Abstract

The Sanaga prospect in the north of Edea is located in the upper Nyong unit of the Ntem complex in Cameroon. The objective of this study is to use geochemical data trends for major and some trace elements to constrain the origin and/or sources of various constituents in the iron-bearing units as well as assess their economic potentials. The rock samples were collected from a single drill core sampled at various depths. Major elements were analysed using X-ray fluorescence spectrometry after powder digestion following. All data were processed with the aid of XLSTAT. The stratigraphic log described revealed from top to bottom two lithological sequences composed of oxidized formations (oxidized cap and oxidized gneiss), and gneissic formations (magnetite gneiss, magnetite amphibolite gneiss and enriched magnetite amphibolite gneiss successions). Detailed examination showed that quartz and iron oxides are the main minerals present. Bulk geochemical analysis of the oxidized and gneissic formations showed that Fe₂O₃ and SiO₂ are the main constituents (averaging 84.40 wt % and 92.54 wt %, respectively), confirming that quartz and iron oxides are the major mineral phases in both the oxidised and gneissic formations. Al₂O₃ averages 9.34 wt % and 3.06 wt %, Na₂O averages 0.04 wt % and 0.59 wt %, K₂O averages 0.26 and 0.53 wt %, and P₂O₅ 0.07 and 0.05 wt %, respectively, in both oxidized and gneissic formations. Concentrations of trace elements in the various lithologies are generally very low (< 100 ppm). Certain correlations of interest in both units include Al₂O₃ with LOI (r > 0.8), and Zr (r > 0.7); LOI with Zr (r > 0.8). From these data it appears that mineralisation at the Sanaga prospect is restricted to the magnetite gneiss. The high concentration of Al₂O₃ (average 9.34 wt %) in the oxidized iron formations is partially due to its introduction during recent chemical weathering. The Sanaga iron formations are metamorphosed chemical sediments formed by precipitation of iron and silica from a mixture of seawater and hydrothermal fluids with a significant terrigenous input.

1. Introduction

There has been extensive investigations on the distribution, stratigraphy and origin of the iron-formations worldwide and in Cameroon ^{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12}. Most of these investigations aim at determinate the origin of iron formations in order to obtain a large data set that can justify their genetic models. Understanding the genesis of iron ore deposits has traditionally been carried out using geochemistry of either: major/trace elements, rare earth elements, fluid inclusion microthermometry, solute chemistry, stable isotopes, radiogenic isotopes, organic matter, active gas chemistry and noble gas isotopes ^{13, 14, 15, 16, 17, 18}. Several geochemical data are present in the literature, and still many factors are detracting from analytical values as contribution to comprehensive understanding of the composition of these rocks despite all these. It is therefore still leading to inadequately knowledge on the chemical composition of iron formations. According to ¹⁹ chemical analysis of iron formations fall into three main categories: (1) a general petrographic and chemical characterization; (2) determination of ore potential; and (3) understanding of origin and evolution.

In this paper, we provide new macroscopic description and geochemical data of iron-bearing mineralisation in a stratigraphic sequence in South Cameroon with the focus on deciphering the geochemical trends for major and trace elements, which may lead to some constraints about the origin and/or sources of chemical components in the iron formations. Furthermore, the chemical composition of this iron formations are compared with others world known deposits to assess its quality and evaluate the iron potential for eventual exploitation.

2. Geological Overview

The Sanaga iron ore prospect is located at the extreme end of the Nyong unit within the Archean Ntem complex, which is at the north-western end of the Congo Craton (Figure 1). The Nyong Unit constitutes the main Paleoproterozoic unit in Cameroon ^{20, 21}. Deformation in the Ntem complex can be divided into two major events: (1) the Archean event marked by the successive diapiric emplacement of Mesoarchean charnockites and TTGs and (2) a Paleoproterozoic transcurrent deformation phase marked by the expansion of N-S to NE-SW trending shear zone and partial melting of the TTG and the greenstone belt ^{22, 23}. The complex is made up of three units namely: Ayina unit, Ntem Unit and the Nyong unit were the study area belong.

The lithologies that are encountered in the Nyong unit include: metamorphosed mafic–ultramafic rocks, expressed as pyroxenites and amphibolites ²⁴, orthogneisses and metaquartzites with a dominance of biotite–amphibole–hornblende–quartz gneiss, pyroxene–amphibole–garnet–gneiss and magnetite-bearing gneiss ²⁵, and the main plutonic rocks made of syn- to late-tectonic charnockites, alkaline syenites and post-tectonic metadiorites ^{20, 21}.

Gneisses in the unit have undergone Pan-African rejuvenation as supported by the presence of similar rocks in the Pan-African mobile belt north of the Congo Craton ²⁰. The biotite-hornblende gneisses are of TTG composition ²⁶ and are spatially related to the orthopyroxene-garnet gneisses (charnockites), garnet-amphibole, pyroxenites and BIF that have been intruded and deformed by the metagranitoids. Magnetite gneiss formations occurrence are lenticular and often appear as quartz–magnetite gneiss, biotite–quartz–magnetite gneiss, biotite–amphibole–magnetite gneiss, biotite–amphibole–magnetite-garnet gneiss and pyroxene–garnet–magnetite gneiss with a characteristic pinch and swell morphology ²⁷.

3. Sampling and Analytical Methods

Chemical sampling of one selected diamond drill hole was done following the various lithologies of iron bearing rocks and a total of 34 samples were taken. The sample core length was taken below 1m at top hole up to 33m, then above 1 m until bottom hole after this depth depending on the lithology. Samples were selected from a diamond core and halved using a core cutting machine, and then quartered at the Cameroon Mining Company (CMC). The sample preparation was completed with drying, crushing, splitting and pulverizing at ALS Laboratory in Yaoundé Cameroon. Furthermore, major and selected trace element (As, Ba, Co, Cr, Cu, Ni, P, Pb, S, V, Zn and Zr) concentration were analysed using the lithium borate fusion technique, coupled with X-ray fluorescence spectrometer with a relatively standard deviation (RSD) of 5% at the ALS laboratory of Johannesburg in South Africa. Loss on ignition (LOI) was determined by weight difference after ignition at 1000°C. Certified reference material, duplicate and quartz blanks were randomly inserted into the sample batches as part of a continuous sample number to ensure quality assurance and quality control.

Chemical data processing was done using XLSTAT and permit to obtained line formulas, summary statistics and correlation coefficient of Patterson for the two lithological sequences.

4. Results

Stratigraphic log sequence of the selected diamond hole is summarized in Table 1. The log described showed from top to bottom two lithological sequence composed of: oxidized formations (oxidized cap and oxidized gneiss), and gneissic formations (magnetite gneiss, magnetite amphibolite gneiss and enrich magnetite amphibolite gneiss successions), (Figure 2). The oxidized cap is the near surface enriched mineralization. They are located at top of the stratigraphic sequence (Figure 2 and Table 1), and composed of the oxidized cap and oxidized gneiss. The mineralogy is mainly composed of iron oxides (magnetite, goethite, and hematite), some quartz and rare clay.

The gneiss formations presented a characteristic pinch and swell morphology, and the transition between the various units is either sharp or gradational. Magnetite is the main iron oxide in the gneissic formation. The magnetite gneiss bands are sometimes observed at different stratigraphic locations with quartz lodes intruded within the gneissic assemblage. Massive magnetite vein-like features are observed toward bottom log in the magnetite gneiss formation.

Whole-rock chemical analyses of all the facies were carried out in one diamond drill hole of 60 m depth in quarter core representing the iron bearing lithological types formed by the oxidized and the gneissic formations. From this it appear clearly that Fe and Si are the main components with 88 wt % of the rock (averaging 42.13 wt % and 45.63 wt % respectively) (Table 2).

Oxidized formations

Bulk analyses indicated that Fe percentage ranges between 6.96 and 58.54 wt % (average 40.40 wt %), while silica varies between 27.60 and 65.80 wt % (average 44.02 wt %). The low Fe is due to the abundance of quartz veins cross-cutting some sections. Al₂O₃ percentage is between 1.08 and 25.40 wt % (average 9.34 wt %), while LOI varies between from 0.38 - 14.54 wt % (Table 3). The contain in this two element reduced with depth and could be indicative of the presence of clays minerals that equally reduce with depth. Again, average TiO₂ is 0.49 wt % with some value greater than 2 wt %, while phosphorous is between 0.02 – 0.28 wt % (average 0.7 wt %). A relative elevated Al and Ti contents in the sequences suggest trace input of detrital materials ²⁸. The alkali contents (Na₂O and K₂O) are significantly low averaging 0.04 and 0.26 wt % respectively. MgO and CaO varies between 0.02 – 1.29 wt % (average 0.19 wt %) and 0.005 – 0.83 wt % (average 0.07 wt %) respectively. The concentration of these elements increase with depth. In general, the oxidized formations are marked by low concentration of major elements apart from iron, silica and alumina. The bottom unit is marked by a slight enrichment in MgO, Na₂O and CaO.

Gneissic formations

Fe₂O₃ and SiO₂ are the main constituent (average 92.54 wt %) with values ranging between 23.32 – 55.40 wt % and 41.90 – 59.10 wt % respectively (Table 3). Al₂O₃ range between 0.48 and 8.69 wt % (average is 3.06 wt %). Na₂O and K₂O respectively range between 0.07 – 2.26 wt % and 0.01 – 2.00 wt %. Phosphorous content is very low varying from 0.04 – 0.07 wt % (average 0.05 wt %). MgO and CaO increase in the gneissic formation compared to the oxidized formations with average value of 1.89 and 1.56 wt % respectively. This enrichment in MgO and CaO with depth associated to the absence of LOI (negative values) is due to the presence of amphibole minerals and facies change. The other major elements display a very low concentration (Table 2). Again, the sharp transition between magnetite gneiss and amphibolite gneiss is marked by an increase in MgO, and CaO.

The concentration of trace elements in the various lithologies is generally very low (< 100 ppm). Compared to major element chemistry that clearly differentiates each lithotypes, trace element concentrations are not as clearly indicative of a particular lithology (Table 4). Oxidized formations are more depleted in trace elements than gneissic. The latter is slightly enrich in Cl (ranging between 0.01 – 35.00 ppm), Sr, (ranging between 3.00 – 63 ppm in amphibole bearing gneiss) and Ba (ranging between 0.01 – 91.00 ppm), while the former has higher contents in Zr (ranging between 7.00 – 43 ppm) and V (ranging between 2.00 – 21.00 ppm).

Oxidized formations are more depleted in trace elements than gneissic. The latter is slightly enrich in Cl, Sr, and Ba, while the former has higher contents in Zr and V.

Mineral Inter-element correlations matrix of the oxidized cap and gneissic formations are given in Table 5 and Table 6 respectively. Besides correlation characteristic similar in the two formations of the lithological sequence (e.g., Al₂O₃ with LOI (r > 0.8), and Zr (r > 0.7); LOI with Zr (r > 0.8), and S with V (r > 0.8)), there are significant correlations specific for the oxidized formations (e.g., Al₂O₃ and TiO₂ (r > 0.7); MgO with MnO (r > 0.7), Ba (r > 0.8), and Cl (r > 0.8); TiO₂ with P₂O₅ (r > 0.9), LOI (r > 0.8), Cr (r > 0.7), S (r > 0.8), V (r > 0.9), and Zr (r > 0.8); LOI with S (r > 0.7), and V (r > 0.8); and S with Zr (r > 0.7)), and the gneissic formations (e.g., Al₂O₃ with SiO₂ (r > 0.9), Na₂O (r > 0.8), Ba (r > 0.8), K₂O (r > 0.8) and Sr (r > 0.7); SiO₂ with Na₂O (r > 0.8), Ba (r > 0.8), K₂O (r > 0.8) and Sr (r > 0.7); CaO with TiO₂ (r > 0.7), and MnO (r > 0.7)). These correlation are indicative of a terrigenous contribution during the BIF deposition ^{17, 29, 30}. The significant positive correlation between Al₂O₃ and TiO₂ (r > 0.7) could be inherent from the common association of Ti and Al in clay minerals from residual weathering.

Beside this positive correlation there is a negative correlation between SiO₂ and Fe₂O₃ (r > 0.5 in both sequences), suggesting the incorporation of Fe₂O₃ and SiO₂ in different mineral phases and inversely reflected the variation of one component to another.

5. Discussion

Many iron occurences are found within the upper Nyong serie in the Ntem complex. Ore bearing rock varies from BIF Koumbo and Elom area to gneissic formations at Ngovayang ^{3, 4, 27}. Iron bearing mineralisation from macroscopic description showed that the mineralisation in the Sanaga prospect is carried by magnetite and magnetite amphibolite gneiss. This is in accordance with data obtained by ²⁷. This fact could permit a bigger comparison in their genesis. However, detail petrographic data are currently prepared for a clear comparison between the Sanaga iron formations and those of the Ngovayang ridge.

Al₂O₃ is enriched at the top of the sequence compared to bottom. Enrichment in alumina has been considered to result from three main sources: (a) immobile behaviour and residual enrichment of Al₂O₃, (b) introduction of Al₂O₃ during the ore-forming event; (3) introduction of Al₂O₃ during geological recent chemical weathering processes, possibly from adjacent shale unit ³¹. Simple residual enrichment of alumina would necessitate very considerable removal of iron during ore formation, rather than it effective residual enrichment as observed elsewhere ³². There is no considerable removal of Fe₂O₃ at the top profile in the oxidized formation when compared to the gneissic formation, and the concentration of Al₂O₃ (average 9.34 %) instead suggest it introduction during recent chemical weathering. The high LOI obtained (with a maximum of 14.54, Table 2) is in favour of the conclusion.

Elevated Al₂O₃ and TiO₂ are often type to be indicator of detrital input in iron formations ³³. Likewise, elevated Al₂O₃ and TiO₂ MgO, CaO and K₂O are regularly present in sediments. The geochemical sequences of gneissic formations showed a relatively high amount of Al₂O₃ MgO and TiO₂ ranging between 0.48 – 8.69 %, and 0.04- 2.05 % respectively. Equally MgO (0.63 – 2.93 %), CaO (0.87 – 3.88 %), K₂O (0.01 – 2.50 %) and Na₂O (0.07 – 2.26 %) showed an increase. Al₂O₃ and TiO₂ are generally used to trace clastic input ^{28, 34}. It has been suggested that, high content of Al₂O₃ together with high amounts of TiO₂ are evidence of the supply of primary materials from deeply weathered source rocks ³⁵. The incorporation of terrigenous sediments in gneissic formations of the Sanaga prospect is suggested by the strong correlation between Al₂O₃ with TiO₂, and Zr, (Table 4), as Zr is an element considered as a detrital component ³⁶. The relatively low content of these elements indicated a low degree of contamination ¹⁷.

Even with the huge amount of data concerning the origin of Fe and Si in iron formations, controversies remain very high. The majority of writers agree that iron formations are chemical sediments formed by precipitation of iron and silica from solutions consisting of a mixture of seawater with hydrothermal fluids. The main impurities are terrigenous sediments carried by rivers or winds, or deposited by volcanic activity ^{37, 38, 39}; or pelagic sediments ⁴⁰. Most workers consider that Fe and Si components are derived from the leaching of basalt and komatiites of the ocean floor by hydrothermal fluids ^{12, 29, 41, 42, 43}. Only a few authors attribute the origin of Fe and Si to the weathering of continental rocks ^{44, 45, 46}. Proposed methods for distinguishing between seawater, hydrothermal, biogenic and detrital sources are based on differences in the mineralogical, chemical and isotopic composition.

Again ⁴⁷ suggested that hydrothermal metal-rich deposits could be distinguished from hydrogenous deposits formed by diagenetic processes on the basis of the relative abundance of SiO₂ and Al₂O₃. Due to the fact that the top sequences were weathered, data plots were taken for the gneissic formation from 20 m depth in fresh samples. SiO₂ and Al₂O₃ diagram (Figure 3) plotting of study samples showed that, they fall into the hydrothermal and hydrogenous zone. Pure hydrothermal deposits contain very little Al and have high Al/Ti ratios ⁴⁸. Contamination of such deposits by pelagic and terrigenous deep-sea sediments enriches them in components such as Ti and Al, resulting drastic lowering of the Fe/Ti ratios and increase in the Al/(Al + Fe + Mn) ratio. On an Fe/Ti vs. Al/(Al + Fe + Mn) diagram of ⁴⁹ , most of the studied core samples cluster on the East Pacific Rise hydrothermal deposits (Figure 4). Furthermore, the Si/Al ratio is commonly used to detect eventual hydrogenous, respectively hydrothermal material supply. It is worth underlining here that the Si/Al ratio has been used to distinguish between hydrothermal Fe–Mn crusts, rich in Si and characterized by a Si/Al ratio >5.1, and hydrogeneous Fe-deposits (nodules), whose typical Si/Al ratio is 3, the same as marine sediments. The average Si/Al ratio of the studied iron formation is 21.17 pointing to the hydrothermal origin of the studied iron formation (Table 2).

However, some samples fall into the hydrogenous field (Figure 3) suggesting chemical precipitation from seawater ⁵⁰. The seawater is equally suggested by the low CaO/(CaO+MgO) ration (between 0.17 – 0.70) indicative of sediments that precipitated from seawater.

Additionally, it is known that Co/Zn ratio is used as tracer of hydrothermal origin ⁵⁶. According to this author, low Co/Zn ratio (0.15) characterizes hydrothermal deposits whereas hydrogenous deposits display high Co/Zn ratio (2.5). The average Co/Zn ratio of the studied iron ores samples is 0.04, which is consistent with trace metals derived from a largely hydrothermal source ⁵¹.

Major elements diagram plotted of the Sanaga iron formations with that of average Agloma and Lake Superior showed a similar mineralogy with a slight enrichment of iron and silica at from ore of the Sanaga prospect (Figure 5). This diagram shows that the ore have an interesting potential for exploitation.

The potential of iron ore mineralisation is generally viewed using the iron percentage, the deleterious elements encountered such as P₂O₅ and S, Al₂O₃, and SiO₂. The deleterious effects of such elements in the production process and metallurgical properties of steel have long been recognized ^{52, 53}. For that reason ⁵⁴ divided iron ore in three main category base on the Fe percentage: (1) high grade iron ores with iron percentage > than 65 %, (2) average grade with iron ranging between 65<Fe %>52, (3) low grade with Fe< 52 %. Based on this classification it is clear that the iron mineralisation of the Sanaga prospect is of low grade (average Fe is 40.94 wt %) but largely remain in the extracting range since the Fe₂O₃ percentage vary between 27.55 – 58.54 wt % for the oxidized iron formation and 23.32 – 55.40 wt % for gneissic formations. The total gangue content (Al₂O₃ and SiO₂ represent the main gangue in the composition of this iron formation with an average of 53.36 wt % and 51.00 wt % in the oxidized and gneissic formation respectively. This total result combine with total iron percentage showed that the Sanaga iron formations are of low grade.

Silicon and Al also impact greatly on the toughness and ductility of cast and deformed steel ⁵⁵. Consequently, understanding and defining the mineralogical association of such deleterious elements is critical in developing strategies on how to best manage and process high-grade ore.

Phosphorous and sulfur represent contamination in steel making process and are among the main target during drilling and iron ore beneficiation. As example, for all the elements commonly present in steel, phosphorus has the most embrittling effect ⁵⁵, the degree of which can be enhanced by the presence of other alloying elements in steel, particularly Cr and Mn ⁵⁶. The present bulk of production is a blend of low P <0.07 % ⁵⁷. Average P value in the Sanga permit is exactly 0.07 %. Again, sulphur content in both oxidized and gneissic formations from Sanaga is very low, below the acceptable 0.1 % in steel making ^{58, 59}. This two deleterious elements values are within the acceptable ranges, making the ore suitable for the production of steel.

6. Conclusions and Perspectives

The current study highlights a number of common feature in the Sanaga prospect base on drill hole chemical data, and the following conclusions could be derived:

(1) Iron ore mineralisation of the Sanaga prospect is hosted by magnetite gneiss similar to that observed in the Ngovayang ridge located in the Nyong Unit.

(2) The high concentration of Al₂O₃ (average 9.34 %) in the oxidized iron formations is due to recent chemical weathering.

(3) Major elements with the couple SiO₂/ Al₂O₃ and Si/Al ratio as well as trace metals ratio average (Co/Zn) confirm the hydrothermal nature of Fe and Si.

(4) The Sanaga iron formations are metamorphose chemical sediments formed by precipitation of iron and silica from a mixture of seawater and hydrothermal fluids with a minor terrigenous input.

(5) P₂O₅ as well as Sulphur percentage are very low making the iron ore suitable for the production of steel.

The low suite of trace elements from the deposit did not permit to increase geochemical evidence regarding the origin of this iron prospect. For the better understanding of the deposit, feature work would be very important. In this, the development of polished thin section, scan electron microscopy, full suite of geochemical analyses, and effective dating of this iron formation to study the various iron oxide, develop a model of iron formation and determinate the various mineral phases bearing some of the deleterious elements could be envisaged. Additionally a detail petrology is required to increase knowledge in the prospect.

Acknowledgements

We gratefully acknowledge the assistance of the Cameroun Mining Company (CMC) for making available the geochemical data and drill core of a hole within their Sanaga permit.

References