Geochemical and Petrological Insights into the Neoproterozoic Moumba Metabasites: Implications for Crustal Processes in the West Congo Belt (Republic of Congo) ()
1. Introduction
The West Congo belt is a pan-African mobile belt located in the western part of central Africa. This belt is part of a segment of the Araçauï-Congo orogen (Teixeira & Figueiredo 1991; Heilbron & Machado 2003; Pedrosa-Soares et al., 2008; Pedrosa-Soares et al., 2011) , which extends parallel to the Atlantic coast from southwest Gabon to northeast Angola over 1400 km long and 100 to 150 km wide (Dadet, 1969; Maurin, 1993) . It consists of a juvenile Neoproterozoic continental crust formed during the Pan-African orogeny, with a heritage of older Paleoproterozoic crustal rocks (Dadet, 1969; Hossie, 1980; Boudzoumou, 1986; Maurin, 1993) . This belt comprises igneous and metamorphic rocks, as well as green rocks affected by two phases of deformation (Boudzoumou, 1986) . Metabasites of the Neoproterozoic age of Mayombe belt crop out in intermediate domain. This domain consists of formations from the lower part of the Western Congo Supergroup (Fullgraf et al., 2015) . It contains basic siliciclastic rocks, volcaniclastic rocks and metasediments (Fullgraf et al., 2015) . The metabasites subject of this study represents a large formation called the Nemba Complex (Fullgraf et al., 2015) .
The study area is located in the northwest part of the Mayombe Belt, consisting of metasediments and metabasites attributed to the Nemba Complex (Bazika et al., 2022) . This study focuses only on the metabasites rocks outcropping in the Moumba River.
The origin and geodynamic setting of the Araçuaí-West Congo Orogen (AWCO) have been the subject of numerous controversies both on the African margin and the Brazilian margin, linked to various geodynamic staging contexts: 1) Vellutini et al. (1983) and Vicat & Vellutini (1987) interpret these metabasites as ophiolitic basalts, resulting from the ocean closure with subduction phenomena. This hypothesis is supported in the Brazilian margin by Pedrosa-Soares et al. (1998) and Alkmim et al. (2006) who suggest the existence of oceanic crust in the Macaúbas Basin. Pedrosa-Soares et al. (2000) and Pedrosa-Soares et al. (2008) interpret the metabasites of the Macaúbas Group in the Araçuaí belt as an ophiolitic remnants, related to a wrench tectonics in a subduction setting; 2) Tack et al. (2001) ; Fullgraf et al. (2015) ; Djama et al. (2018) and Bazika et al. (2022) reject this interpretation and consider them as intraplate basalts. Fullgraf et al. (2015) and Djama et al. (2018) suggest that these metabasites are part of intraplate basalts linked to the opening of Rodinia. Tack et al. (2001) and Bazika et al. (2022) interpret these metabasites as trap basalts (CFB) due to crustal extension. In the Brazilian margin, Cavalcante et al. (2019) also consider the metabasites of the Macaúbas Group in the Araçuaí belt as intraplate basalts. Despite the rifting episode presented by these two hypotheses, neither of them makes it possible to convincingly integrate all the petrographic and geochemical data of all the Neoproterozoic métabasites of this orogen. As a result, several questions remain: what are the nature, origin and exact geodynamic context of the metabasites of the Araçuaï-West Congo orogen? It is with the perspective of answering this question that we undertook to constrain the nature, origin, and geodynamic context of the Nemba complex of the West Congo belt from the petrographic and geochemical study of the metabasites of Moumba.
2. Geological Overview
The Araçuaí-West Congo Orogeny resulted from the collision between the Sao Francisco craton and the Congo craton (Figure 1(A), Figure 1(B)). It amalgamated during the late stages of the Neoproterozoic (Tack et al., 2001; Noce et al., 2007; Nsungani, 2012; Affaton et al., 2016) .
The West Congolian belt is constitutes a prominent topography with an NNE orientation consistent with the initial structural trend of the belt (Hossie, 1980; Boudzoumou, 1986; Djama, 1988; Maurin, 1993; Vicat & Pouclet 2000; Tack et al., 2001; Frimmel et al., 2011, Thiéblemont et al., 2018) (Figure 1(C)). This high domain generally measures less than 200 km and is followed to the east by a low and flat domain, known as the Niari-Nyanga Basin in Gabon and Congo, where it is bordered to the northeast by the basement of the Ntem-Chaillu block (Thiéblemont et al., 2018) (Figure 1(C)). This morphological differentiation also reflects a geological division between an internal domain (Mayombe) and an external domain represent by the basin Niari-Nyanga Basin.
Form lithostratigraphy point of view, the West Congo belt consists of deformed and metamorphosed igneous and sedimentary rocks (Hossie, 1980; Boudzoumou & Trompette, 1988; Maurin et al., 1991; Affaton et al., 2016; Fullgraf et al., 2015) (Figure 2), which can be subdivided into two types of formations: 1) Paleoproterozoic formations represented by the Loeme Supergroup (Fullgraf et al., 2015; Affaton et al., 2016) , corresponding to the western internal domain of the belt.
They consist of metasedimentary and reactivated granitoids during the Pan-African orogen; 2) Neoproterozoic formations composed of rift and post-rift deposits (Fullgraf et al., 2015) forming the external, eastern domain of the belt.
The Paleoproterozoic formations are characterized by a gneissic substrate overlaid by volcanic and volcano-sedimentary deposits, as well as sediments related to Neoproterozoic tectonic extension. They consist of schists, ortho- and para-gneisses intersected by the Bilinga and Bilala metagranites, of U-Pb zircon ages of 2048 ± 12 Ma and 2014 ± 56 Ma, respectively (Affaton et al., 2016; Djama et al., 1992) .
In the Brazilian segment of the Araçuaí-West Congo orogen, this domain is represented by the Paleo-mesoproterozoic Esinohaço Supergroup. This Supergroup comprises volcanics, quartz sandstones, conglomerates, pelites and subordinate carbonates (Dussin & Dussin, 1995; Uhlein et al., 1998; Martins-Neto, 2000) .
The Neoproterozoic formations are defined by a metamagmatites and metasediments substrate and sedimentary rocks. They consist of the Sounda Group, corresponding to the intermediate domain, represented by the lower part of the West Congolian Supergroup (Tack et al., 2001; Fullgraf et al., 2015) , including the Nemba Complex and the Kakamoeka and Mvouti Subgroups. The Nemba Complex is composed of metagabbros, metabasalts, amphibolites, and green schists. It is dated 915 ± 8 Ma, U-Pb zircon (Fullgraf et al., 2015) . In the DRC, its equivalent (the basic rocks of Gangila from the Zadinien Group) has been dated between 999 ± 7 Ma and 920 ± 8 Ma (Tack et al., 2001) .
The Kakamoeka Subgroup, or Mayumbian Group in the DRC, consists of volcano-sedimentary deposits, conglomerates, quartzites, graphitic schists, tuffs, pyroclastics, and mafic volcanics rocks. It has roughly the same age with the Nemba Complex (930 Ma to 910 Ma) (Tack et al., 2001; Fullgraf et al., 2015) . The Mvouti Subgroup reflects the development of clayey sedimentation of turbiditic nature, originally rich in organic matter (Fullgraf et al., 2015) , deposited in an aulacogenic basin (Boudzoumou & Trompette 1988) ; The Mayombe Group, including the Moussouva Subgroup in the intermediate domain, and the Lower Diamictite and Louila Subgroups in the external domain, representing the upper part of the Western Upper Congolian Supergroup. The Moussouva Subgroup is composed of quartzites and schists. The Lower Diamictite is composed of black clayey schists, shale, or massive rocks containing heterogeneous pebbles and cobbles (quartzite sandstones, quartz, subangular black jaspers, granites, coarse pyrite fragments), often stretched and parallel to the schistosity (Boudzoumou & Trompette, 1988; Affaton et al., 2016; Fullgraf et al., 2015; Delpomdor et al., 2019) . Locally, this congolomerate contains metric lenses of quartzites (Boudzoumou, 1986) . It is considered a glaciogenic conglomerate formed after a temporary emersion phase (Dadet, 1969; Boudzoumou & Trompette, 1988; Alvarez & Maurin, 1991) . The Louila Subgroup is composed of silty and quartzose schists, arkosic meta-sandstones, occasionally oolitic limestones and marls (Boudzoumou & Trompette, 1988; Affaton et al., 2016; Fullgraf et al., 2015; Delpomdor et al., 2019) . This subgroup is overlain by glaciomarine sediments (Upper Diamictite) (Boudzoumou & Trompette, 1988) of Marinoan age (Mickala et al., 2014) , followed by the Schisto-Calcaire Group consisting of thick carbonate platform deposits (Dadet, 1969; Delpomdor et al., 2019; Mfere et al., 2020) . The Schisto-Calcaire contains intercalations of shales and siltstones.
In the Araçuaí belt, the Neoproterozoic formations are represented by the Macaubas Group dated at approximately 1000 Ma and 850 Ma (Pedrosa-Soares et al., 2000; Uhlein et al., 1998) . The Macaubas Group consists of tillite, sandstone, pelites, and subordinate basic volcanics rocks, distal pelites with intercalations of sandy turbidites (Uhlein et al., 1998; Martins-Neto et al., 2001) and an ophiolitic assemblage. The sedimentary rocks are considered deposits into a rift evolving into an oceanic basin (Pedrosa-Soares et al., 2000) .
Structurally, the Mayombe Belt is a typically folded and thrust-belt towards the external domain. The internal domain is affected maintly by the Pan-African D2 deformation (Hossie, 1980; Boudzoumou & Trompette, 1988; Affaton et al., 2016; Le Bayon et al., 2015; Bouénitéla, 2019) , marked by a strong NW-SE-trending S2 crenulation schistosity filling the Eburnean foliation (Boudzoumou & Trompette, 1988; Bouénitéla, 2019) . Foliation is represented by the S1 flow schistosity. Locally, the S2 crenulation schistosity is overprinted by a third D3 deformation characterized by the S3 schistosity in the Loeme Group (Boudzoumou, 1986) .
The intermediate and external domains of the Mayombe Belt are affected by two main Pan-African deformation stages primarily oriented NW-SE (Hossie, 1980; Maurin et al., 1991; Boudzoumou & Trompette, 1988; Affaton et al., 2016; Le Bayon et al., 2015; Bouénitéla, 2019) . The first deformation stage (D1) is characterized by the S1 schistosity with eastward vergence and cross-cutting cleavage, affecting isoclinal folds in the intermediate domain and upright folds in the external domain. The second deformation stage (D2) is characterized by the S2 crenulation schistosity carried by upright folds in the intermediate domain. The S2 schistosity appears discreetly in the external domain.
The grade of metamorphism decreases from the internal domain to the external. In the internal domain, metamorphism is characterized by the amphibolite facies, greenschist facies in the intermediate domain, and by the anchizonal facies in the external domain.
3. Methodology
The methodology used includes: 1) fieldwork based on the systematic description and sampling of various volcanic facies outcropping along the Moumba River (Figure 1(C)). Twenty-five (25) samples were collected and seventeen (17) representative samples of fresh rock were selected for laboratory analysis and studies. The geographic coordinates of all sampling points were determined using the global positioning system; 2) petrographic study has been done on thin section made at the University of Rennes 2, France. Microscopic petrographic descriptions were carried out at the Geodynamics Laboratory of Geosciences of Marien Ngouabi University.
Geochemical analyses of major, trace, and rare earth elements were performed at the Laboratory of Petrographic and Geochemical Research (CRPG) of Nancy, France.
The samples were powdered, fused with LiBO2, diluted with HNO3, and analyzed by Inductively Coupled Plasma Mass Spectrometry (ICP-MS) and Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES). Quality control was carried out using international standards.
4. Results
4.1. Macroscopic Petrographic
The field descriptions show that the Moumba metasediments (Figure 3(A), Figure 3(B)) consist mainly of quartz schists and green schists. The quartz schists and green schists are characterized by aggregates and/or lenses of quartz, exhibiting an S2 schistosity, trending N150˚-30SW; N115˚-45SW.
They are rich in both micas and rich in chlorite, giving them a micaschist appearance. The green schists alternate with quartz schists and form schistose layers, spanning several tens of meters. These metasediments belong to the Paleoproterozoic unit that constitutes the basement of the Mayombe belt.
The Moumba metabasites are green rocks rich in plagioclase and epidote, oriented along a crenulation schistosity (S2) dipping N˚140-50 SW. They generally appear massive with a medium to fine grain. They include metagabbros, epidotites, amphibolites and chlorite schists (Figure 3(D), Figure 3(E), Figure 3(C) and Figure 3(F)).
![]()
Figure 3. Outcrops of metasediments from the Bikossi unit and metabasites from Moumba. (A) Greenschist at the edge of the Moumba River; (B) Quartz schist containing quartz aggregates in the Moumba River; (C) Fractured amphibolites; (D) Metagabbros; (E) Epidotite containing nodules of gabbro; (F) Amphibolite with stretching lineation.
The amphibolites (Figure 3(C), Figure 3(F)) are massive and locally exhibit a penetrative schistosity highlighted by biotite and amphibole. Locally, this plane carries a stretching mineral lineation (L1) dipping 40˚ towards the SSE, emphasized by hornblende porphyroblasts (Figure 3(F)). They are predominantly green, due to the abundance of amphibole and green to yellowish when they are rich in epidote.
The epidotites are greenish-yellowish epidote balls packed within the metagabbros. The metagabbros (Figure 3(D)) present, either massive appearance, with medium grain size or a foliated appearance with fine grain size.
4.2. Thin Section Description
The Thin section descriptions of the Moumba metabasites show that (Figure 4(A), Figure 4(B)) the metagabbros show a porphyroblastic texture including hornblende phenocrysts and phantom pyroxene, oriented within a matrix of plagioclase, epidote, actinolite, quartz, calcite, sphene, and iron oxides (Figure 4(A), Figure 4(B)). The primary metamorphic assemblage is characterized by Hbl + Pl + Bt, while the secondary assemblage consists of Pl (alb) + Ep + Chl ± Sp ± Op-Cal.
The amphibolites display a porphyroblastic texture (Figure 4(C), Figure 4(D)). They are composed of hornblende, chlorite flakes, epidote grains, quartz-plagiotaxis associated with plagioclases, and some relics of biotite. All these minerals are oriented. The texture is characterized by a primary assemblage of Hb + Pl + Bt and a secondary assemblage of Pl (alb) + Chl + Act + Ep ± Op. Hornblende locally appears as stretched phenocrysts along the elongation direction, giving rise to pressure shadow structures (Figure 4(C), Figure 4(D)).
The epidotites are characterized by a porphyroblastic texture. The foliation is rudimentary and defined by the orientation of actinolite needles and chlorite flakes (Figure 4(E), Figure 4(F)).
They are primarily composed of epidote, constituting the matrix in which porphyroblasts of amphibole are embedded. Epidote occurs as fine isolated grains or clusters. Actinolite appears in acicular needle-like forms associated with chlorite. Hornblende persists as epidotized relic crystals, partially replaced by the secondary phase marked by actinote. The plagioclases form the matrix on which the other minerals tapise and constitute together with hornblende the primary metamorphic assemblage of Pl + Hbl + Bt. These epidotites contain numerous veins of minerals such as chlorite, epidote, quartz, and calcite, of hydrothermal origin, characterizing thus the secondary minerals assemblage of Ep + Act + Chl ± Cal ± Op.
The chlorite schists show a lepidogranoblastic texture and contain chlorite, quartz, epidote, and iron oxide (Figure 4(G), Figure 4(H)). The schistosity is emphasized by the orientation of chlorite flakes. The overall crystal association defines a paragenesis of Pl + Chl + Ep + Qz ± Op.
4.3. Geochemistry
The results of geochemical analysis are reported in Table 1(A), Table 1(B).
![]()
Figure 4. Microscopic images of Moumba metabasites. Metagabbros [(A) Mou12, (B) Mou30]; Amphibolites [(C) Mou13, (D) Mou14]; Epidotite [(E) Mou16, (F) Mou06]; Green schists [(G) Mou01, (H) Mou01a]; Minerales: Act-actinolite, Pl-plagioclase, Ep-epidote, Chl-chlorite, Bt-biotite, Qz-quartz, Hbl-hornblende, Sp-spinel, Ap-apatite, Px-pyroxene, Op-opaque.
![]()
Table 1. (A): Analysis results of major and trace elements for samples of greenschist (01), epidotite (07). (B): Analysis’s results of major elements and trace elements of amphibolite (04) and metagabbros (05) samples.
The metabasites from Moumba exhibit a varied chemical composition, with SiO2 content ranging from 40.65% to 52.79%, Al2O3 ranging from 9.86% to 17.46%, Fe2O3 varying from 12.63% to 16.37%, MnO ranging from 0.18% to 0.30%, MgO ranging from 2.86% to 8.51%, CaO ranging from 4.87% to 18.74%, Na2O varying between 1.09% and 4.42%, K2O ranging from 0.05% to 0.22%, with low TiO2 values < 2, except for three samples (Mou3, Mou7b, and Mou23) showing TiO2 > 2, and P2O5 ranging from 0.14% to 0.67%. The sum of Fe2O3 + MgO + MnO + TiO2 varies from 18.37% to 27.54%, and Na2O + K2O from 1.15% to 4.6%. The contents of certain transition elements such as Co, Cr, Ni, and V range respectively between 24.7 and 66.4 ppm, 52.0 and 90 ppm, 66.1 and 146 ppm, and 277 and 378 ppm.
The Harker diagram of major elements vs Zr (Figure 5) shows positive linear correlations for MgO, Fe2O3, TiO2, and P2O5, whereas SiO2 shows a negative linear correlation. The concentrations of CaO, Na2O, K2O, and Al2O3 are relatively variable and exhibit a wide scatter of points.
![]()
Figure 5. Harker diagrams showing random correlations between oxides (Al2O3, CaO, K2O, P2O5 Na2O3) with Zr and linear correlations between oxides (SiO2, Fe2O3, TiO2, MgO) with Zr in the metabasite rocks from Moumba. The x-axis represents the Zr content taken as an index of magmatic differentiation. The y-axis represents the concentrations of oxides and other trace elements. The oxide contents are expressed in weight percentage.
The LILE vs Zr display a wide scatter of points and do not display any linear correlation. In contrast, transition elements such as Cr and Ni, HFSE, and REE plotted against Zr, define relatively immobile trends with positive linear correlation with the differentiation index (Zr) (Figure 6).
The diagram by Winchester & Floyd (1977) modified by Pearce (1996) (Figure 7(A)), plots all the metabasites of Moumba in the field of subalkaline basalts. The diagram by Ross & Bédard (2009) (Figure 7(B)) groups all the samples in the field of transitional basalts and shows Th/Yb and La/Yb ratios ranging respectively from 0.2 to 0.4 and from 4.40 to 6.6.
These metabasites have rare earth element (∑REE) contents ranging from 51.63 to 135.58 ppm. In the chondrite-normalized diagram by Nakamura (1977) (Figure 8), the studied samples show nearly identical and parallel patterns with a negative slope inside LREE, and flat patterns inside HREE.
![]()
Figure 6. Harker diagrams showing random correlations between Rb, Sr and Ni with Zr and linear correlations between transition elements Cr, La, Ce, Y, Nb and Ta with Zr in the metabasites of Moumba. The x-axis corresponds to the Zr content taken as an index of magmatic differentiation. The y-axis corresponds to the concentrations of oxides and other trace elements. Trace elements are expressed in ppm.
![]()
Figure 8. Chondrite normalized rare earth elements (REE) patterns of the Moumba metabasites Values used are those of Nakamura (1977) . The metabasites exhibit a negative trend enriched in LREE/HREE with positive Eu anomalies.
They are significantly enriched in LREE compared to HREE. The different fractionation ratios are high: (La/Yb) N = 2.97 - 4.49 and (La/Sm) N = 1.94 - 2.28. The REE patterns show positive Eu anomalies, except the sample MO28 which exhibits a negative Eu anomaly (Eu*/Eu = 0.95).
The Moumba patterns normalized to the primitive mantle of McDonough & Sun (1995) (Figure 9) are characterized by the absence of negative anomalies in Nb-Ta and Ti. They are enriched in LILE and show a low fractionation rate in HFSE, strong positive anomalies in Ba, Sr, Pb, and negative anomalies in Rb.
The projection onto the Zr vs Zr/Y diagram by Pearce & Norry (1979) (Figure 10(B)) and Yb vs. Th/Ta diagram by Schandl & Gorton (2002) (Figure 10(C)) plots the majority of samples in the within-plate basalts field, except two samples located outside the distinguished fields.
![]()
Figure 9. Trace-element diagram of Moumba metabasites normalized to the primitive mantle of McDonough & Sun (1995) . All metabasites exhibit positive Pb anomalies.
![]()
Figure 10. Discrimination diagrams for trace elements of the Moumba metabasites. (A) Zr/4, Ti/100, and Y*3 triangular diagram Pearce and Cann (1973) , the metabasite samples tend towards within-plate continental alkali basalts, but two metagabbros show tholeiitic affinity of island arcs. (B) Zr vs. Zr/Y Pearce & Norry (1979) . (C) Ta/Hf vs. Th/Hf Schandl & Gorton (2002) . (D) (Th/Ta)N vs. (Tb/Ta)N Thiéblemont et al. (1994) . The metabasites plots in the within-plate domain. Abbreviations: E-MORB: enriched mid-ocean ridge basalt, N-MORB: normal to mid-ocean ridge basalt, CAB: volcanic arc basalt, WPB: within-plate basalt; WPAB: within-plate alkali basalt, BAB: back-arc basin basalts; IAT: island arc tholeiite, CFB: continental flood basalt.
In the Th/Ta and Tb/Ta diagram by Thiéblemont et al. (1994) (Figure 10(D)), the metabasites fall in the CFB field, whereas in the Zr/4-2Nb-Y diagram by Pearce and Cann (1973) (Figure 10(A)), they are distributed in the within-plate lavas field, except two samples Mou23 and Mou01, which fall into the field of island arc tholeiites, MORB, and continental alkali basalts.
5. Discussion
5.1. Petrochemical Characteristic
The results of the petrographic study reveal that the study area is characterized by a well-preserved upper crustal volcano-felsic assemblage represented by interstratified metabasites in metasediments. The coexistence of Hb + Ep + Pl + Bt in the metabasites indicates that the rocks of Moumba underwent lower amphibolite grade of metamorphism. The secondary paragenesis Pl (Alb) + Act + Chl + Ep ± Cal ± Op reflects a retromorphose towards greenschist grade. The Moumba metabasites are affiliated with the Nemba Complex dated between 915 Ma and 1000 Ma (Le Bayon et al., 2015; Fullgraf et al., 2015) . The heterogeneity of textures and rocks nature suggest a gabbroic or doleritic protolith.
The geochemical characteristics show that Moumba metabasites are mainly composed of transitional subalkaline basalts poor in TiO2 (Low-Ti) and MgO with a strong enrichment in Fe2O3 (t) (13.95% - 16.37%). The fairly significant variations in alkali contents Na2O (1.09% - 4.42%) and the high values of Al2O3 suggest extensive feldspar albitization (Talbi et al., 2005) .
5.2. Tectonic Discrimination
The content of immobile HFSE elements used to constrain the tectonic setting of volcanic protoliths, discriminates the Moumba metabasites in an intraplate context. High Y and Zr contents, along with high Zr/Y ratios > 1.5, characterize, according to Pearce & Norry (1979) ; Pearce (1983) , an intraplate continental basalt environment of the CFB type (Figure 10). Similarly, the values of the La/Sm and Gd/Yb ratios of REE normalized to N-MORB vary from 2 to 23 and from 0.5 to 3.5, respectively. They are similar to those reported for Deccan-type basalts by Lightfoot et al. (1990) ; Mahoney et al. (2000) . Furthermore, the absence of negative anomalies of Nb, Ta, and Ti (Figure 9), characteristic of a subduction or arc environment, excludes the emplacement of Moumba metabasites in the context of an active margin related to a subduction zone forming volcanic arcs. Instead, it suggests emplacement in a passive margin setting.
The diagram of Pearce & Cann (1973) (Figure 10(A)) suggests a continental arc character with enrichment in MORB. The arc signature of these rocks may suggest crustal involvement and a decrease in mantle melting degrees (Wilson, 1989) . These characteristics are in line with the interpretation suggested by Bazika et al. (2022) for the equivalents basic roks of Loukounga in the southern part of Mbena and by for the Gangila basalts of the Zadinian Group in the D.R. Congo and dated between 930 Ma and 920 Ma Tack et al. (2001) . Our results also conform to the Tack et al. (2001) model proposed by Cavalcante et al. (2019) for basalts of the Macaùbas group in the Brazilian margin in the Araçuaí belt. Cavalcante et al. (2019) question the possible existence of oceanic crust. According to Cavalcante et al. (2019) , the model involving about 50 Ma of subduction of oceanic crust and the development of the related arc cannot be compatible with the emplacement of Neoproterozoic magmatic rocks in the Araçuaí-West Congo Belt. They propose an intracontinental hot orogeny model, where significant melting of the middle crust could have caused the propagation of the upper crust in an orogenic setting created by collisions along the N, W, and S margins of the São Francisco craton from 630 Ma onward.
In conclusion, this magmatism reflects heterogeneity in the composition of the lithospheric source and probably different rates of partial mantle melting. Despite the calc-alkaline nature (Boudzoumou, 1986; Djama et al., 1988; Fullgraf et al., 2015; Djama et al., 2018) of the acid-basic rocks emplaced during this period, it does not seem that a genetic link with oceanic crust subduction can be invoked. The Neoproterozoic paleogeography (Rodinia) is marked by intracontinental rifts (Kampunzu & Cailteux, 1999; Trompette, 2000) , indicating an intraplate continental environment. The calc-alkaline signature may be due to partial melting in an extensional setting of a metasomatized mantle (McCulloch & Gamble, 1991) during an ancient, presumably Eburnean subduction (Piqué et al, 1998) . Thus, the volcanism of the metabasites of Moumba was effectively established in a context of continental flood basalts (CFB).
5.3. Petrogenesis
5.3.1. Fractional Crystallization
The trends of major and trace elements in the Harker diagrams (Figure 5, Figure 6) indicate that positive linear correlations may be related to the fractionation of olivine, pyroxene, and plagioclase. This fractionation is consistent with fractional crystallization and forms, for the Moumba metabasites, a continuous cogenetic evolutionary trend similar to those observed in continental basalts (Bellieni et al., 1984; Piccirillo et al., 1988; Marques et al., 1999; Rocha-Júnior et al., 2013) . However, some scattering of points indicates a random evolution. This suggests that, simple binary mixing mechanisms must have played a role in the origin of differentiation lineages. The positive anomalies in Eu (Figure 8) in the majority of samples suggest an accumulation of plagioclase in the magma and the presence of garnet in the residue.
5.3.2. Crustal Contamination
Some chemical characteristics of volcanic rocks can be acquired through interactions with crustal rocks during the transfer of magma from the mantle to the surface and also during the storage of these magmas in crustal magma chambers at different levels and their differentiation through crystalline fractionation (Kamgang et al., 2008) . These interactions have already been observed and characterized by Bazika et al. (2022) in the Loukounga metabasites.
In the Zr/-Ti/100-3*Y diagram (Pearce & Cann, 1973) (Figure 10), the random correlations of concentrations of certain major elements (CaO, Na2O, K2O, and Al2O3) and trace elements (Rb, Sr) seem to indicate possible crustal contamination of the magma. The absence of Nb-Ta anomalies in the chemical composition of the igneous rocks may, suggest that they have not been affected by crustal contamination and/or its effect is negligible (Nagudi & Coll, 2003; Fatehi & Ahmadipour, 2018) . However, Frey et al. (2002) , Wooden et al. (1993) , Xia & Li (2019) suggest that high ratios of (Th/Ta) > 1.5, (La/Ta) > 22, and low ratios of (Th/La) < 1, (Nb/La) < 1 would indicate crustal contamination of the magma. The majority of metabasites show high ratios of (Th/Ta) > 1.5, (La/Ta) > 22, and low Th/La, Nb/La ratios, respectively less than 1. Figure 11 shows positive correlations between Zr vs. Nb and (Sm/La)N vs. (Gd/Yb)N, indicating clearly that the chemical variations observed in the metabasites of Moumba were mainly acquired through interaction with crustal rocks during the differentiation of the magma.
![]()
Figure 11. Crustal contamination discrimination diagrams. All diagrams indicate the contribution of crustal contamination to the metabasites; (A) (Sm/Yb) vs (Th/Nb) diagram from Wang et al. (2007) normalized to the primitive mantle of Sun & McDonough (1989) ; (B) Zr/Yb vs Zr diagrams of the Moumba metabasites. D) La/Nb vs Zr diagram of the metabasites. (C) Nb vs Zr diagram of the metabasites. The indicative trends of fractional crystallization and crustal contamination confirm that the basic magmas of Moumba interacted with continental crustal rocks.
The Zr/Yb and La/Nb ratios are also well correlated with Zr contents, compatible with contamination by a high La/Nb component, typical of continental crustal rocks (Figure 11).
This crustal component also has a very high Pb content. This process of contamination by a crustal component with a high Pb content has already been observed in other trap basalts (Mahoney et al., 2000; Beard et al., 2017) . This indicates that the Moumba metabasites are contaminated by a crustal component.
5.3.3. Mantle Source Nature
The ratios of immobile trace elements such as Th and Yb are independent of mineral sources, the degree of partial melting, and fractional crystallization, meaning that their ratios reflect the mantle source (Aldanmaz et al., 2000; Rolland et al., 2000) . In the Nb/Yb vs. Th/Yb diagram (Pearce, 1983; Pearce, 2008) (Figure 12(A)) and Nb/Th vs. Zr/Nb (Condie, 2005) (Figure 12(D)), the metabasites fall into the field of mantle-derived fusion enriched parallel to the intraplate enrichment trend (W trend).
In the Zr/Y vs. Nb/Y diagram (Fitton et al., 1997) (Figure 12(B)), the Moumba metabasites trace in the enriched MORB field. This configuration suggests that
these metabasites originate from an enriched mantle plume. The Ti/Y and Nb/Y ratios in intraplate basalts are higher than in other basalts. These differences indicate that the mantle source of these rocks is more enriched compared to MORB and volcanic arc basalt sources (Rollinson, 1996) . High Ti/Y and Nb/Y ratios in the studied metabasites suggest an enriched mantle source of their primitive magmas. According to Sun & McDonough (1989) , Zr/Y ratios > 2.46 and Zr/Nb ratios < 15.71 separate enriched sources from depleted sources in basalts. The Zr/Y ratio (3.76 to 6.44) and Zr/Nb ratio (10.46 to 19.32) in the metabasites indicate their derivation from an enriched mantle source. On the Zr vs. Y diagram (Coban & Flower, 2007) (Figure 12(C)), the metabasites fall into the enriched field. This configuration suggests that the metabasites of Moumba derive from an enriched mantle plume head.
5.3.4. Condition and Degree of Partial Melting
The gradient of REE spectra in any suite of igneous rocks helps to understand the depth and degree of melting conditions (Cullers & Graf, 1984; Fram & Lesher, 1993; Hirschmann et al., 1998; Patel et al., 2021) . The REE abundances
![]()
Figure 13. Discrimination diagram to determine the conditions and degree of magma partial melting. (A) Sm Vs (Tb/Yb)PM, (B) Al2O3/TiO2 vs (Sm/Yb)PM, (C) Gd/Yb vs. LA/Sm (Wang et al., 2007) ; the Moumba metabasites fall into deep low-degree melting and shallow high-degree melting; (D) Sm VS; Sm/Yb (Aldanmaz et al., 2000) , all metabasites fall on the curve of partial melting ranging from 10% to 30%, containing 10% to 30% lherzolite with spinel and garnet.
normalized to chondrites define the presence or absence of residual garnet in the source region (Jung, 2000; Patel et al., 2021; Hirschmann et al., 1998; Green, 1994) . The REE spectra of Moumba are enriched in LREE and inclined, showing a negative trend (Figure 8). They have relatively high (Gd/Yb)PM ratios (1.44 to 2) and (Tb/Yb)PM ratios (1.10 to 1.53). This indicates the presence of residual garnet during partial melting (Green, 1994) . Since garnet retains HREE from the mantle during partial melting, YbN values < 10 (4.24 and 8.91) in the metabasites suggest, according to Wilson (1989) ; Oliveros et al. (2007) , the presence of garnet as a residual phase in the source region. Diagrams Sm vs. (Tb/Yb)PM, Al2O3/TiO2 vs (Sm/Yb)PM, and Gd/Yb vs. La/Sm by (Wang et al., 2007) (Figures 13 (A)-(C)) shows that the Moumba metabasites fall between the fields of deep mantle low-degree melting and shallow mantle high-degree melting fields.
However, the relatively moderate Al2O3/TiO2 ratios (4.72 - 11.42) and high (Sm/Yb)PM ratios (1.37 - 2.33) indicate lower degrees of partial melting from a relatively shallower mantle source (Figure 13). In the Sm/Yb vs Sm diagram (Figure 13(D)), these metabasites plot on the lherzolite - spinel + garnet curve, confirming thus the presence of garnet in the source region, and hence, the magma likely resulted from a shallower partial melting of the mantle containing 50% spinel and 50% regolith. It is highly probable that the Moumba metabasites were generated from a significantly enriched mantle, which began melting in the garnet facies stability zone and continued melting in the spinel facies stability zone. The degree of partial melting, following the parameters of Aldanmaz et al. (2000) (Figure 13), was probably less than 30% (Figure 13(D)).
6. Conclusion
This study, which aims to constrain the geochemical signature of the Nemba complex of the West Congo belt, constitutes a contribution to the reconstruction of the geodynamic context of the metabasites of the Pan-African Araçuaï-West Congo belt. It shows that Moumba metabasites correspond to Nemba complex of West Congo Belt and are made up of amphibolites, metagabbros, epidotites and green schists interstratified in the Eburnean metasediments. The geochemical signature highlights continental flood basalts (CFB), emplaced in an intraplate geodynamic context from magma coming from an enriched mantle plume having undergone partial melting of low degrees. This magma was subsequently contaminated by continental lithospheric during their ascent. This present study calls into question the possible existence of oceanic crust and the presence of the ophiolite complex in the West Congo belt. We support the intracontinental model for the formation of the West Congo belt. Our results suggest that the West Congo belt was formed following the closure of a continental rift due to continental collision. This hypothesis is in agreement with recent data obtained throughout the Araguaï-West Congo Belt.
The understanding would only be better if we added isotope, microprobe and structural analyses to the protocol to better understand the geochemical signature, the P-T-t paths and the markers of Pan-African tectongenesis of the metabasites of the Nemba complex and the associated lithological units. These additional analyses will make it possible to precisely establish correlations on the scale of the West-Congolian Belt, and also on the scale of the Araçauï-West Congo orogen.