The Kaolinitic clay (Kc) used in this work was from Dibamba, Littoral Region of Cameroon. Some chemicals such as sucrose pellets (99 wt%), hexane (96 wt%), hydrochloric acid (32 wt%), ammoniac (99.5 w%) acquired from PROLABO Ltd. were also used in minor quantities. The chemicals were used without any further purification.

The clay material was characterized by several techniques described as follow: the chemical composition was carried out by ICP analysis and the X-ray diffraction (performed with Bruker D8 Advance with Cu Kα1 radiation equipment). The differential coupled with gravimetric thermal (DTA/TGA) analyses using a Setsys 2400 de SETARAM apparatus, heating rate 5˚C/min). The phase composition was investigated by Fourier transform infrared spectra recorded in the range of 4000 - 300 cm⁻¹ using Bruker FT-IR Tensor 27 spectrophotometer.

The mineralogical phase was enriched by wet sieving at 75 µm, and it was dried to obtain a powder fraction labelled as Kc. The synthesis process was in few steps, starting with the thermal activation of the clay at 600˚C, 700˚C, and 800˚C for 1 hour in a model LH60/14 muffle furnace with a heating rate of 5˚C/min to deliver precursors according with the following equation.

${\text{Al}}_{\text{2}}{\text{O}}_{\text{3}}\cdot {\text{2SiO}}_{\text{2}}\cdot {\text{2H}}_{\text{2}}\text{O}\left(\text{Kc}\right)\stackrel{600\text{-}800˚\text{C}}{\to }{\text{Al}}_{\text{2}}{\text{O}}_{\text{3}}\cdot {\text{2SiO}}_{\text{2}}\left(\text{metakaolinite}\right)+2{\text{H}}_{\text{2}}\text{O}$ (1)

The samples obtained were labelled MK600, MK700, MK800, where MK stands for Metakaolin. Those precursors were used for the synthesis of alumina nanoparticles following a modified sol-gel associated with the bottom-up process [19] [20] . Indeed, each precursor was chemically treated with a 2.5 N solution of hydrochloric acid for 2 hours at 90˚C, (5 > pH > 3) as shown in Equation (2).

$\begin{array}{l}{\text{Al}}_{\text{2}}{\text{O}}_{\text{3}}\cdot {\text{2SiO}}_{\text{2}}+\left({\text{Fe}}_{\text{2}}{\text{O}}_3,{\text{TiO}}_2,{\text{SiO}}_2\right)+\text{HCl}\\ \to 2{\text{SiO}}_2\left(\text{residue}\right)+\left(n{\left[\text{Al}\left({\text{H}}_{\text{2}}\text{O}\right)\right]}^{3+}+3n{\text{Cl}}^{-}+\left[{\text{Fe}}^{3+}+{\text{Ti}}^{4+}\right]\right)\left(\text{aqueous}\text{filtrate}\right)\end{array}$ (2)

The mixture was filtered and the filtrate was then kept for further use. The amphoteric character of aluminum (III) species is preferentially shifted to basic to facilitate its solubility in the water together with other oxides. The liquid phase recovered, typically alumina was submitted to react with ammonium hydroxide (NH₄OH) to form Aluminum hydroxide (Al(OH)₃) which is then soaked in hexane to purify it and dried. Sucrose was added to minimize the particle agglomeration.

$\begin{array}{l}AlC{l}_3\left(aq\right)+C_{11}{H}_{22}{O}_{11}\left(s\right)+3N{H}_{4}OH\left(l\right)\\ \to Al{\left(OH\right)}_3+3N{H}_{4}Cl\left(g\right)+10C\left(s\right)+10H_2\left(g\right)+C{O}_2\left(g\right)+H_{2}O\left(l\right)+4O_2\left(g\right)\end{array}$ (3)

The Aluminum hydroxide (Al(OH)₃) obtained is then soaked in an appropriate quantity of hexane for 24 hours, then dry after filtration. The role of hexane is to rid the product of more organics impurities. The calcination at 900˚C of product in the last step permits the transformation in amorphous phase following the Equation (4).

$\text{Al}{\left(\text{OH}\right)}_{\text{3}}\stackrel{900˚\text{C}}{\to }\gamma {\text{-Al}}_{\text{2}}{\text{O}}_{\text{3}}$ (4)

The obtained powders were analysed with Scanning Electron Microscopy (SEM) using A Zeiss Ultra Plus FEG for surface morphology analysis, equipped with an Oxford detector EDX at 20 kV, which uses Aztec software for elemental analysis.

The chemical bonds in the final products were investigated by Fourier transform infrared spectra recorded in the range of 300 - 4000 cm⁻¹ using Bruker FT-IR Tensor 27 spectrophotometer in addition to powder diffraction patterns which were recorded in the high angle 2 theta range (5˚ - 70˚) using a Bruker D8 Advance with Cu Kα1 radiation. The scan speed and step sizes were 0.3/min and 0.001˚ respectively.

The specific surface area was obtained by BET method, from N₂ adsorption and desorption isotherms at 77 K, obtained from a Carlo Erba Sorptomatic 1990 volumetric device, after outgassing the samples overnight at room temperature at a pressure lower than 10⁻⁴ Pa.

The yield of the synthesis process (Y) was calculated following the equation [14] [21] .

$Y\left(\%\right)=\frac{m_{\text{NKi}}}{m_{\text{Ki}}}\times 100$ (5)

where m_NKi is the mass of Al₂O₃ obtained in the final step of procedure and m_Ki is the initial mass of Al₂O₃ in calcined kaolinitic clay (600˚C, 700˚C, 800˚C).

For the determination of crystallite size, the X-ray patterns of the three alumina powders were considered. Some related peaks were then used to evaluate the average size (D) in the hkl direction, expressed in Å of the crystallites using Debye Scherrer equation [1] [22] [23] .

$D=\frac{K\times \lambda }{HL\times \cos \theta }$ (6)

where k is apparatus correction constant, or Scherrer constant (0.89); HL the angular width at half-height lengh of the diffraction peak relative to the reticular plane (hkl) expressed in radians. HL = 2θ_(L2) − 2θ_(L1); λ: wavelength of the KαCu = 1.5406 Å = 0.15406 nm; θ: half the wave deviation (half the Bragg angle, at the peak of the reflection under consideration, expressed in degrees).

The chemical composition of clay material (Kc) obtained by ICP analysis was shown in Table 1 .

The clay material has a high mass proportion of silica: SiO₂ (76.06%), while alumina is comparatively low: Al₂O₃ (14.65%). This gives a SiO₂/Al₂O₃ molar ratio of 5.19, classifying the clay as siliceous [24] [25] . In general, alkaline oxides (Na₂O, K₂O) and alkaline earth oxides (CaO, MgO) were present in very small quantities. This indicates that the material is a low-melting point [4] [26] . The coloring oxides TiO₂ and Fe₂O₃ are present in very small quantities.

Figure 1 shows the (XRD) patterns of the clay material (Kc). Some peaks of crystalline phases were revealed.

a. detection Limit; b. loss on fire.

Kaolinite (Al₂Si₂O₅(OH)₄) (JCPDS 14-164) is the main mineral phase of the clay material studied, associated with other phases as quartz (JCPDS 5-490) (SiO₂), lepidocrocite (FeOOH) (JCPDS 8-98), and anatase (TiO₂) (JCPDS 4-447).

The infrared spectrum of Kc presented in Figure 2 showed the typical appearance of a kaolinite spectrum [3] [27] . The absorption bands at 3440 and 3370 cm⁻¹ corresponds to the vibrations of the O-H bonds of the adsorbed water molecules, while those at 3689, 3647, and 3619 cm⁻¹ are attributed to the symmetrical elongation vibrations of the free OH groups on the external surface and the elongation vibrations of the internal -OH groups of the kaolinite [14] [27] . The absence of the absorption band between 3665 and 3670 cm⁻¹ reflects an inversion of the band intensities at 3647 and 3670 cm⁻¹. This is only observed when there is disorder within the material [25] [26] [28] . The absorption bands located around 1114 - 1116 cm⁻¹, 998-1025 cm⁻¹ on the spectrum correspond respectively to the deformation vibrations of the Si-O and symmetrical Si-O-Si and asymmetrical Si-O-Al bonds [25] [29] , while the bands at 908 and 909 cm⁻¹ as well as those at 788 - 792 cm⁻¹, 747 cm⁻¹ and 663-677 cm⁻¹ were attributed to the vibrations of the Al-O-Si and Al-OH bonds in kaolinite. The shoulders around 521 - 523 cm⁻¹, and 458 cm⁻¹ shows the deformation vibrations of the Si-O-Si and Si-O-Al bonds [30] [31] . Shoulders around 1150, 714, and 695 cm⁻¹ indicated the presence of quartz [25] .

Figure 3 showed the (DTA and TGA) thermograms of the clay material and revealed that the shapes of the two curves were typical to those of kaolinitic clays [16] [26] [28] [32] .

An endothermic peak on the DTA thermogram of clay around 80˚C accompanied by a small loss of mass on the (TGA) curve, this thermal incident would correspond to the endothermic phenomenon of the departure of the water of hydration molecules [32] [33] . An endothermic peak was observed around 260˚C, with a mass loss of around 1.84% on the (TGA) curve corresponding to the dehydroxylation of lepidocrocite (γ-FeO(OH)) into maghemite (γ-Fe₂O₃) according the equation [25] [28] .

$\underset{Lepidocrocite}{2\gamma \text{-}FeOOH}\stackrel{259˚\text{C}}{\to }\underset{Maghemite}{\gamma \text{-}F{e}_2{O}_3}+\underset{Water}{H_{2}O}$ (7)

The content of this mineral from the proportion of loss on ignition related to this mineral gives 10.18% Lepidocrocite. Another endothermic phenomenon occurs between 400˚C and 600˚C, corresponding to the dehydroxylation of kaolinite to form metakaolinite [25] [32] .

$\underset{Kaolinite}{S{i}_2{O}_{5}A{l}_2{\left(OH\right)}_4}\stackrel{500˚\text{C}}{\to }\underset{Metakaolinite}{2Si{O}_2\cdot A{l}_2{O}_3}+\underset{waterofconstitution}{2H_{2}O}$ (8)

Around 574˚C there is an endothermic phenomenon with no loss of mass that reflects the polymorphic transformation of alpha quartz into béta quartz as shown in following equation.

$\underset{quartz\alpha }{\alpha \text{-}Si{O}_2}\stackrel{573˚\text{C}}{\to }\underset{quartz\beta }{\beta \text{-}Si{O}_2}$ (9)

Between 900˚C and 1000˚C, the exothermic peak on the DTA curves corresponds to the reorganization of metakaolinite and the formation of pseudo-spinel [34] .

$\underset{Metakaolinite}{2\left(2Si{O}_2\cdot A{l}_2{O}_3\right)}\stackrel{925\text{-}1000˚\text{C}}{\to }\underset{Pseudo\text{-}spinelle}{2A{l}_2{O}_3\cdot 3Si{O}_2}+\underset{crystalizedSilica}{Si{O}_2}$ (10)

These observations confirm that the treatment of Kaolinitic clay (Kc) up to 700˚C, transformed it into metakaolinite which is a more reactive amorphized form. At this temperature, no reorganized phase was yet detected, which proves that the whole system was indeed activated so that hydrated minerals can decompose into primitive oxides [25] [32] .

The X-ray patterns ( Figure 4 ) of the three products synthesized on the basis of precursors MK600, MK700, and MK800 respectively permitted the determination of the mineralogical phases. The three diffractograms exhibit peaks which characterize γ-nano alumina (2θ = 37.15; 46.15, and 66.75˚). These peaks correspond respectively to the (311), (400), and (440) planes on JCPDS file No. 29-63 [14] [35] [36] . The phases were therefore in a very poorly crystallized state given the amount of amorphous phase. The areas under the peaks were of similar sizes and do not really depend on the pre-heating temperature of the precursor. This can be explained by the further changes due to chemical reaction during the synthesis process. The so-called amorphous phase may be linked to the residual organic matter.

The zoom of hkl (440) peak has been assumed to be representative due to its highest intensity and used in the calculation of crystallite size ( Figure 5 ).

Table 2 shows the different values of the calculation parameters used to calculate the crystallite size Equation (6).

The (FTIR) spectra ( Figure 6 ) of the three synthesized products (NK600, NK700, and NK800) showed that the spectra were all similar and display characteristic adsorption bands of nano alumina [14] [15] [35] . Around 450 and at 1090 cm⁻¹ are the stretching vibration bands of the Al-O-Al bond. Around 1646 cm⁻¹ was located a peak related to the bending mode of physically adsorbed water molecules. At around 3358 cm⁻¹ were the stretching vibrations of bonded OH groups and isolated OH groups [8] [14] [35] . Despite the difference in the pre-heating temperature, the bands are similar showing that the bonds present in the three synthesized products are the same.

The SEM images of the alumina nanoparticles from precursor previously heated at 600˚C, 700˚C, and 800˚C presented in Figure 7 revealed that all the products have heterogeneous morphology, with large particles dominating in the case of NK600 and NK700, compared with NK800, where the particles were less large and closer. Moreover, the particles seem to be isolated from each other, independently of their respective sizes. These images were similar to those of alumina nanoparticles found by some authors using different precursors [3] [8] [27] . In addition, EDX analysis of the contrasts in each of the products clearly shows that the species present in high proportion are aluminum (Al) and oxygen (O). This suggests that the chemical substance present was aluminum oxide (Al₂O₃).

According to the surface morphology analysis, the specific surface area (SSA) of NK600, NK700, and NK800 were 141.7, 146.8, and 137.7 m²/g for precursor pre-treated at 600˚C, 700˚C, 800˚C respectively. The more reactivity of the precursor, higher the SSA of the nano alumina obtained was observed. Furthermore, the values were challenging with those found in the literature [3] [8] [12] .

Table 3 gives the synthesis yield from the three precursors. The synthesis yield, calculated through Equation (5) and presented in Table 3 , was found to be 17.88 % for the amorphization temperature of 700˚C compared with 14.4% and 12.24 obtained at 600˚C and 800˚C respectively. These values were well below those given by industrial precursors [14] [34] [35] . The results were in accordance with the respective reactivity of the precursors. Indeed, at the lower of 600˚C the clay material was not completely amorphized and the bonds between the tetrahedral and octahedral planes of the clay sheets were not sufficiently weakened to optimize oxide dissolution during acid leaching and promote good synthesis yield. One part of the chemicals (HCl and NH₄OH) might be used to dissolve the network.

The precursor obtained at 700˚C (MK700) was the most reactive due to the transformation of the clay material into primitive oxides [25] [32] . In fact, as amorphization was linked to dehydroxylation, it has been shown by some authors that residual hydroxyl groups, around 10% of the -OH present in metakaolinite after its formation, were gradually eliminated [16] [35] . In addition, infrared studies of kaolinite dehydroxylation kinetics have shown that it was the -OH of the less firmly bonded outer octahedral sheets that first react with each other, releasing around three-quarters of the water of constitution, followed by the elimination of the better-protected -OH of the tetrahedral and octahedral sheets. Whereas at 800˚C, the reorganization process has already begun, involving the alumina species, since some authors believe that the reorganization observed on the (TGA/DTA) thermogram has a combined effect resulting on the one hand, from the demixing of the system into Al₂O₃ rich domains and, on the other, from the diffusion of small, anisotropic particles at the grain boundary interfaces [20] [35] .

The metakaolinite in MK800 tends to give rise to a pseudo-spinel phase and the amorphous silica to a more crystalized form of quartz, or cristobalite. This leads to a more organized structure that finally decreases the reactivity of the precursor subsequently to useless consumption of reactants. Based on these results, the optimum temperature for amorphization of the material is 700˚C.

A comparison of the present nano alumina particles with other obtained from different precursors has been made and the results are presented in Table 4 .

According to Table 4 above, the nano alumina synthesized in this work exhibits a higher SSA compared with many other adsorbents obtained from chemicals reagents and this is in accordance with the smaller crystallite size. Also, it might be of great economic interest since it was obtained from low-cost available materials. Such good properties may broaden the applications fields of the product.