The raw sampled clay was wet sieved at 40 μm, air-dried, crushed and sieved at 40 µm to give BN (Bana). The capacity to be exchange of cations in the interlayer space of smectic clays depends on its hydratation degree; this later itself depends on ionic ray. The most the cation has a small ionic ray, the most its exchange is easy. After Lithium, Na⁺ has a small ray and literature suggest that its echangeability by others cations will be easy. [7] We proceed to Na-saturation of Bana to maximise intercalation and release of Ciprofloxacin. 40 g of BN were introduced into 240 ml of 1 M NaCl and then stirred for 8 hours; operation was repeated four times. Then repeated washing with deionized water and centrifugations at 3500 rpm for 40 min was carried out until the supernatant water is free of chloride ions as tested with silver nitrate solution. The obtained cake is dried in oven at 105˚C for 24 hours, crushed and sieved at 40 µm to give BNNa (Bana saturated with Na⁺). For being around the clay fraction, with an assay of elimination of non-clay fraction, an extraction of the ≤2 µm has been done. Another 40 grams of BN were also saturated with NaCl and introduced into 200 mL test tubes containing deionized water; thereafter, ≤2 μm particles was extracted after sedimentation pre-calculated time according to the Stockes’ law. [21] [22] The suspension obtained undergoes extensive centrifugations at 3500 rpm for 1 hour; the slurry is dried at 105˚C for 10 h, ground and sieved at 40 µm to obtain BNNa- F (fine fraction of Bana preceded by Na⁺ saturation).

As quoted in section 2.1, CFX positive form is achieved at pH < 5.9; hence, the intercalation was implemented in batch mode at room temperature in a 10⁻² M HCl with an average medium pH of 2.5. In view to optimize the process, contact time and the initial amount of CFX effect have been calculated; the various operations during the test were replicates three times.

1) Variation of the contact time

The effect of the stirring time on the amount of adsorbed CFX is evaluated by introducing 100 mg of BN, BNNa or BNNa-F in 50 mL of CFX 600 mg/L under a stirring of 500 rotations per minute (rpm) with stirring times between 5 and 240 min. At each time, pre-calculated volume of solution is collected, filtered through PTFE (polytetrafluoroethylene) filters of 0.2 µm and then diluted for the determination of the equilibrium (residual) amount of CFX by using UV-visible spectrophotometry at λ_max = 277 nm.

2) Adsorption isotherms

150 mg of BN, BNNa or BNNa-F are introduced into 50 ml of CFX concentrations range 200 to 1000 mg/L. The mixture is stirred at 500 rpm for equilibrium contact time at room temperature and samples are filtered through PTFE filters of 0.2 μm. Maximum amount of CFX adsorbed is determined by measuring residual CFX using UV-visible spectrophotometry at λ_max = 277 nm.

Characterizations will approve the smectitic nature of the materials used and confirm CFX intercalation.

X-ray diffractometry (XRD) was performed on powders using a Bruker diffractometer under the following conditions: CuKα radiation, 40 KV voltage, 40 mA intensity and scanning angle from 3˚ to 70˚ by steps of 0.02˚ with a duration of 1 s.

Fourier-transform infrared spectroscopy (FTIR) was performed in ATR mode at 4000 - 600 cm⁻¹ on ZnSe crystal using a PerkinElmer infrared spectrophotometer.

The thermo-gravimetric analysis (TGA) was carried out using thermo-gravi- meter Q5000 between 20˚C and 800˚C under an oxygen atmosphere at a heating rate of 5˚C/min and a flow rate of 25 mL/min of oxygen.

The specific surface areas of the samples were determined from nitrogen adsorption and desorption isotherms at 77 K, obtained from Micrometrics TriStar II 3020 V1. The specific surface area (SSA) was calculated from adsorption data by applying the Brunauer-Emmet-Teller (B.E.T) method, which only measures the external surface accessible by molecules.

BN, BNNa, BN-CFX and BNNa-CFX samples are sterilized under UV radiation for 15 min; the samples (5 mg) were released in 10 mL of PBS, under agitation at 80 rpm at 37˚C. At times 1 h, 2 h, 4 h, 6 h, 24 h, 48 h, 72 h, 96 h, the media of each well is taken by syringe and filter needle (BD Blunt filter Needle, Beckton Dickinson). The collected medium is renewed by 10 mL of fresh PBS, using the needle that was used to rinse it, and restocked to 80 rpm at 37˚C. Samples were analyzed for CFX content by HPLC at λ_max = 270 nm. These studies were performed in triplicate for each sample and the average values were reported.

The sample mediums (BN, BNNa, BN-CFX and BNNa-CFX) have been analyzed in microbiology by a diffusion test.

9 mm diameter Petri dishes containing 18 mL Muller Hinton Agar are seeded with 0.1 mL of bacterial suspension at 10⁴ cfu/mL (E. coli K12 or S. aureus CIP224). Carrots of 6mm diameter are made in the agar to obtain wells allowing the deposit of 50 µL of solution to be tested.

Petri dishes are incubated at 37˚C for 24 hours. After 24 h of incubation, the seeded bacteria yielded a homogeneous bacterial mat, and the solutions provided an inhibition zone with a diameter related to the amount of active ingredient contained in the solution tested. The diameter of these inhibition zones is measured, and graphically expressed over time.

Figure 1 shows the X-ray patterns of clays with basal reflection d₀₀₁ at 14.10 Å,

12.33 Å and 12.41 Å respectively for BN, BNNa, BNNa-F assigned to montmorillonite. [23] The difference between basal reflections at 14.10 Å of BN with 12.33 or 12.41 Å of BNNa and BNNa-F could be explained by hydrated radius of naturally intercalated cations of BN in comparison of Na⁺ in homoionic forms. Associated minerals are kaolinite (7.20 Å), mica (4.47 Å), quartz (3.34 Å), anatase (3.51 Å) and orthoclase (3.78 Å). [19] [24] BNNa-F associated mineral are less important probably due to their elimination during the process of fine fraction extraction. BN, BNNa and BNNa-F are smectites clays according to their X- ray patterns. [19]

Figure 2 below presents the visual aspect of Clay-CFX loaded compounds prepared; this aspect is similar to clay mineral.

In Figure 3, it can be clearly seen that basal distance at 14.10 Å of BN rises to 15.71 Å for BN-CFX, the one at 12.33 Å of BNNa rises to 16.01 Å for BNNa-CFX and that at 12.41 Å of BNNa-F rises to 16.52 Å for BNNa-F-CFX. The increase of basal spacing could be due to intercalation of CFX by ionic exchange between the hydrated interlayer cations of clays and CFX molecules [11]. The increase in the basal distance for clay-CFX is 1.61 Å, 3.68 Å and 4.11 Å respectively for BN-CFX, BNNa-CFX and BNNa-F-CFX. BNNa and BNNa-F possess in their interlayer a more readily exchangeable cation Na⁺, while the raw BN may have more tightly linked cations in its interlayer that could limit the CFX intercalation because Na⁺ favours ionic exchange in interlayer space. Cationic exchange capacity (CEC) represents the ionic capacity of interlayer cations to be exchange by others cations and Na-saturation of clay increase this CEC. [8] So, after Na-saturation, CEC of BNNa and BNNa-F become certainly high than raw BN cationic capacity exchange. Which precede could be explain the higher basal spacing variation after CFX loading of BNNa and BNNa-F than BN. The thickness of montmorillonite sheet is 10 Å [25] and the theoretical estimated value of longitudinal molecular length of CFX is 8.5 Å while its lateral molecular length of CFX is 11.9 Å [16]. The interlayer space represents the basal spacing minus

the layer thick. The experimental values of interlayer space of BN-CFX, BNNa- CFX, BNNa-F-CFX are 5.71 Å, 6.01 Å and 6.52 Å respectively. Those values are smaller than the theoretical estimated value of longitudinal molecular length of CFX (8.5 Å). This presumes that after intercalation, CFX in interlayer space has been compressed by electrostatic interactions in clays. [16] The variation of the basal spacing, indicative of CFX intercalation within the clay platelets will be confirm by infrared spectroscopy results.

Figure 4 shows the infrared spectra of BN, BNNa, BNNa-F (a), CFX (b and c). From raw material spectra (Figure 4(a)), the vibration band at 3700 cm⁻¹ corresponds to the elongation vibration of the -OH bond characteristic of phyllosilicate. [26] The band at 3623 cm⁻¹ is characteristic of internal –OH clays structure and particularly for kaolinite. The 3381 cm⁻¹ band corresponds to the stretching

vibration of -OH bond of hydration water; this band is better defined in BN than BNNa or BNNa-F because in those latter, some water molecules were displaced during saturation with NaCl. The peak at 1630 cm⁻¹ is attributed to in plane deformation of adsorbed water molecule.The bands at 1118 cm⁻¹ and 1024 cm⁻¹ are for stretching vibrations Si-O bond of phyllosilicates. [10] [27]

In Figure 4(b) and Figure 4(c), important CFX characteristic bands are mostly between 1700 cm⁻¹ and 1200 cm⁻¹. The band at 1613 cm⁻¹ corresponds to C=O bond of quinolones and peak at 1585 cm⁻¹ is due to asymmetric stretching vibration of COO⁻ bond; peak at 1540 cm⁻¹ is attributed to stretching vibration of aromatic C=C. Bands at 1485 cm⁻¹ and 1440 cm⁻¹ represent respectively stretching vibration of C-N and inclination vibration of -C-H or -C-O. Peak at 1372 cm⁻¹ corresponds to protonation of amino group in piperazine part. The 1260 cm⁻¹ band is assigned to the coupling of -C-O elongation vibration with O-H deformation of COOH. [28] [29] [30]

Due to the similarity among the IR spectra of all loaded compounds, only BNNa-F spectra and its loaded compounds BNNa-F-CFX have been presented in Figure 5 together with that of CFX. For better visualization of changes occurred, spectra have been zoomed in CFX functional domain 2000 - 1000 cm⁻¹. On the spectra of loaded materials, appearance of new bands is observed. The peak at 1613 cm⁻¹ (C=O stretching vibration) of CFX shifted to 1630 cm⁻¹ and mix up with the one (-OH inclination vibration) already presents in clay. This mixing made the peak more intense. Also, the shifting from 1613 cm⁻¹ in CFX to 1630 cm⁻¹ in the loaded sample is associated to strong interactions between C=O of CFX and the oxygen atoms of the basal surface of the clay materials. [31] The vibration at 1580 cm⁻¹ attributed to carboxylate COO⁻ disappears after adsorption certainly because of the loaded compounds elaboration conditions (pH < 5.9) which induce the cationic form of CFX. The band at 1540 cm⁻¹ is attributed to stretching vibration of aromatic C=C; peaks at 1485 and 1460 cm⁻¹ designer respectively stretching vibration of C-N and inclination vibration of -C-H or –C-O. The displacement of the protonation vibration of amino group from 1372 cm⁻¹ to 1386 cm⁻¹ suggests existence of electrostatic attractions between protonated amino group and negative surface charges. The 1260 cm⁻¹ band moves to 1273 cm⁻¹ because of the strong hydrogen bonds between COOH and the hydroxyl surface group of montmorillonite in an acidic medium. [11]

The appearance of CFX characteristic bands on the loaded compounds proves the CFX adsorption on the raw materials and indicates intercalation of CFX in the interlayer space of raw materials.

Figure 6(a) shows dTG/TG curves of clay materials. Between 20˚C and 150˚C, mass loss of 12.11%, 4.8% and 8.5% respectively for BN, BNNa, BNNa-F is attributed to loss of water moisture and water bound to interfoliar cations. [32] This loss is lower in Na-saturated materials than BN because of reduced intercalated water molecules as proposed from XRD. Around 430˚C for BN, 470˚C for

BNNa and BNNa-F respectively, a maximum mass loss of 6.3%, 6.4% and 6.4% correspond to de-hydroxylation of clay minerals up to 700˚C. [32] The mass loss is almost the same, which means that clay content in the prepared materials have the same structure. For CFX (Figure 6(b)), it is observed that maximum decomposition occurred at two different temperatures (300˚C and 400˚C). It is proposed that at 300˚C the less stable functions such as carboxylic and amino-basic are decomposed and the most stable phenyl cycle are decomposed later around 400˚C.

Figures 7(a)-(c) show the dTG/TG curves of clay-CFX materials. The CFX decomposition is marked by two peaks temperature around 300˚C and around 400˚C (413˚C, 395˚C, 420˚C respectively for BN-CFX, BNNa-CFX and BNNa-F- CFX). Those two stages of decomposition are observed for pure CFX (see Section 3.1.3-c). Around 300˚C, decomposition of CFX is accompanied by mass loss of 17%, 18% and 19.5% respectively for BN-CFX, BNNa-CFX and BNNa-F-CFX; finally, around 560˚C, mass loss could be due to complete disintegration of CFX- clay structure. [1]

The comparison of the mass loss is coherent with the observed changes in basal distance from XRD which indicates increase intercalation as follow BNNa- F-CFX > BNNa-CFX > BN-CFX. The same temperature of CFX-clay structure disintegration (560˚C) suggests that BN-CFX, BNNa-CFX and BNNa-F-CFX have the same structure.

As the total mass loss can be associated to CFX decomposition, the estimation of CFX mass loss at 700˚C was calculated. The experimental CFX quantities have been compared to UV estimation of CFX fixed by the clay. In this way, 146 mg/g, 153.4 mg/g, 156.9 mg/g of CFX adsorbed obtained by UV estimation respectively for BN-CFX, BNNa-CFX, BNNa-F-CFX is closed to 133.5 mg/g, 147.4 mg/g, 150.7 mg/g of CFX adsorbed calculated by TGA methods. The obtained results are in agreement and the UV-visible values can be use latter for others calculations.

The N₂ adsorption-desorption isotherms of the natural clays on Figure 8 show very similar shapes with presence of a hysteresis that indicates existence of mesopores. [19] The measured specific surface area of BN, BNNa, BNNa-F were about 39.3 m²/g, 42.8 m²/g, 50.1 m²/g respectively. The high value for BNNa-F could be explained by the fact that this sample in the fine fraction is enriched in clay minerals. Those values of the specific surface area (SSA) are confirms the high smectite mineral content in clay. The high SSA area is important factor when the clay minerals are used in adsorption experimentations. [19]

Only BNNa/BNNa-CFX adsorption-desorption shapes are reported for comparison. Measured specific area decreases from 42.8 m²/g for BNNa to 16 m²/g for BNNa-CFX apart of Figure 9; this decreasing could be attribute to occupation of BNNa-F pores by CFX molecules during adsorption process, expressive of CFX presence on loaded compounds.