Abstract

The mafic enclaves from Paleoproterozoic domain are considered to be the results of large-scale crust-mantle interaction and magma mixing. In this paper, petrography, mineralogy and geochemistry were jointly used to determine the origin of the mafic enclaves and their relationship with the host granitoids of the Kan granite-gneiss complex. This study also provides new information on crust-mantle interactions. The mafic enclaves of the Kan vary in shape and size and have intermediate chemical compositions. The diagrams used show a number of similarities in the major elements (and often in the trace elements) between the mafic enclaves and the host granitoids. Geochemical show that the Kan rock are metaluminous, enriched in silica, medium to high-K calc-alkaline I-type granite. The similarities reflect a mixing of basic and acid magma. Mafic enclaves have a typical magmatic structure, which is characterized by magma mixing. The genesis of these rocks is associated with the context of subduction. They result from the mixing of a mafic magma originating from the mantle and linked to subduction, and a granitic magma (type I granite) that arises from the partial melting of the crust.

Keywords

Magma Mixing, Mafic Microgranular Enclaves, Host Granitoids, Kan Granite-Gneiss Complex, West Africa

Share and Cite:

1. Introduction

The concept of two magmas mixing was first mentioned in the 1850s [1]. This is a process that occurs when two chemically distinct magmas mix to form a hybrid magma. Mafic microgranular enclaves (MME), schlierens, minerals out of equilibrium in their host rock, bedded rocks and hybrid rocks are the main observations that attest to magma mixing [2]. These characteristic features of magma mixing are observed in intermediate and felsic calc-alkaline granitoids [3]-[8].

The first (mafic microgranular enclaves) show (i) preliminary evidence of mantle heterogeneity and (ii) the role of mafic magmas in the initiation and evolution of calc-alkaline granitoid magmas [3] [9]. Mafic microgranular enclaves are excellent choices for interpreting the history of calc-alkaline pluton genesis. The enclaves also reflect incomplete mixing, a low proportion of injected basic material and poor representation in the plutonic rock [10]. According to the work of [3] [4] [11]-[14], the disintegration of dykes in host magmas influenced by the rheology of the materials involved is also at the origin of enclaves.

Schlierens represent the second case of magmatic mixing. They are important structures for deciphering the geological history of a region. They provide valuable insights into mineral crystallization conditions and the processes underlying rock formation and their chemical composition. Schlierens can have several origins and differ from enclaves in their elongation intensity [2]. However, some schlierens originate from enclave stretching [15]. Schlierens are found in the plutonic rocks described by [3] [15] [16].

Minerals formed during the cooling of magma are sometimes sensitive to physical and chemical conditions [17] [18]. The association of quartz and olivine minerals is impossible without invoking magmatic mixing. This is why [19] support the case of magmatic mixing between rhyolite and basalt to justify the presence of quartz and olivine in dacites. The bedded rocks correspond to the juxtaposition of different magmas during reservoir recharge. Hybrid rocks are rocks in which the initial constituents are no longer identifiable [2].

In this paper, the evidence of magmatic mixing that will be studied are the mafic microgranular enclaves. These enclaves, which vary in shape and size, outcrop in the granitoids of the Kan complex. The aim is to establish the field relationships, mineralogical composition and geochemical interaction of the mafic microgranular enclaves with the Kan granite-gneiss complex. Geochemical analysis will help to identify the geotectonic environment of the Kan complex.

2. Geological Background

2.1. West African Craton

The West African craton is made up of the Reguibat Rise and the Man-Leo Rise (Figure 1(a)). These two rises are separated by the Taoudéni Basin (Figure 1(a)). The studied area is located within the Man-Leo Rise.

Figure 1. Geological maps. (a) Simplified geological map of the West African craton [34]. (b) Geological map of Kan granite-gneiss complex (modified after [29]).

The rocks comprising the Man-Leo Rise have undergone a complex geological history. The Man-Leo Rise is subdivided into two domains separated by the Sassandra shear zone. The Kénéma-Man domain located in the western part of the rise is essentially composed of Archean rocks including migmatites, gneisses, granulites and charnockites [20]-[23]. These rocks are dated between 3.2 Ga and 2.75 Ga [20]-[22]. The eastern part of the Man-Léo Ridge is occupied by the Paleoproterozoic domain. It is characterized by tholeiitic and calc-alkaline volcanics, granitoids, vocano-sediments and sediments. These rocks were emplaced between 2.25 and 2.05 Ga [24]-[26].

2.2. Kan Granite-Gneiss Complex

The geological history of the Kan granite-gneiss complex is intertwined with that of the West African craton (Figure 1(b)). The Kan granitoids were first defined by [27]. In this part of the Birimian domain, the author distinguishes three major groups composed of the Kan River Group, the volcanic group and S-type intrusive granitoids [27] [28]. The Kan complex consists of amphibolite-facies TTG, mylonitic orthogneiss, mafic-ultramafic rocks and mylonitic gneiss [6] [27] [28]. These rocks are affected by intense ductile deformation. The Kan granitoids formed after the belts of Birimian rocks [29]. The volcanic group includes clastic metavolcanites, rare lavas and volcanosediments. It lies unconformably on the Kan rocks. The western part is occupied by the Comoé basin [29] [30]. The Kan complex and the Comoé basin are separated by the Dimbokro shear corridor [28]. In recent works, the granitoids of the Kan complex are interpreted to originate from juvenile crustal growth within a Paleoproterozoic orogen at the boundary of a convergent plate [31].

3. Field Observations of Mafic Enclaves

The Kan granite-gneiss complex contains several mafic microgranular enclaves of varying sizes, shapes and lengths (Figure 2). The mafic enclaves used in this

Figure 2. Different shapes and sizes of mafic microgranular enclaves. (a) Feldspar and amphibole phenocrysts in the enclave; (b) Cluster of mafic enclaves of varying shapes; (c) Sub-circular mafic magmatic enclave in a host granitoid; (d) Elongated mafic enclave; (e) Ovoid-shaped mafic enclave; (f) Different relationships between mafic enclaves. The enclaves show a sharp contact with the host granitoids.

study come from the localities of Ahouakro, Assounvoué, Bendressou and Sakassou (Figure 1(b)). As their name suggests, these are dark rocks with fine-grained microgranular textures. However, some enclaves show phenocrysts of feldspars and amphiboles derived from the host granitoids, suggesting the intermingling of several magmas (Figure 2(a)). The enclaves encountered are sometimes grouped in a single location and take several forms (Figure 2(b)). They range in size from centimetres to decimetres (Figure 2). The enclaves observed have sub-rounded to rounded shapes (Figure 2(c)). Rounded shapes correspond to partial assimilation in the magma [32]. Others are elongated, forming spheroids (Figure 2(d)). The most common are ovoid (Figure 2(e)). Several stages in the creation of enclaves have been identified. In some outcrops, for example, a rock contains an enclave of rock B which in turn contains an enclave of rock C (Figure 2(f)). The multitude of enclaves present reflect the frequency of interactions between different magmas. The contacts between the different enclaves may be rounded or angular. These contacts are generally very sharp with no evidence of deformation.

4. Analytical Methods

4.1. Mineral Chemistry

The chemical composition of amphibole and biotite minerals in the mafic microgranular enclaves was determined using a CAMECA SX 100 microprobe at the IFREMER laboratory in Brest (France). The operating conditions are 15 kV for the voltage and 20 mA for the intensity. Spot analyses were carried out on four thin sections that were initially metallized. Biotite and amphibole standards are used in the analytical procedure for Si, Al, Fe, Mg, Ti, Cl and F. Calibration was done according to [33].

4.2. Geochemistry

To determine the state of the chemical composition, analyses were performed by the BUREAU VERITAS laboratory in Vancouver, Canada. These analyses were conducted on eight fresh rocks, including four samples of mafic enclaves and four samples of host granitoids. The concentration of major oxides was determined using X-Ray fluorescence (XRF) spectrometry. Trace element concentrations were analyzed using inductively coupled plasma-mass spectrometry (ICP-MS). Table 1 summarizes the major and trace element data for the analyzed samples.

5. Results

5.1. Samples Description

Microscopic observations reveal that the mafic enclaves sampled from different localities exhibit nearly similar mineralogical assemblages. The mineralogy is composed of amphibole, biotite, plagioclase, quartz and accessory minerals

Table 1. Major (wt%), trace and rare earth element (ppm) analyses of mafic magmatic enclaves and granitoids hosts in Kan granite-gneiss complex.

(alteration mineral and oxides).

Amphibole is the predominant mineral found in the MMEs (Figures 3(a)-(e)). These amphiboles show variations in both shape and size (Figures 3(a)-(c)). They are green in color and display significant pleochroism. Amphibole is the main ferromagnesian mineral in MMEs. The amphiboles are mainly found alongside biotite and constitute around 40% of the minerals present in the rocks (Figure 3(a), Figure 3(b), Figure 3(e)). The amphibole crystals often form aggregates up to 6 mm in size (Figure 3(b)). They are often macerated, with frequent inclusions of biotite and oxides (Figure 3(a)). Some crystals show epidote alteration

Figure 3. Representative photomicrographs of mafic microgranular enclaves. (a) Amphibole crystals of various shapes and sizes containing oxide inclusions and altered to epidote; (b) Amphibole aggregate and associated fine biotite band; (c) Elongated amphibole crystals; Plagioclase containing quartz and biotite inclusions; (d) Zoned plagioclase with core altered to sericite surrounded by amphibole often altered to epidote; (e) Clusters of quartz and sphene developing aureoles; (f) Oxide inclusions in sphene and alteration of plagioclase feldspars to sericite. Qz: Quartz; Spn: Sphene; Amp: Amphibole; Bt: Biotite; Ser: Sericite; Pl: Plagioclase; Ep: Epidote; Ox: Oxides.

(Figure 3(a), Figure 3(c), Figure 3(d)).

Biotite and amphibole form the ferromagnesian minerals of the mafic enclaves. They are frequently observed together in the thin sections. Biotite forms fine, medium-sized flakes in the microgranular matrix ((Figure 3(b)). It accounts for around 25% of the mineralogy. Inclusions of zircons and oxides are common in the biotite crystals.

Plagioclase crystals, which vary in size, account for an estimated 20% of the overall mineralogy. They frequently form subhedral phenocrysts up to 5 mm long with an average width of 2 mm (Figures 3(a)-(d), Figure 3(f)). Plagioclases are zoned with sericite-altered cores (Figure 3(d)). Quartz and biotite inclusions can also be observed (Figure 3(c)).

Quartz mineral makes up a relatively small proportion of the rock (around 10%). It forms small, clear crystals and occupies an interstitial position between plagioclase and ferromagnesian minerals (Figures 3(a)-(c)). Quartz crystals are often grouped in sub-rounded clusters (Figure 3(e)). Accessory minerals include epidote, sphene and zircon (Figure 3(a), Figures 3(d)-(f)). Sphene develops large aureoles (Figure 3(e)) and oxide inclusions (Figure 3(f)).

5.2. Mineral Chemistry

5.2.1. Amphiboles

Amphiboles are one of the main halogen minerals capable of recording magmatic events [35] [36]. Microprobe analyses of amphiboles are given in Table 2. The amphiboles originate from mafic microgranular enclaves in the samples. They have intermediate contents of SiO₂ and low contents of TiO₂ and alkalis (Na₂O + K₂O). These contents are typical of amphiboles in subduction zones [35]-[38]. The Si content (Si in formula) makes it possible to distinguish magmatic amphiboles (Si between 5.1 - 7.4) from secondary amphiboles (Si between 7.4 - 8.1). These secondary amphiboles in the mafic enclaves could be the result of a change in the chemical composition of the magmatic amphiboles during the emplacement of the host granitoids [39]. The amphiboles were projected in the classification diagram of [40]. The diagrams show that the mafic enclaves are composed of amphiboles with values (Na + K)_A ≥ 0.5 and (Na + K)_A ˂ 0.5 (Figure 4(a) & Figure 4(b)). Thus, we distinguish edenites, magnesio-hornblendes and actinolites with a wide range in the magnesio-hornblendes domain. Amphiboles are rich in Mg and depleted in Fe with Mg/(Mg + Fe) ranging from 0.51 to 0.61. However, some amphibole crystals show a high Fe content, with Fe/(Fe + Mg) ratios of 0.51.

5.2.2. Biotites

Biotites can be used to estimate the conditions of genesis of the magmatic host rocks: oxygen fugacity of the parent magmas, liquidus temperature, typological classification of granitoids, and dating of the thermal events undergone by these rocks [41]. The results of microprobe analyses of biotites are given in Table 3. These results show high concentrations of MgO (12.00 wt.% - 13.71 wt.%), low concentrations of Al₂O₃ (14.92 wt.% - 16.25 wt.%) and FeO (16.48 wt.% - 18.45 wt.%). Mg/(Mg + Fe²⁺) ratios of between 0.54 - 0.60 indicate that these are magnesio-biotites [42]. Using the MgO vs. Al₂O₃ diagram of [43], we can dissociate all the biotites according to their peraluminous, calc-alkaline or alkaline magmatic origin. Analysis of this diagram shows that the biotites in the mafic magmatic enclaves were generated by calc-alkaline magmatism (Figure 4(c)). The MgO-FeOt-Al₂O₃ ternary diagram uses biotites to distinguish igneous rocks crystallized from A, C or P magmas type [43]. Enclave biotites lie on the boundary between the realms of orogenic suites and S-type magmas (Figure 4(d)).

Table 2. Representative microphobe analyses of amphiboles from the Kan mafic microgranular enclaves.

Structural formula based on 23 oxygen atoms.

Table 3. Representative electron-microphobe analyses of biotite from Kan mafic microgranular enclaves.

Structural formula based on 23 oxygen atoms.

Figure 4. Diagrams discriminating amphiboles and biotites in the mafic microgranular enclaves. (a) Nomenclature of amphiboles from the study area in the Si vs. Mg/(Mg + Fe) diagram, with Na + K)_A ≥ 0.5 [40]; (b) Nomenclature of amphiboles from the study area in the Si vs. Mg/(Mg + Fe) diagram, with (Na + K)_A ˂ 0.5 [40]; (c) Discrimination of biotites in the MgO vs. Al₂O₃ diagram [43]; (d) Distribution of biotites in the MgO-FeOt-Al₂O₃ tectonomagmatic diagram [43].

5.3. Geochemistry Data

5.3.1. Major Element Geochemistry

The chemical compositions in SiO₂ vary from 53.5 wt.% to 64.5 wt.% for the mafic microgranular enclaves with an average of 59.06 wt.%. For the host granitoids, the SiO₂ composition ranges from 62.4 wt.% to 70.8 wt.% with an average of 68.08 wt.% (Table 1). These compositions are intermediate and intermediate to felsic respectively. The MgO (3.16 wt.% - 5.52 wt.%), MnO (0.13 wt.% - 0.16 wt.%), TiO₂ (0.44 wt.%-0.91 wt.%) and P₂O₅ (0.13 wt.% - 0.32 wt.%) contents of the MMEs are higher than those observed in the Kan host granitoids where MgO ranges from 0.9 wt.% to 2.54 wt.%; MnO from 0.04 wt.% to 0.1 wt.%; TiO₂ from 0.24 wt.% to 0.6 wt.% and P₂O₅ from 0.07 wt.% to 0.22 wt.% (Table 1). The classification of the mafic enclaves investigated and their host rocks are shown in the SiO₂ (wt.%) vs. Na₂O + K₂O (wt.%) diagram after [44]. According to this nomenclature, MMEs are composed of tonalite and monzodiorite (Figure 5(a)). The host granitoids range from tonalites to granodiorites and granites (Figure 5(a)).

The SiO₂ versus alkali-ratio diagram according to [45] shows that the microgranular mafic enclaves and their host granitoids have calc-alkaline affinities (Figure 5(b)). In detail, these rocks belong to the medium-K to high-K calc-alkaline series

Figure 5. Chemical classification diagrams of the mafic enclaves and host granitoids of the Kan complex. (a) SiO₂ (wt.%) vs. Na₂O + K₂O (wt.%) from [44]; (b) SiO₂ versus Alkali ratio diagram from [45]; (c) K₂O versus SiO₂ diagram with solid lines from [46] (dashed lines from [47]); (d) A/CNK vs. A/NK diagram from [48]. Abbreviations: FMs: Foid Monzosyenite; FMd: Foid Monzodiorite; Mz: Monzonite; MD: Monzodiorite; MG: Monzogabbro; GD: Gabbroic Diorite; PG: Peridotgabbro.

according to the diagram of [46] (Figure 5(c)). However, some enclaves show compositions that fall into the shoshonitic series (Figure 5(c)). K₂O contents vary from 1.43 wt.% to 3.34 wt.% for the MMEs and from 1.76 wt.% to 2.85 wt.% for the host granitoids (Table 1).

Al₂O₃ contents ranged from 14.20 wt.% to 17.10 wt.% for the MMEs with an average of 15.30 wt.% and from 15.20 wt.% to 16.80 wt.% for the host granitoids with an average of 15.70 wt.% (Table 1). The relatively low A/CNK values, ranging from 0.62 to 1.00, show that the MMEs and their hosts are metaluminous (Table 1, Figure 5(d)). This metaluminous character and the absence of peraluminous minerals (e.g. cordierite, sillimanite and andalusite) demonstrate that all the rocks studied are type-I granitoids (Figure 5(d)).

5.3.2. Trace Element Geochemistry

The rare earth and multi-element patterns can be seen in Figure 6. The former is the chondrites-normalized [49] and the latter is the primitive mantle-normalized [50]. Overall, the rare earth patterns all show a negative slope, with varying degrees of LREE/HREE fractionation. In detail, variations appear between the patterns of each petrographic entity.

The mafic microgranular enclaves and host granitoids show patterns moderately enriched in light rare earths with an average of 65.87 and 62.22 La_N respectively. In the mafic enclaves, the rare earths are slightly fractionated (La_N/Yb_N = 7.15 - 13.60) whereas the host granitoids show high La_N/Yb_N fractionation rates ranging from 13.21 - 30.88 (Table 1). The rare earth elements show comparatively similar and almost flat heavy rare earth profiles with an average Dy_N/Lu_N of

Figure 6. Chondrite-normalized REE patterns [49] of mafic microgranular enclaves (a) and hosts granitoids (b) and primitive mantle-normalized trace element patterns [50] of mafic microgranular enclaves (c) and hosts granitoids (d) of Kan complex.

1.36 for the mafic enclaves and 1.27 for the host granitoids (Figure 6(a), Figure 6(b)). The MMEs (Eu/Eu*= 0.80 - 0.93) show profiles with negative Eu anomalies except for one mafic enclave sample which shows a positive Eu anomaly (Eu/Eu* = 1.12) (Figure 6(a)). In granitoids, rare earth patterns do not show negative anomalies (near zero). However, two samples of host granitoids show slight positive Eu anomalies (Eu/Eu* = 1.07 - 1.13) (Figure 6(b)). These positive anomalies suggest that the plagioclases did not fractionate during magmatic differentiation and that they are present in the form of cumulates [51] [52].

The primitive mantle-normalized trace element patterns (Figure 6(c), Figure 6(d)) show that mafic microgranular enclaves and host granitoids are enriched in large ion lithophile elements (LILEs) such as Ba, K, Sr and La. MMEs and host rocks are depleted in high field strength elements (HFSEs) such as Nb, Ta, Ti and P (e.g. Late Triassic MMEs and granitoids, [53]). However, there are differences in the concentrations of trace elements. MMEs show higher concentrations of Ba up to 2572 ppm, Sr up to 1211.7 ppm, Cs up to 71.1 ppm, U compared to the host granitoids (Table 1, Figure 6(c), Figure 6(d)). The mantle compatible elements (Co, Cu, Ni and Cr) have higher contents in the mafic enclaves than in the host granitoids (Table 1).

6. Discussion

6.1. Characteristics of Magma Mixing

The granitoids of the Kan complex are composed of abundant mafic microgranular enclaves, double enclave and plagioclase megacrysts visible in both MMEs and host granitoids. These field observations indicate the role played by magmatic mixing in the generation of these granitoids [3] [53]. Another fact illustrating magma mixing process in the Kan complex is the presence of sharp boundaries in the MMEs with their granitoid hosts. The variable shapes and sizes of MMEs could suggest variations in the viscosity of magmas, and MMEs and host granitoids have distinct parent magmas [54]-[58]. The mafic enclaves have a typical magmatic structure, which is characterized by magma mixing between high-temperature basic magma and low-temperature acidic magma through injecting [59].

Geochemically, similarities (alkaline, metaluminous and type I granite affinities) were observed between MMEs and host granitoids (Figure 5). [3] explains this similarity between MMEs and hosts as the result of diffusion and percolation processes. For [56], these similarities result from chemical exchange between mafic enclaves and rock hosts during re-equilibration. The use of biotite chemistry in the ternary diagram of [57] shows that the biotites in MMEs have undergone more or less complete chemical re-equilibration by a late-magmatic hydrothermal fluid (Figure 7). Bi-element SiO₂ versus Oxides Harker diagrams show a continuous and linear distribution of points representative of MMEs and granitoid host samples (Figure 8). This distribution indicates negative correlations between SiO₂ and Al₂O₃, Fe₂O₃, CaO, MgO, MnO and TiO₂. There is a

Figure 7. Disposition of biotites in mafic microgranular enclaves in the diagram of [57].

positive correlation between SiO₂ and Na₂O, but this is less clear-cut with K₂O, which appears to be negative. The approximately linear distribution of MMEs and host granitoids suggests a magma mixing phenomenon between acidic and basic magmas [58]-[62].

However, huge variations, reflecting magma mixing, are observed when MMEs and hosts are studied in detail. Harker diagrams also show that MMEs are enriched in Fe₂O₃, CaO, MgO, MnO, K₂O and TiO₂ compared to the host granitoids (Figure 8). Conversely, alkali (Na₂O) content in host granitoids is higher than in MMEs (Figure 8). The Al₂O₃/CaO vs. Na₂O/CaO and K₂O/CaO vs. SiO₂/ CaO diagrams show linear trends of representative points of host MMEs and granitoids (Figure 9(a), Figure 9(b)). Hyperbolic mixing arrays can be observed in the CaO/Al₂O₃ vs. K₂O/MgO plot (Figure 9(c)). These two trends reflect the occurrence of magma mixing (e.g. [38] [53] [59]). In the MgO vs. FeOt diagram by [62], the major oxide content shows a distribution along the hybridization trend (Figure 9(d)).

Finally, the similarities observed in the chondrite-normalized REE patterns (LREE/HREE enrichment), confirmed in the primitive mantle-normalized patterns, imply a certain homogenization between MMEs and hosts in the trace elements during magma mixing [63] (Figure 6).

6.2. Petrogenesis and Magma Sources

The linear distribution of MMEs and host granitoids in Harker diagrams also reflects fractional crystallization processes [64] [65] (Figure 8). Furthermore, the transition elements P₂O₅, TiO₂ and Cr₂O₃ each account for less than 1% in the mafic microgranular enclaves and host granitoid and there is a negative correlation between TiO₂ and SiO₂. This is due to the effect of fractional crystallization processes from the start of magma formation with the early genesis of rutile,

Figure 8. Distribution of MMEs and host granitoids in Harker diagrams. The inserts with solid lines show the trend in host rocks and mafic enclaves.

Figure 9. Geochemical correlation diagrams for mafic microgranular enclaves and host granitoids. (d) after [62].

ilmenite and apatite minerals [66]. The low Eu and Ti anamoly reflect the crystallization of plagioclase and sphene minerals respectively. The significant negative Nb anomalies indicate the role played by titaniferous and amphibole phases [67]. The presence of hornblende explains the low Yb content of magmas [68] [69].

In magmatic differentiation where fractional crystallization is the only process, the points representing trace elements follow a curve [70]. The Rb vs. Sr and Ba vs. Sr spaces show a large dispersion of the trace element points (Figure 10(a), Figure 10(b)). This dispersion demonstrates that the relationship between the different petrographic entities is not controlled exclusively by a crystallization differentiation process, but also by partial melting (Haïmeur et al., 2004). High La_N/Yb_N ratios (up to 30.88) reflect a higher partial melting rate in the rocks and a different source [71]. The MMEs and host granitoids are metaluminous with acid to intermediate signatures and belong to type-I granite (Figure 5(d)). Type-I granites have been generated by partial melting of mafic to intermediate crustal rocks [72]-[74].

Figure 10. Plots of some pairs of trace elements. (a) Sr vs. Rb; (b) Sr vs (Ba); (c) Zr/Sm vs. Zr; (d) Ta/Tb vs. Ta.

Nb/Ta ratios range from 10.57 to 26.75 for mafic enclave samples and from 9 to 16 for host granitoids. The maximum value (26.75) is higher than that for basaltic oceanic crust, which is 17. The high field strength elements (Nb and Ta) are used in the genesis of magmas. The high ratios indicate that MMEs may represent the final basic element in the magma mixing [59]. MMEs and host granitoids show enrichment in LILEs and depletion in HFSEs, which implies that the ascent of mafic magmas was contaminated by continental crust [75].

Binary diagrams comparing the ratios of incompatible elements such as Zr vs. Zr/Sm and Ta vs. Ta/Tb can be used to investigate potential genetic relationships between magmas [76]. When the points representing the samples align along a straight line passing through the origin of the graphs, it indicates a partial melting process within the crust. Conversely, if the points align along straight lines parallel to those passing through the origin of the graphs, it reflects the evolution of magma originating from multiple sources and involving variable modalities of magma mixing or crustal contamination [77]. The analysis of the Zr vs. Zr/Sm diagram indicates that all the samples of enclaves and host granitoids are distributed almost entirely along two parallel lines that do not pass through the origin of the graph (Figure 10(c)). This suggests that the different rocks have distinct parent magmas. These magmas result from the mixing of acid and basic magmas. The Ta vs. Ta/Tb diagram shows that partial melting processes underlie the crust and, to a lesser extent, confirm multiple sources and magma mixing (Figure 10(d)).

6.3. Tectonic Implication

The average Nb/Ta ratio is 17.16 (close to the primitive mantle, 17.5 ± 2, according to [78] for the MMEs and 11.45 for the host granitoids. The enclaves therefore reflect the contribution of the lithospheric mantle (e.g. [59]). The La/Yb vs. Nb/La diagram [79] confirms the lithospheric mantle origin of the mafic enclaves, far from the average lower crust (unlike the host granitoids, average lower crust La/Yb = 30.26 and Nb/La = 0.27; Figure 11(a)). However, the MMEs field extends into the mixed lithospheric-asthenospheric mantle domain (Figure 11(a)).

Figure 11. Tectonic discrimination diagrams for MMEs and host granitoids. (a) La/Yb vs. Nb/La diagram to distinguish of the lithospheric and asthenospheric mantle (after [79]); (b) TiO₂/Al₂O₃ vs. Zr/Al₂O₃ diagram for mafic microgranular enclaves and host granitoïds (after [82]); (c) Rb/30-Ta x 3-Hf diagram (after [83]); (d) Y + Nb vs. Rb diagram [84].

The primitive mantle-normalized patterns highlight negative anomalies of Nb, Ta and Ti which are typical of magmas from subduction zones [80] [81]. This subduction zone environment is reflected in the TiO₂/Al₂O₃ vs. Zr/Al₂O₃ diagram [82] where mafic enclaves and host granitoids fall within the field of continental arcs and at the boundary of oceanic arcs (Figure 11(b)). It results from the convergence of an oceanic lithospheric plate and a continental lithospheric plate. Furthermore, the host granitoids were inserted into the Rb/30-Ta x 3-Hf and Y+Nb vs. Rb tectonic discrimination diagrams according to [83] and [84] respectively (Figure 11(c), Figure 11(d)). Due to their low Ta, Y and Nb contents, all the host granitoids fall within the fields of volcanic arc granites (VAG), thus confirming the subduction environment. The absence of Eu anomalies in some host granitoids is thought to be due to the melting of a hydrated subducted oceanic crust. This occurred within a moderately deep subduction zone that reacted with the mantle wedge at the base of the adjacent continental crust. Under such conditions, no Eu anomalies are expected to appear [67].

7. Conclusions

In this paper, the evidence of magma mixing lies in the presence of the mafic microgranular enclaves. A combination of field data, petrographic and chemical (minerals and whole-rocks) methods demonstrate the mixing of two magmas.

1) Field and petrographic data consisting of mafic enclaves provide evidence of magma mixing. This evidence includes enclaves of various shapes and sizes, sharp contact surfaces between the host and the enclaves, and the presence of phenocrysts of feldspars belonging to the host granitoids found in the mafic enclaves.

2) Chemically, the MMEs studied are composed of tonalites and monzodiorite with intermediate compositions. The host granitoids are composed of granite, granodiorite and tonalite with intermediate to acid compositions. Although embedded in granites, granodiorite, tonalite and highly mafic, the MMEs show almost similar chemical affinities (alkaline, metaluminous, type I granite) to their hosts.

3) Host mafic enclaves and granitoids result from the mixing of basic and acid magmas. They were formed by processes dominated by partial melting of the lithospheric mantle and granitic magma derived from the continental crust in a subduction context.

Further work will focus on zircon U-Pb dating to prove that the host rocks and mafic enclaves were emplaced at the same time.

Acknowledgements

This work is the result of thesis work by T.K.R. and P.K.K.J.-M. It received financial support from the T2GEM project (Geophysical and Geochemical Technologies for Mining Exploration). We would like to thank Kouamé Junior for his involvement in the production of this paper.

Conflicts of Interest

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

References