 wt%; FeO = 30.1 - 31.6 wt%; Fe₂O₃ = 66.4 - 69.9 wt%) and pseudobrookite + rhombohedral phase (TiO₂ = 46.0 wt%; FeO = 33.2 wt%; Fe₂O₃ = 11.0 wt%) were observed in addition to the rhombohedral phase (TiO₂ = 9.40 wt%; FeO = 8.46 wt%; Fe₂O₃ = 75.7 wt% and TiO₂ = 47.4 wt%; FeO = 40.5 wt%; Fe₂O₃ = 9.76 wt%) (Table 5 and Figure 7).

5.6. Epidote

The structural formula of epidote was recalculated on the basis of 12 oxygen atoms. Absent in most of the studied rocks and rare in the Antsahakely granite, epidote has concentrations of SiO₂ = 37.72 wt%, TiO₂ = 0.10 wt% Al₂O₃ = 25.14 wt%, FeO = 11.33 wt%, MnO = 0.61 wt%, and CaO = 22.42 wt% which is aluminum rich compared to that reported by Anthony (1990) of Westfield, Hampden County, Massachusetts, USA. The presence of epidote could be an indicator of a certain degree of metamorphism or hydrothermal activity in the Antsahakely granite.

6. Geobarometry and Oxygen Fugacity Estimates

6.1. Aluminum in Hornblende Geobarometer

The aluminum in hornblende geobarometry was applied to the Tsarasaotra monzonitic rocks in which amphiboles were identified and analyzed. The method proposed by Schmidt (1992) gives pressure estimates in the range of 4.7 to 5 kbars (±0.6 kbars) based on total aluminum in formula (Al^tot = 1.62 - 1.69), whereas Anderson & Smith (1995) provides pressures around 5 kbars (Fe/(Fe + Mg) = 0.53 - 0.55) (Figure 8).

6.2. Titanite Geobarometer

Since titanite minerals were identified and analyzed in the Ibity granite, the Tsarasaotra monzonitic rocks, and the Antsahakely granite, the titanite geobarometer was applied to these rocks by using the method recently proposed by Erdmann et al. (2019). Based on Al₂O₃ content in weight percent, average pressures of 2.5 - 3.2 kbars, 2.9 kbars, and 7.1 kbars were obtained for the Ibity granite, the Tsarasaotra monzonitic, and the Antsahakely granite, respectively (Table 6).

6.3. Oxygen Fugacity

On the basis of Amphibole

An oxygen fugacity estimate was applied to the Tsarasaotra monzonitic rocks by using the method proposed by Anderson & Smith (1995). According to the Fe/(Fe + Mg) ratio versus the Al^IV expressed in atom per formula unit (apfu), regardless of the Al^IV content (0.37 - 0.45), the Tsarasaotra monzonitic rocks have Fe/(Fe + Mg) ratio less than 0.6 (0.53 - 0.55), indicating a formation condition with high oxygen fugacity (Figure 9(a)).

On the basis of Fe-Ti oxides

Oxygen fugacity and temperatures were also estimated by using the method proposed by Spencer & Lindsley (1981) according to the magnetite-ilmenite mineral assemblage. Based on this method, the Tsarasaotra rocks give oxygen fugacity estimates ranging from −18.2 to −17.5 for the monzonitic rocks and from −19.9 to −17.7 for the granite dyke. These results are above the NNO buffer, which indicates a formation under a highly oxidized environment. For comparison, the Tsarasaotra monzonitic rocks were formed under a similar oxidation environment as that of granitic rocks around the Kamaishi Cu-Fe skarn deposit as reported in Rakotondravaly (2018). The Tsarasaotra monzonitic rocks and granite dykes were both formed under much higher oxygen fugacity relative to the Yerington batholith (Dilles, 1987), which might indicate a porphyritic environment affinity (Figure 9(b); Table 7).

7. Conclusion

Plagioclase and alkali feldspars are among the main minerals found in the Ibity granite, the Tsarasaotra monzonitic and granite dykes, and the Antsahakely granite, while biotite, amphiboles, Fe-Ti oxide minerals, and titanite are accessory minerals.

Regarding the studied rocks, plagioclases are mainly distributed in albite and oligoclase, while alkali feldspars have a composition of orthoclase. Phlogopite-mica is distributed mostly in and around the biotite fields, and accessorily in and around the muscovite fields. Biotite is mostly distributed in the primary and re-equilibrated primary biotite fields. Amphibole is hornblende for the Tsarasaotra monzonitic, which is rare and absent for other studied rocks. Igneous titanite is observed in the Ibity granite and Tsarasaotra monzonitic, while titanite is not identified in the other studied rocks. The Fe-Ti oxide minerals of the studied rocks are found in both titanium-rich and titanium-free phases. An indication of either metamorphism or hydrothermal activity was observed in the Antsahakely granite.

The aluminum in hornblende geobarometer gave pressure estimates higher than the titanite geobarometer for the Tsarasaotra monzonitic rock. The Antsahakely granite most likely formed at a lower depth relative to the Ibity granite and the Tsarasaotra monzonitic, which formed at a similar depth.

The high oxygen fugacity estimated in the Tsarasaotra rocks could indicate an affinity for a porphyritic environment.

This work enlightens us regarding the main and accessory minerals in the granitic rocks from the studied area as well as the potential affinity of these rocks for a mineralization process. Additional investigations should be carried out, especially fluid inclusion studies, in order to better understand the possible fluid-mineral interaction relating to the emplacement of these rocks.

Acknowledgements

This research work was financially supported by the Japan International Cooperation Agency (JICA) as part of the human resource development capacity program for the mineral sector in developing countries. We gratefully thank Professor Daizo Ishiyama of the Faculty of International Resource Sciences, Akita University, Japan, for his advice on using EPMA equipment. We are also grateful to anonymous reviewers for their comments and suggestions, which contributed to the improvement of the final version of the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest regarding the publication of this paper.

References