The concentrations of metals and elements were measured by inductively coupled plasma optical emission spectroscopy (ICP-OES) on a Varian apparatus (St-Laurent, Canada), model Vista Ax. A solution of cesium chloride (to improve atomization) and yttrium (1 mg/l) as an internal standard was mixed with the sample. A certified control solution (ICP-OES & ICP-MS standard) PlasmaCal Multielement Standard 900-Q30-100 (SCP Sciences, Baie-D’Urfé, Quebec, Canada) was used to ensure the accuracy of the analyses.

X-ray diffraction (XRD) was performed to reveal information about the structure of the analyzed powders. For this, a D8 Advance diffractometer (Bruker) with an X-ray source (copper anticathode Kα1 (λ = 1.5406 Å) and Kα2 (λ =

1.5445 Å)) is used. The measurements were taken in the Bragg-Brentano θ/θ configuration. The K powder samples were analyzed over a range of 2θ angles between 5˚ and 60˚ with a scan rate of 0.04˚.

Analysis by Fourier Transform Infrared Spectroscopy (FT-IR) is used to study the type of bonds formed within materials. All the K powder samples were mixed separately with potassium bromide powder (KBr, Fisher Chemical) in order to obtain a ratio of 1% mass of K powder mixed with the KBr and for a total mass of 0.2 g per sample. This mixture of K and KBr powder is first ground using a pestle and mortar to ensure the homogeneity of the mixture. The latter is then pressed in the form of pellets ~1 cm in diameter using a manual press. The FT-IR analyses of the different K powders were carried out using a Nicolet 6700 spectrometer (Thermo Electron) in absorbance mode. The FT-IR measurements were made with a spectral resolution of 4 cm⁻¹ over the spectral range (250 - 4000) cm⁻¹. A number of 512 scans are typically used for these measurements. Before each series of measurements, a pellet of pure KBr (0.2 g) is used as a reference for the subtraction of the bottom line of the final spectra of the K powders.

Thermogravimetric analysis (TGA) consists of heating a sample under a controlled atmosphere and determining its mass loss as a function of temperature. The TGA analyses of the various K powders were carried out using the TGA-Q500 system (Thermal Advantages of the company TA-Instruments). For a typical TG analysis, 5 mg of K powder is placed in a small flat cuvette made of platinum (Pt). The latter is then introduced into the vertical oven, where an air flow of 60 mL/min circulates. The temperature range studied is between room temperature (25˚C) and 1000˚C. The heating rate is fixed at 10˚C/min. The mass loss curves as a function of temperature are differentiated with respect to temperature to bring out the transition peaks corresponding to mass losses and/or phase changes.

The scanning electron microscope (SEM) is a device that uses high-voltage accelerated electron beams to observe sample surface morphology at the microscopic (or even nanometric) scale. SEM analysis of the different K powders was carried out using a VEGA3 system (TESCAN) with an electron source subjected to a high voltage fixed at 20 kV. The system is equipped with a Tescan Low Vaccum Secondary Electron Detector (LVSTD). For analysis purposes, the samples are installed on a sample holder which itself is introduced into the SEM enclosure where a vacuum of 5 × 10⁻² Pa must be maintained (using a pumping bench primary and turbo-molecular) to be able to carry out SEM observations.

Analysis by energy dispersive X-ray spectroscopy (EDX) was done in the same SEM system and used the same beam of accelerated electrons at high voltage (20 kV), except that the major difference concerned the X-rays produced by the incident electron beam. It is thus possible to characterize the composition of a sample from its X-ray emission fingerprint. The EDX analysis of the different K powders was carried out with the same VEGA3 system (TESCAN) composed of a source of electrons (accelerated to 20 kV) and an X-ray detector (XFlash, Bruker).

The results of chemical analysis by ICP-OES are recorded in Table 2 (major

elements) and Table 3 (minor elements). The clays studied are characterized by:

· A comparable loss on ignition for samples K1, K2 and K3, but it is lower for sample K7, which confirms its siliceous nature in agreement with a previous study carried out in Tunisia [11] ;

· Contents of silica SiO₂, alumina Al₂O₃ and iron oxide Fe₂O₃ for all samples (K1 to K7) similar to Katiola clay [12] .

The determination of minor elements shows the absence of arsenic, cadmium and lead in all materials. Moreover, clay samples contain chromium (873 to 150 ppm), barium (676 to 143 ppm), nickel (450 to 40 ppm), zirconium (320 to 111 ppm), strontium (276 to 11.4 ppm), zinc (257 to 23 ppm), vanadium (242 to 105 ppm), copper (154 to 38 ppm), lanthanum (85 to 26 ppm), cobalt (78 to 6 ppm), yttrium (54.2 to 16.1 ppm) and scandium (46.9 to 13.9 ppm).

The derivative thermogravimetric analysis (DTG) curves obtained (Figure 2) look the same and show all the phenomena characteristic of kaolin-type clays [13] [14] which are:

· An endothermic peak around 50˚C corresponding to the loss of absorbed water;

· An exothermic peak centered at 450˚C due to the dehydroxylation of kaolinite towards the formation of a non-crystalline phase (meta-kaolinite), indicating the loss of constitution water (450˚C - 650˚C).

Figure 3 presents the X-ray diffraction patterns of the samples, showing different intensity peaks. The major crystalline phases detected are mainly made up of:

· Si₂O₅Al₂(OH)₄ kaolinite (peaks at 12.5 Å; 20 Å; 24.9 Å; 34.9 Å; 38.2 Å; 39 Å; 62.5 Å) in the majority of the samples (K1 to K7);

· Muscovite (lines at 8.9 Å; 17.8 Å) for samples K4, K5 and K6;

· Smectite (lines at 6 Å; 29 Å; 36.8 Å) for samples K2 and K6;

· Quartz SiO₂ as impurity (lines at 21 Å; 26 Å; 39.5 Å; 50 Å; 60 Å) for all samples (K1 to K7).

The infrared spectra are almost similar (Figure 4) and include the following absorption bands:

· Between 3750 and 3500 cm⁻¹, the bands observed are those of the elongation of the hydroxyls; those at 3695; 3652 and 3618 cm⁻¹ are characteristic of the valence vibrations of the OH groups of kaolinite [15] [16] ;

· Between 1200 and 700 cm⁻¹, Si-O stretching vibrations are observed. The band at 1112 cm⁻¹ is the characteristic band of kaolinite. The shoulders at 693 and 788 cm⁻¹ can be attributed to the presence of quartz [17] ;

· Between 950 and 600 cm⁻¹, the Al-OH deformation vibrations appear, the band of which at 912 cm⁻¹ is characteristic of kaolinite [17] ;

· At 1036 cm⁻¹, a Si-O-Si stretching vibration characteristic of kaolinite [15] is observed;

· At 1007 and 538 cm⁻¹, a Si-O-Al stretching vibration characteristic of kaolinite [15] is observed;

· At 428 cm⁻¹ a deformation vibration of the Si-O or Al-O bonds can be suspected.

Scanning electron microscope observation of the samples (Figure 5) revealed:

· The presence of smectite in the form of piles in superimposed sheets with undulations [18] ;

· Plane and sub-hexagonal crystalline forms corresponding to kaolinite [19] ;

· The presence of quartz grains embedded in the matrix [19] ;

· The presence of muscovite in the form of flakes [20] .

The different EDX spectra (Figure 6) show the presence of elements such as silica, aluminum and iron in the different clay samples analyzed. These results are similar to those of some authors, with different composition percentages [21] [22] .