The materials used were essentially sampling and laboratory equipment. For the sampling campaign, the equipment consisted of a 1/200,000 topographic map of the Gagnoa sheet [15] , a Garmin Etrex 10 GPS, a hand shovel and plastic bags for packaging the samples taken. In the laboratory, sample preparation equipment included a Memmert-Saftel oven and a ball mill. Sample analysis equipment consisted of a Niton XL3t X-ray fluorescence spectrometer (XRF) and a GBC-EMMA diffractometer (XRD). According to Thermo Fisher Scientific Inc [16] , manufacturer of XRF spectrometers, limits of detection (LOD) depend on test duration, interference/matrices and statistical confidence level. Research on Niton XL3t, XLi, XLp and XLt analyzers, with measurement times ranging from 30 seconds to 2 minutes, depending on the type of analyzer, yielded the following limit values: As (2 - 10 ppm), Cr (2 - 50 ppm), Cu (20 - 60 ppm), Ni (40 - 90 ppm), Zn (9 - 80 ppm) and V (20 - 60 ppm).

The work methodology can be subdivided into four parts, namely sampling, preparation and XRF and XRD analysis of samples in the laboratory, statistical analysis of content data and determination, on the basis of indices, of levels and possible sources of contaminants and/or metal pollutants.

The study area abounds in a multitude of sites, but not all of them are active. In fact, the activity is subject to the phenomenon of a rush to the richest areas. During the sampling period, twenty (20) localities ( Figure 2 ) accounted for most of the mining activity in the Bouaflé region. They are: Alekran, Aley, Allaiyaokro,

Angovia, Benou, Bozi, Degbezre, Kaviessou, Kouassi Périta, Koudougou, Ndenou, Toumbokro, Zagouta, Ouanzanou, Krigambo, Okabo Kouakougnanou and Pakouabo with its major gold mining camps of Bel chantier and Lakrikro. Each of these localities has several active sites which were sampled. For a given locality, samples from several relatively close sites were blended to form a composite sample representative of the locality.

The sampling campaign took place from February 05 to 18, 2023, during the dry season. For the XRF spectrometer analyses, samples of mine tailings were taken from the sites using a plastic hand shovel. They consisted of solid rejects (rocks, spoil heaps) and sluice slurry, and came from a variety of deposits (alluvium, eluvium, quartz veins, etc.). For XRD analysis, the samples are made up of more or less altered rocks (basalts, gabbros) (saprolite and bedrock), clays and ore processing sludges as shown in Figure 3 . All samples collected were packaged in plastic bags and coded according to their nature and origin.

Samples were dried at 60˚C for three days in a Memmert-Saftel oven, then ground to 80% passing 75 µm. As mentioned above, an XRF spectrometer was

used for concentration determination, coupled with a stand to avoid the risk of X-ray irradiation. Cups covered at their bases with 06 µm-thick Mylar sheets are filled with sample powders and placed in the stand’s housing. The instrument offers the possibility of adapting measurement modes to suit the nature of the sample and the type of elements sought (major or trace elements).

In the case of our analysis, we chose the “Soil” mode, better suited to the detection of MTEs. This mode features 03 filters spaced 30 seconds apart, giving a measurement time of 90 seconds per sample. After analysis, the content data are collected on the computer using the Niton Data Transfer (NDT) software associated with the spectrometer.

The samples collected undergo the same mechanical treatment as those used for XRF analysis. They are then mounted, which consists of placing the powder in a glass sample holder and aligning it precisely in the XRD measuring device. Calibration is then carried out using a reference sample, crystalline silica. Finally, the DRX device is switched on to collect diffraction data. Ten (10) samples selected on the basis of their heterogeneity were subjected to XRD analysis.

Raw assay data were sorted by eliminating major elements (except titanium, which will be used as a normalizing element in index calculations) and trace elements that were not or only partially detected, to retain only those elements with assays above the instrument’s detection limits at all sites. It should be noted that several measurements were carried out on the same sample to retain the average content.

Next, monovarietal statistics were performed using XL-STAT software to determine the basic descriptive parameters of the population studied (central tendencies and dispersions). Multivariate descriptive statistics were used to determine correlations between chemical elements and to perform a principal component analysis (PCA). PCA is used to summarize the information provided by the study’s MTEs (quantitative variables) and localities (individuals). To do this, the eigenvalues of the variables were determined in order to identify the main components, using Kaizer’s criterion (positive values). Next, a cone diagram was designed to better visualize the eigenvalues. The analysis continued with the correlations between the variables. All these parameters led to the representation of correlation circles and visualization of the importance of each explanatory variable for each axis of representation, in particular, the relationships between mining sites and ETMs.

When X-rays interact with matter, they cause a displacement of the electron cloud relative to the nucleus in atoms. These oscillations induce the re-emission of electromagnetic waves of the same frequency, a phenomenon known as Rayleigh scattering. The analysis parameters are as follows: Current: 28 mA; Voltage: 20 kV; Power: 1 kW and filament amperage: 4.5 mA. The interference of the scattered rays is alternately constructive or destructive, forming diffraction peaks. The relationship between the position of the peaks and the crystal lattice is described by Bragg’s law, according to the following formula:

$n\lambda =2d\sin \left(\theta \right)$ (1)

In this formula,

- n is the order of reflection,

- λ is the X-ray wavelength,

- d is the inter-reticular distance (between two crystallographic planes), and is the half-angle of deviation.

Crystalline phases are identified by analyzing the diffraction signature specific to each crystalline phase, each forming a characteristic diffraction pattern.

Four (04) indices were calculated on the basis of reference concentrations in the Upper Continental Crust (UCC) of the ETMs selected for the study [18] . These are the Geoaccumulation index (Igeo), the Enrichment Factor (EF), the Contamination Factor (CF) and the Polymetallic contamination degree (CD). According to [19] , the notion of “background noise” or “back ground”, which refers to the concentration of a chemical element without human input, poses a number of problems in more or less anthropized environments. Citing [20] , he recommends that the use of the term “geochemical background”, which is too complex, should therefore be abandoned in favor of the concentration of trace metal elements in the upper continental crust as the geochemical reference background as shown in Table 1 .

The Igeo is a criterion for assessing pollution intensity [21] [22] . This empirical index compares a given concentration with the geochemical background. Igeo is calculated according to the following formula:

$\text{Igeo}={\log }_2\left[\left(\text{Cn}\right)/1.5\left(\text{Bn}\right)\right]$ (2)

With:

- Cn: measured element concentration,

- Bn: reference concentration of the element.

The coefficient 1.5 is a correction factor to account for heterogeneity in background levels caused by lithological effects.

For the interpretation of Igeo values, Müller proposed 07 classes defined in Table 2 .

The enrichment factor highlights the origin of an element in sediments [23] , and has been proposed to discriminate between anthropogenic and natural terrigenous inputs. In this work, the normalizing element is titanium due to its low mobility [24] and abundance in the samples. The enrichment factor is expressed by the following formula:

$\text{EF}=\left[\left(\text{CMs}/\text{CNs}\right)/\left(\text{CMr}/\text{CNr}\right)\right]$ (3)

With:

- CMs = concentration of metal M in the sample;

- CNs = concentration of the normalizing element in the sample;

- CMr = concentration of metal M in reference material;

- CNr = concentration of normalizing element in reference material.

For the interpretation of EF values, [23] proposed 05 classes defined in Table 3 .

CF is an excellent tool for monitoring the contamination of an environment by trace metals. It is an index that highlights the presence or absence of contamination, but also provides information on the level of contamination [25] . It is calculated using the following formula:

CF = (Cn/Bn) (4)

With:

- Cn: concentration of the element in the sample;

- Bn: reference concentration of the element.

For the interpretation of CF values, Hänkanson proposed 04 classes defined in Table 4 .

This index is derived from the determination of contamination factors. It is used to estimate the polymetallic, in advance, for each sampling site. For a given site, the CD is the sum of the CFi of the n chemical elements at that site. It is calculated using the following equation:

CD = CF1 + CF2 +... + CFn (5)

For the interpretation of CD values, Hänkanson proposed 04 classes defined in Table 5 .

This index determines the ecological sensitivity and toxicity of pollutants [25] [26] . It has been calculated specifically for arsenic and chromium. It is calculated by the following formula:

Eri = Tri × CFi (6)

- CFi: contamination factor of component i;

- Tri: toxic response factor of a given component. The Tri values used are 10 (As), 5 (Cu) 2 (Cr) and 1 (Zn) [25] [26] .

For the interpretation of Eri values, Hankanson proposed 05 classes defined in Table 6 .

This index takes into account the combined effect of several parameters, namely the toxic level, the concentration of heavy metals and the ecological sensitivity of biological communities to heavy metals [27] . The RI is calculated using the following equation:

$\text{RI}={\displaystyle {\sum }_{i=1}^n\text{Eri}}$ (7)

where “Eri” represents the ecological risk index specific to metal i.

For the interpretation of Eri values, Singh proposed 05 classes defined in Table 7 .

The following parameters were used to achieve a relative classification between TMEs:

For the classification of sites by contamination level, the polymetallic contamination degree (CD) has been used.

As indicated above, a total of twenty (20) mining sites in the study area were subject to composite sampling (ore extraction and processing zones). Most of the processing areas are located near water sources, but for safety reasons they are also located in villages. Figure 2 shows the mappable localities hosting the mining sites in the study area.

The results of the XRF analysis are presented in Table 8 .

Table 9 shows the statistical parameters for the six (06) MTEs that met the

*Normalising element for calculating the enrichment factor.

selection criteria after XRF analysis, namely abundance, complete detection at all sites and proven toxicity naturally or by accumulation in receiving environments.

The ratio of crustal mean and abundance (M/UCC) is greater than 1 for the six (06) TMEs analyzed, indicating enrichment in these elements [28] .

The following table shows the results of linear relationships between MTEs, based on Pearson correlation coefficients. Values in bold indicate a significant correlation. This matrix shows some pairs of elements with correlations ranging from strong to very strong (Pearson classification; Table 10 ).

This finding is confirmed by the Principal Component Analysis (PCA) and the scree plot (or cone diagram; Figure 4 ) below represented from the eigenvalues ( Table 11 ) which indicate the following 02 components: Component 1: As, Cu, Cr and V; Component 2: Zn and Ni.

The PCA carried out enabled us to visualize the variables, i.e. the MTEs targeted by the study, as well as the mining sites concerned as shown in Figures 5-7 .

On these representations, we can see that the MTEs are well represented on the correlation circle (far from the center). Their relationships with the mining localities are described as follows:

Table 12 shows the results of Igeo and EF calculations for each site.

The Igeo table shows that all sites are polluted with at least four (04) TMEs, notably, As, Cr, Ni, Cu and V. Igeo values range from 0.52 to 4.59 (As), −1.85 to 0.24 (Zn), −0.11 to 2.67 (Cu), 0.61 to 2.72 (Ni), 1.81 to 3.69 (Cr) and 0.40 to 2.64 (V). The Igeo is therefore more pronounced for As and Cr, in contrast to Zn, whose Igeo is below zero at almost all sites. The same trend is observed for the Enrichment Factor, where As and Cr are still the most pronounced at almost all sites. Values range from 1.67 to 24.82 (As), 0.22 to 1.65 (Zn), 1.08 to 5.75 (Cu), 1.26 to 6.61 (Ni), 2.83 to 10.76 (Cr) and 1.45 to 5.21 (V).

According to [29] , a distinction must be made between contamination, which is the introduction of “a potentially dangerous substance, whatever its content” into the natural transfer and exposure media of soil, air and water, and pollution, which is present when “the content is potentially dangerous or reaches the limit values set by standards”. Contamination is therefore not synonymous with pollution. Table 13 shows the results of the CF and CD.

calculations for each site and for the targeted chemical elements. It should be remembered that the results of the Igeo and EF calculations enabled us to distinguish As and Cr as major pollutants. The eventual solubilization of these two (02) TMEs and their accumulation in receiving environments such as water can affect the food chain. For this reason, ecological risk parameters for these two (02) ETMs were also calculated.

A look at the coloring of the tables in relation to contamination/pollution intensities provides information on the severity of the sites’ situation, except for zinc contamination, where CF values range from 0.42 to 1.77. The CF values for the other TMEs range from 2.15 to 36.07 (As), 1.39 to 9.56 (Cu), 2.29 to 9.92 (Ni), 5.27 to 19.36 (Cr) and 1.98 to 9.38 (V).

The approach adopted for the relative classification of the MEs targeted has enabled us to determine the preponderance levels in Table 14 . This table takes into account the average content of each TME across all sites and the maximum values of three (03) indices, Igeo, EF and CF per TME.

The following classifications emerge from these graphs ( Figure 8 and Figure 9 ):

Arsenic and chromium are therefore the dominant TMEs and the main contaminants and/or pollutants in the study area.

The classification of contamination/pollution levels by locality is based on potential ecological risk indices, specifically for arsenic and chromium, which are the most abundant, and degrees of contamination. The results are shown in Figure 10 and Figure 11 .

The graphs showing the classification of contamination or pollution levels by locality illustrate the extent of the adverse effects of mining activities throughout the study area.

It should be pointed out that these localities are all located at the foot of the Yaoure Mountains. Angovia and Zagouta, the localities of greatest concern, are former gold-panning sites that are still active. The same applies to Kaviéssou, Alékran, Aley, Toumbokro and Kouassi Périta. On the other hand, the N’Denou, Benou and Dégbézré sites are recent but very active. All these sites require environmental monitoring.

The analysis of ten (10) samples consisting of clays, saprolites, slightly altered rocks and sludges yielded the following diffractograms.

By interpreting the results of the XRD analyses, it is possible to identify the minerals that make up the various samples ( Table 15 ). Dominant minerals are shown in bold.

XRD analysis of the samples clearly shows that the mineralogy of the samples is complex but mainly dominated by stenhuggarite (CaFe³⁺(As³⁺O₂)(As³⁺SbO₅), periclase (MgO), pyrrhotite (Fe₁₋ₓS (where x varies from 0 to 0.20), carlinite (TI₂S), filipstadite ((Mn,Mg)( ${\text{Sb}}_{0.5}^{\text{5}+}$ ${\text{Fe}}_{0.5}^{\text{3}+}$)O₄), calcite (CaCO₃) and Antigorite ((Mg, Fe II)₃ Si₂O₅ (OH)₄)).

The elements Fe, Mg and Si are not too problematic for the environment, but sulfide minerals such as pyrrhotite are potentially acidogenic and arsenates, potentially toxic. It should be noted that the base rocks contain carbonates (calcite and dolomites) that can act as neutralizers in the event of Acid Mine Drainage (AMD) [30] .

As mentioned above, the geology of the Yaouré furrow is made up of an assemblage of magmatic volcanic, intrusive and sedimentary rocks. Most of the mining sites incriminated by potential ecological risks and/or high levels of contamination are located in the following lithological formations ( Figure 11 ):

It should be noted that a study by [6] showed that the metalliferous paragenesis in the Yaoure furrow is mainly represented by pyrite (abundant) and chalcopyrite (frequent). The presence of pyrrhotite, pentlandite and magnetite, as well as traces of sphalerite, cubanite, bravoite, lineite, molybdenite, Bi, Ni, Hg tellurides (tetradymite, melonite, coloradoite) and cassiterite was reported by [31] . Their work also showed that Angovia’s gold mineralization is linked to intense hydrothermal alteration with quartz, chlorite, carbonates and sulfides.

Among other phenomena, this hydrothermal alteration led to corrosion of the sulfides, the mineral carriers of the contaminants studied. This corrosion was followed by soil reworking due to intensive mining activity in the area, as demonstrated by the high levels of enrichment factors in several localities. Chromium and vanadium may also have originated from ultrabasic rocks reported in the study area.

The risk of a chemical element’s impact on the environment depends on its pollutant capacity, linked to its concentration and reactivity, which in turn depends on several parameters, including its chemical form and mobility in the environment. Here we provide some informations on chromium and arsenic, the main pollutants in the study area.