Mineralogical, Physico-chemical and Geochemical Characterization of Three Kaolinitic Clays (Ne Algeria): Comparative Study

Forteen clay samples collected from three kaolin deposits (Tamazert, Hadj Ali and Chekfa; NE Algeria) are characterized by several techniques in order to compares them to somes kaolins used in industry especially that used in ceramics. All the samples were investigated by X-ray diffraction, Infrared absorption spectroscopy, thermal analysis (TG), plasticity, environmental scanning electron microscopy and chemical major elements analysis. The bulk mineralogical composition of all clays samples is dominated by kaolinite (21-75%), illite/ muscovite (33-76%) and quartz (7-21%). K-feldspar and plagioclase are only present in Chekfa and Hadj Ali clays with small amounts. Clay fraction (< 2µm) dominated by kaolinite and illite (98%). Chlorite and smectite are present in some samples of Chekfa and Hadj Ali clays with insignicant amount ( (cid:0) 1%). The particles-size distribution of all samples showed the abundance of sandy silt fraction (28-63%) and silty sand (39-64%) with moderate clayey fraction (2-7%). The chemical composition showed variable amounts of SiO2 (59-68%), Al2O3 (18-39%), Fe2O3 (.26-1.38%) and TiO2 (0.34-0.69%) in accordance with the free quartz in all studied samples. Plasticity-index (7.5-7.9%), Specic surface (28-47m2) and Cation exchange (5-11meq/100g) values are moderate in all samples. Given these properties, these clays may be suitable in bricks and ceramic product. comparison between three two methods: eld methods and laboratory methods. Fieldwork consists of collecting samples according to N-S direction proles and their description. The laboratory methods consist of various analytical techniques performed on all the samples taken. Particle size analysis shows the presence of a small amount of the clay fraction (2–7%), the sandy silt fraction in proportions of (28–63%) and the sandy fraction in large proportions (39– 64%). The Atterberg limits show that the clay materials of Tamazert and Hadj Ali are moderately plastic clays, with plasticity characteristics varying between (27–39%) and a plasticity index between (8–9%).


Introduction
Clays are abundant raw materials on the surface of the Earth. They were widely used by ancient civilizations to store food, water (pottery), and to produce building materials (tiles, bricks) (Caillere et al,. 1989). Currently, clay materials nd place in many industrial, agricultural, civil, environmental and sanitary applications (Ciullo, 1996;Murray, 1999;Carretero et al., 2013;Awad et al., 2017bAwad et al., , 2018. They participate in economic (Kühnel, 1990;Ekosse, 1994;Murray, 2000) and technological development (Njopwouo, 1984;Martin, 1994Martin, , 2005Harvey and Murray, 1997). Before the use of clay, knowledge of the mineralogical, physico-chemical and geotechnical properties is essential for a better use of this clay in the appropriate industrial eld. However, the effective valorisation of monomineral clays as kaolin, talc, smectite for industrial applications (raw materials for ceramic, paint, paper…) often needs knowledge on the processes The grade and quality of economical kaolin deposits are mainly determined on the basis of kaolinite content and its degree of structural disorder, as well as quartz and Fe-Ti mineral impurities and heavy metals (Vie et  The physical properties of kaolin such as color, opacity and whiteness, compactness, plasticity and rheology, and physicochemical properties, including sorption or cation exchange capacities are largely affected by chemistry and the mineralogy of the parent rock Lyons, 1955, 1959;Vasilev et al., 1976;Cabrera and Eddleston, 1983;LaIglesia and Aznar, 1996;Fialips et al., 2000;Awad et al., 2017a).
Clayey materials are widespread in Algeria, important reserves of kaolin are presented in the Northern part of the country. This part is characterized by highland terranes with intense drainage and Mediterranean climate with a dense vegetal cover. This topographic and climate promote degradation of minerals and rocks with development of kaolin deposits associated with the Kabyle basement. Among the three main deposits (Tamazert, Chekfa and Hadj Ali), Tamazert and Chekfa deposits are under exploitation by the El Milia complex ceramic industry. However the quality of the nished ceramic products is insu cient for exportation. The third deposit of Haj Ali is used at an artisanal level for local pottery production. In this work we conduct a mineralogical and chemical characterization of the three northern Algerian kaolin deposits. Note the fourth kaolin deposits of economic potential from Djebel Debbagh (Guelma province) will not be treated in this study because of its different origin. The mineralogical composition has a major in uence on uses of clays. Whiteness and quality decreases drastically with mineral mixture and/  The three kaolin deposits are located in the Kabylie of Collo and El Milia. this region is characterized by: a hot Mediterranean climate in summer (average 25 ° C) and a mild and rainy winter (average 1200mm / year); a very rugged mountainous relief where mountains occupy more than 80% of the region; a diversi ed plant cover (cork chain, maritime pines) and very dense and a highly developed hydrographic network which drains the region in a permanent regime (Oued El Kebir, Oued El Guebli, Oued Nile), and the abundance of sources which supply the villages drinking water. The Tamazert deposit is located about 17 km north of the town of El Milia (West of Jijel). Hadj Ali deposit is located at 14 km of the Ain Kechera town (West of Skikda) not far from Tamazert. Chekfa deposit is located about 23 km east of Jijel (Fig.1).

2.2-Geological condition
The Kabyle massifs are traditionally considered as the substratum of the internal zones of the Tellian chain, an Algerian section of the Maghreb alpine chain (Fig. 1). The latter extends linearly along the Mediterranean coast over 1200 km from the Rif to Calabria. The massif of Less Kabylie is located in the northern position of the inner domain of the Tellian range (Fig. 2). It stretches along the Mediterranean coast for more than 150 km between the region of Jijel in the West and the massif of Fil la (Skikda) in the East, with a NS extension of 50 km (Fig.1), it constitutes the outcrop of most important crystallophyllian lands of the Algerian coast. Previous work on Less Kabylie shows the existence of a major abnormal contact between the crystallophyllian Kabyle massif (basement layer) and the infra-Kabyle complex which is constituted by the Mauritanian, Massylian and Tellian units (Fig.1). (D-Delga, 1969;Bouillin, 1978;Vila, 1978

3.1-Material and sampling
The samples on which this study is based are collected from three kaolin deposits (Tamazert, Hadj Ali and Chekfa) according to N-S direction pro les. The clay materials of the three deposits present a similar organization. of the pro le shows a succession of three vertical zone from top to bottom (Fig. 2): a strongly kaolinized zone (D); a moderately kaolinized zone (C1, C2, C3); a weakly kaolinized zone (B) and an intact zone (A) represent the parent rock. The passage between the parent rock and the weakly kaolinized level is not linear. Fourteen samples were taken (Table 1) from the different levels of the three pro les. Several preparation steps (grinding, sieving and washing) were carried out in order to obtain different information on the mineralogical composition of clay materials.

Description of the pro les
The pro les show identical morphologies and variable dimensions. Field observations and examination of thin sections in a polarizing microscope made it possible to follow the evolution of weathering from healthy rock to totally weathered rock. From bottom to top, there are three distinct sets: set of alterites, nodular set and clay-sandy loose set. This vertical succession characterized classic equatorial zones (Fig.   2).

X-ray diffraction (XRD)
Bulk minerals compositions were identi ed by X-ray diffraction (XRD) carried out with D8 Bruker Advance diffractometer using Cu Kα1 radiations (λ=1.5418). All XRD bulk data were collected in same experimental conditions, in the angular range 2θ from 2 to 45°, 40 kV, 30mA, with step scan 0.020° and a step time 0.6s for all bulk samples. The clay fraction (< 2µm) is obtained by sedimentation according to the stocke's law and centrifugation after removel of carbonate by HCl (10%). Oriented slides are analyzed under 3 treatments (natural, glycol and heated in 500°C for four hours). XRD data from clay fraction ((< 2µm) were carried in the range 2-15° 2θ under the same conditions mentioned above.

Plasticity test
The plasticity of a clay is a fundamental technological parameter which in uences the characteristics of ceramic materials (Sadik et al., 2012). It was carried out on the fraction less than 63 μm. In this study, we opted for the method known as "Atterberg limits". The Atterberg limits, were measured at the laboratory of Materials and Structures Mechanics at the University of Liege. Measurements have been done on the on raw samples using the method described by (Casagrande, 1947 and Andrade et al., 2011). The plasticity index PI is calculated by the difference between LL and LP according to (Casagrande, 1947). The results of the tests carried out are presented in (Table 2).

Infrared spectroscopy (FTIR)
Infrared spectra (FTIR) were recorded with a Nicolet NEXUS spectrometer at laboratory of mineralogy and crystallography at the University of Liege. For each sample, 32 scans were recorded with a resolution of 1 cm -1 over an interval ranging between 400 and 4000 cm -1 (corresponding to middle-infrared). Two mg of clays were mixed with 148 mg of KBr and then pressed up to 9 tones in order to obtain a homogeneous pellet. The pellet is then dried for several hours at 120°C. To avoid any water contamination the measurements were carried out under vacuum.

Thermal analysis
Gravimetric analysis (TGA) was performed in the LCIS Green-Mat laboratory (Department of chemistry University of Liege using a NEZSCH STA 449 PC instrument. TGA analysis was performed on the clay and slit fraction (< 63µm) over a temperature range at 2 to 1000°C. The "spot test" was used to determine surface area as prescribed by (Johnson 1957) and (Worall 1958). The Cation Exchange Capacity (CEC) was calculated by distillation. An amount of 0.5g of the sample was mixed with 2 ml of NaOH. Then titration is done with HCl and titrated by the Kjeldahl method (Mackenzie et al., 1954). The exchangeable cations were measured using ammonium acetate method at the soil laboratory of Catholic University of Louvain.
Chlorite is identi ed by the persistence of the second re ection at (7Å) during heating. Smectite is evidenced by its characteristic peak at 14Å under natural condition that migrate to16Å after glycolation and collapses to 10Å after heating. The semi-quantitative abundance of the minerals in all samples was estimated from the height of a diagnostic peak multiplied by a corrective factor as reported (Table 3) by (Cook et al.,1975; (Boski et al,.1998 andFagel et al., 2003). The semi-quantitative data of the clay minerals of the TM, CH and HA deposits are shown in (Fig.4).These data show that, TM samples are enriched in clay minerals (57% on average) in regard with the samples of CH and HA (14% and 35% on average).

Plasticity test
Plasticity of clays is one of the most important parameters affecting the determination and process of clay production (Murray, 2007;Barış et al., 2020). The Evaluation of the plasticity index (PI) is very important in determining the suitability of clays for the ceramic industry. The plasticity index of Tamazert and Hadj Ali clays was calculated from the arithmetic difference of LL and PL. these Clay materials are characterized by moderate plasticity index 9% and 8% respectively. we could not perform the plasticity test on the Chekfa samples because of the high quartz content. In the diagram of Holtz and Kovacs (1981) (Fig. 5), the clays of Tamazert and Hadj Ali are ploted in the zone of midium plasticity,the samples plot as kaolintic clay. The PI of all samples clay is < 10%, samples were found unsuitable for bulding-related ceramics production due to the risk craks during extrusion (Nyakairu et al., 2002)

Scanning electron micrscopy
The micrographs (Fig. 8) obtained by the different samples showeds the dominance of kaolinite pseudohexagonal particles organized as platelets or tight packages. EDS microanalyses reveal some small relics of feldspars and quartz in TM kaolinite whereas fresh or slightly altered K-feldspars minerals are evidenced in both CH and HA kaolinites. The better crystallization of TM and HA samples evidenced by ESEM is consistent with X-ray diffraction and IR spectroscopy data.

Physico-chemicals properties (SSA and CEC)
The CEC of Tamazert clay (11.58 meq/100g) is higher than for Chekfa and Hadj Ali (5.42 meq/100g and 6.46 meq/100g (Table.2). Such CEC are in the range of the values reported in literature for kaolin deposits (3 to 15 meq/100g, in Grim 1968). Speci c surface areas of CH and HA kaolinites (~45.5 m 2 /g) are higher than that of TM kaolinite (28.7 m 2 /g). This re ects the small grain size of Chekfa and Hadj Ali kaolinites compared to Tamazert kaolinite.

Particles-size distribution
The suitability of clays for different industrial applications is based on their particle size distribution.
For ceramic products, the ner fraction (<2 μm) is of particular attention (Mahmoudi et al., 2008). The particle size distribution of the different samples shows that these clays are poor in ne fraction (2-7%), while the silty sands and sandy loam fractions are abundant (28 to 63% and 34 to 64%, respectively). The samples from Chekfa and Hadj Ali show low ne fraction contents (<2 μm) compared to those from Tamazert. In the ternary diagram widely used in the ceramic industry (Dondi et al., 1992), the clay samples studied were classi ed as silty sand (samples H1, H3 and C1) and sandy silt clay (sample T2, T3, T6 , C1, C2 and H5, Fig. 9). On Winkler's (1954) ternary diagram (Fig.10), almost all of the samples show an aptitude for making construction products such as common bricks.

Chemical analysis
The chemical composition of the clay samples are showen ( within clays depends on the intensity of hydrolysis. More hydrolysis gives more kaolin minerals and, therefore higher Al 2 O 3 content (Yanik, 2011). Fe 2 O 3 is low for TM (<1%) and ranges between 0.85 to 1.63% for HA and CH. CaO elements, MgO, TiO 2 and P 2 O 5 are present in small quantities often lower than the detection limit, in all sites. Na 2 O presents small variations from one site to another (< 1% for TM and CH, > 2 % for most HA samples). In contrast, K 2 O presents higher uctuations, ranging between 1.8 and 5.1% for the 3 sites. LOI values from kaolin samples ranged from 5.7 to 7.0% for Chekfa, 4.2 to 5.0% for Hadj Ali and 6.6 to 13.8% for Tamazert. The low CaO, MgO, Na 2 O values re ect an important leaching of calcoalcaline elements. It is probably due to their high mobility during kaolinization process and it is compatible with an advanced argillic alteration system close to hydrothermal kaolin deposits (Meyer and Hemley, 1967;Meunier et al., 1983. Inoue, 1995Dill et al., 1997Dill et al., , 2000. However, the enrichment in K 2 O observed in all samples is probably due to the release of the potassium content by feldspar dissolution. 5 Discussion And Perspective 4.1. Weathering stage of the NE Algeria kaolinitic-rich deposits Kaolinite, illite and quartz represent the main minerals in the 3 studied kaolin deposits from N Algeria. However the 3 sites are characterized by different properties that in uence the quality of the kaolin ores. TM clays is more weathered than CH and HA as evidenced by the observed relics of K-feldspar and high crystallinity, high CEC (11,6 meq / 100 g) and Low SA (28,7 m 2 /g). In addition the presence of 2 bands at 3668 cm -1 and 3653 cm -1 on the IR spectrum of TM attests for a better crystallinity of its kaolinite (Cases et al., 1982;Petit, 1994;Filiaps, 1999). The high SA in HA and CH re ects their low abundance of ne clay particle and the occurrence of impurities such as fresh feldspars. Likely the lower CEC in CH and HA may be explained by the presence of higher amount of illite (Tschapek et al., 1974;Yong et al., 1992) T2  14  56  2  -28  46  46  -T3  11  61  7  -21  65  65  -T4  23  13  --64  33  33  -T5  5  8  --87  77  77  -T6  10  11  -1  78  23  23  -H1  14  34  2  19  31  33  57  10   H2  9  44  -18  29  30  67  3   H3  11  33  2  10  44  29  43  28   H4  11  29  3  18  39  40  7  53   H5  12  30  -20  34  57  37  6   C1  9  34  18  25  14  21  74  5   C2  3  47  3  26  14  22  52  26   C3  10  47  9  16  15  60 40 -