Insight into Structural Features of Magnetic Kaolinite Nanocomposite and Its Potential for Methylene Blue Dye Removal from Aqueous Solution

An in-depth understanding on the structural features of engineered magnetic adsorbent is important for forecasting its efficiencies for environmental clean-up studies. A magnetic kaolinite nanocomposite (MKN) was prepared using Malaysia’s natural kaolinite via co-precipitation method with a three different clay: iron oxide mass ratio (MKN 1:1, MKN 2:1 and MKN 5:1). The morphology and structural features of the magnetic composites were systematically investigated using techniques, such as: Fourier transform infrared spectroscopy (FTIR), scanning electron microscope (SEM), surface area analysis (BET), Vibrating Sample Magnetometer (VSM), and zeta potential measurement. The removal efficiencies of the adsorbent for Methylene Blue (MB) dye were studied in batch method as a function of pH and initial concentration. MKN1:1 demonstrated the highest magnetisation susceptibility (Ms) of 35.9 emu/g with four-fold-increase in specific surface area as compared to the pristine kaolinite. Preliminary experiment reveals that all MKNs showed almost 100% removal of MB at low initial concentration (<50 ppm). The spent MKN adsorbent demonstrated an easy recovery via external magnetic field separation and recorded maximum adsorption capacity of 18.1 mg/g. This research gives an insight on the surface characteristics of magnetic clay composite for potential application as an effective and low-cost adsorbent in treating dye contaminated water. Copyright © 2022 by Authors, Published by BCREC Group. This is an open access article under the CC BY-SA License (https://creativecommons.org/licenses/by-sa/4.0).


Introduction
Untreated dyes effluent has become a major source of water pollution especially in developing countries. Methylene blue (MB) is an example of cationic synthetic dye commonly used in the textile and pharmaceutical industry [1]. Exposure to excessive MB could cause harmful ef-Clay minerals like bentonite, palygorskite, and kaolinite have received great interest either as a catalyst or pollutant adsorbent due to its non-toxicity, inexpensive and high availability [3,5,6]. Malaysia has approximately 112 million tons of kaolinite reserve [7,8] with a total export value of Malaysian Ringgit 26 million (approximately USD 6 million) in 2013 [8].
Under the Malaysia's National Mineral Industry Transformation Plan 2021-2030 (MIT 2021-2030 Plan), the government aim to prioritise local minerals like kaolinite over imported mineral commodities for domestic industries and research and development (R&D) [9]. The local kaolinite is mostly used for paper coating, skincare products and engineering applications, whereas only a few scientific works reported its use as high value industrial products [7] including as material for water remediation. As a non-expanding clay, kaolinite in general have a lower adsorption capacity for removal of water contaminants as compared to other swelling clays like montmorillonite. Nevertheless, research works on kaolinite's modification, especially in developing countries, are wellprogressing and have demonstrated a substantial efficiencies improvement in environmental clean-up studies [1,10] that worth for further investigations.
The development of hybrid material like magnetic clay composite for water treatment application is blooming fast due to the attractive surface properties, less complicated preparation method and facile separation capabilities [11,12]. With the incorporation of magnetic particles like iron oxide into the clays' gallery or surface [13], the magnetic clays are easy to be separated from the reaction media through the external applied magnetic field. This modification can address the current recovery issue for spent clay adsorbents which require conventional filtration or centrifugation method that are either cost-or time consuming.
In recent literature, scientific works that carefully discuss the suitable composition ratio required for the synthesis of hybrid composite material are still lacking. Selecting a good compositional ratio is important from both structural and efficiencies perspective. The knowledge will ensure a balance engineering strategy between achieving the desirable structural features (e.g., strong magnetic susceptibility and high surface area) and maximising its performance (high percent removal of pollutant or adsorption capacity). This research aims to investigate the morphology and structural features of the magnetic kaolinite composite prepared via co-precipitation method with different kaolinite: iron oxide mass ratio. Spectroscopic and macroscopic techniques will be used to characterise the composites. The performance of magnetic composites against Methylene Blue removal will be investigated.

Synthesis of Magnetic Kaolinite and Iron Oxide
The iron oxide (IO) particle was prepared by dissolving a 3.1 g of FeCl3.6H2O and 2.4 g of FeSO4.7H2O in 50 mL of deionized water, separately [14]. Next, the Fe 2+ solution was added into Fe 3+ solution then was placed into a water bath and agitated (200 rpm) at 60 ℃ in 30 min. The agitation speed was further increased (400 rpm) and NH4OH (25%v/v) was added dropwise into the suspension until the solution turned to alkaline (pH 8.9 to 9.1) to precipitate the iron hydroxide. The suspension was aged for another 1 hour at the same temperature. The precipitation was collected using filtration and was thoroughly washed with deionized water and ethanol until a neutral pH was achieved. The IO precipitate then was dried at 110 ℃ for 3 hours, ground and sieved (200 mesh).
The preparation procedure of MKN was similar with the preparation of IO as described earlier, except that kaolinite suspension is added into the Fe 2+ -Fe 3+ suspension via in situ approach according to our previous work [15]. Three calculated mass ratios of kaolinite: iron oxide which were 1:1, 2:1 and 5:1 was chosen to prepare the MKNs; later denoted as MKN 1:1, MKN 2:1 and MKN 5:1, respectively. First, a specific amount of kaolinite was added into the Fe 3+ solution at 40 ℃ for 15 minutes. The Fe 2+ solution was then added to the kaolinite -Fe 3+ suspension and agitated. The addition of NH4OH (25%v/v), aging, collection of MKN and the drying process were as similar as the procedure in the preparation of IO described earlier.
The black precipitate of MKNs were ground, sieved (200 mesh) and kept in a dark and airtight container for further use.

Characterizations
Fourier-transform infrared spectroscopy (FTIR) spectrum was collected using FTIR-Spectrum 400 spectrometer in the range of 4000 cm −1 -450 cm −1 at resolution of 4 cm −1 with 32 scans using the KBr pressing method. KBr pellets were prepared with a 1:200 mass ratio of sample and KBr, respectively. The Xray powder diffraction XRD analysis was performed using D8 Advance X-Ray diffractometer (Bruker AXS, Germany) with Cu Kα radiation (λ = 0.15406 nm) and the 2θ range from 5 to 80°. Scanning electron microscopy (SEM) images were acquired on a Quanta 450 FEG operated at 15 kV accelerating voltage to observe the morphology and estimate the particle size of samples. The samples were prepared by direct deposition on an aluminium stub covered by a carbon grid and then coated with a thin layer of platinum (~10 nm thick film). The Energy Dispersive X-Ray (EDX) analysis was performed simultaneously to determine the elemental composition (Al, Si, O, Fe). To determine the surface porosity, the N2 adsorptiondesorption isotherm was performed at −195 ℃ using Sorptometric1990 series. The Brunauer-Emmett-Teller (BET) method was used to calculate the specific surface area and pore size of samples. Meanwhile, Barrett, Joyner and Halenda (BJH) method was used for the pore volume and size distribution. Prior analysis, a 20 mg sample was placed into a clean and dry tube for degassing at 90 ℃ for 1 hour and then increase to 180 ℃ for 4 hours [16]. The weight of the sample before and after degassing will be recorded. Meanwhile the magnetization measurement of MKN was carried out using 7400 Series VSM System (Lake Shore, Ohio). The hysteresis measurements were performed at 300 K with magnetic field up to 0.9 T to obtain the demagnetisation corrections. Zeta potential measurements were performed using the Zetasizer Nanoseries instrument (Malvern, United Kingdom). The suspensions were prepared by mixing a 0.15 g sample in 15 mL electrolyte (deionized water) with selected pH val- ues (pH 2 to 10). Next, the samples were homogenized for 1 hour at room temperature prior analysis and measurements were made in triplicates. The average value was reported.

Adsorption Studies
The Methylene Blue (MB) removal was studied using the batch equilibrium adsorption method. A stock solution (300 ppm) of MB was prepared in the deionized water followed by subsequent dilution. The experimental parameters such as initial concentration and pH were optimised to determine the equilibrium concentration. Pristine kaolinite was used as the control. A 0.125 g amount of adsorbate in MB solution (25 mL) of desired initial concentration (5 to 150 ppm) was agitated at 130 rpm for 4 hours. Throughout the experiment, the samples were fully wrapped in foil to keep the dark condition for avoiding any potential lightdegradation.
After reaction completion, the adsorbent was separated by an external magnetic field using a magnetic bar. The dye concentration of supernatant was determined using the UV-Vis spectroscopy (T80+, PG instrument Ltd). The removal efficiency, E (%) was calculated (Eq. 1), as well as amount of dyes adsorbed, qe, (Eq. 2), where the [M]i and [M]f denotes the concentration of MB before and after treatment (mg/L), respectively, W is the mass of adsorbent (g), and V is the volume of the solution (L). (1) The adsorption equilibrium data were then fitted to the Langmuir and Freundlich adsorption isotherm model.

Characterization of Adsorbent
The FTIR spectra of raw kaolinite, iron oxide and magnetic kaolinite composites are presented in Figure 1(a). Both the raw kaolinite and magnetic kaolinite nanocomposite (MKN) showed two bands at 3695 cm −1 and 3620 cm −1 , which represent the Al−OH stretching vibration in the clay's structure [17][18][19][20]. These bands were less intense in the MKNs spectrum indicating the possible exchange of Al 3+ with Fe 3+ on the surface of kaolinite [21]. Furthermore, acidic Fe 2+ /Fe 3+ solution during the synthesis of MKN might promote partial Al dissolution from clays' gallery thus allowing the interchange of cations. The spectral bands at 1112 cm −1 and 1006 cm −1 in raw kaolinite denotes the Si−O bending and stretching vibrations [15,19,22]. Meanwhile the band appeared at 910 cm −1 was corresponded to the Al−O bending vibration [19,20], whereas the band at 795 cm −1 and 755 cm −1 represents the Si−O−Al stretching and bending vibrations, respectively [17,20]. In MKN, those bands were weaker in contrast to those observed in raw kaolinite which suggest the interaction of Al/Si−O of kaolinite with Fe−O bond for MKN's formation [3]. The Al−O−Si skeletal vibrations in kaolinite were denoted at 530 cm −1 and 460 cm −1 [19]. The spectrum of MKN2:1 and MKN5:1 showed a sharper band at 460 cm −1 than those in the MKN1:1 due to the larger kaolinite content in the former composites. For the iron oxide spectrum, the bands at 627 cm −1 to 542 cm −1 were associated with the Fe−O bond stretching vibration [23]. The band at 627 cm −1 was almost diminished especially in MKN5:1 due to limited Fe−O−Al or Fe−O−Si bond formed within the composite [23]. Overall, FTIR analysis shows that all important bands of functional groups from the pristine material (kaolinite and iron oxide) were well-preserved in the composite. Meanwhile, the MB-loaded MKN (Figure 1(b)) showed a broad band at 3432 cm −1 associated with hydroxyl bonding with nitrogen (O−H---N) in MB molecule [2,24]. The new band at 1625 cm −1 (Figure 1(b)) could be corresponds to the vibrations of unsaturated bond of C=N + (CH3)2 previously observed in MB spectra (Figure 1(c)). In addition, a strong intensity band at 1591 cm −1 (referring to C=C or C=N bonds) [25] and at region 1110 to 1006 cm −1 (heterocycle skeleton (C−N and C−S−C)) from MB dye was also present in MB-loaded MKN (Figure 1(b)) [24,25]. The presence of these bands in MB loaded-MKN hence attributed to the successful adsorption of MB towards the composites' surface.
All MKNs demonstrated a fast magnetic separation (in less than 120 seconds, on average) as compared to kaolinite which need a high-speed centrifugation to settle down. The plot of IO and MKN produced a "S" shaped curve (Figure 2) which shows the near-superparamagnetic characteristic [26]. The magnetization (Ms), retentivity (Mr), and coercivity (Hc) value was shown in Table 1. The decrease of magnetic strength is attributed to the nonmagnetic characteristic of kaolinite, which is typical for most clays. All composites demonstrated a magnetic susceptibility value at par or above the previously reported value for magnetic clay [2,12]. The MKN1:1 composite showed the highest Ms among others due to a higher magnetic iron oxide content. Nevertheless, the non-zero coercivity value indicates the existence of magnetic storage, variation in particles size and the effect of cluster growth of iron oxides particles [27]. For example, although the Ms value of MKN5:1 was almost one-third of those obtained by MKN1:1, their difference in coercivity value was minor; suggesting an almost consistent particle size in each composite.
The surface area analysis demonstrated that the kaolinite has the lowest porosity but with the largest pore diameter ( Table 2). According to the IUPAC classification, both kaolinite and iron oxide demonstrated a type IV isotherm (Figure 3(a)) [3,28] usually associated with the mesoporous structure. In addition, the pristine materials exhibit a H3 type hysteresis loop related to a slit-shape pore. These surface porosity features of pristine Kao and IO are well preserved within the MKNs (Figure 3(a)).  The high BET surface area of MKN 1:1 is contributed by the iron oxide nano-features evidenced by the high dV/dw pore volume of iron oxide in the pore size distribution (Figure 3(b)).

The incorporation of porous iron oxide in MKNs
has resulted up to 4-fold increment of surface area (MKN 1:1) as compared to the pristine kaolinite. This finding hence signified the role of iron oxide in enhancing the surface porosity of clays' composite. Meanwhile, the zeta potential measurements were used to describe the surface charge and the elucidation of the adsorption mechanism especially for electrostatic interaction [14,29]. Kaolinite has an isoelectric point (IEP) at around 4.6 ( Figure 4). Kaolinite was known to have a positive surface charge at low pH but shifted to a negative surface charge as the pH increased [30]. The IEP for IO obtained in this study is relatively high (around 8.2) (Figure 4), but still closer to the previously reported value [31]. The MKNs recorded a different IEPs; in which at around pH 5.3 (MKN 1:1), pH 8.5 (MKN 2:1) and pH 3.7 (MKN 5:1). The zeta potential values of MKN are a mixed and ''inbetween'' [31] of the zeta potential curve for the positive iron oxide surface and the neu-   Figure  4). As more kaolinite component was present in MKN 5:1, this composite exhibits the least positive surface charge. However, at above pH 4, a variety surface charge profile was observed. For example, at pH 6 and 8, only MKN 1:1 exhibited a definite negative surface charge; despite it has the least kaolinite ratio as compared to the MKN 2:1 and MKN 5:1. This variation narrated the non-uniform surface chemistry in MKNs that is differs from the pristine components. The scattered and uneven distribution of iron oxide within the clays' surface could possibly cause a different protonationdeprotonation behaviour of the surface functional groups in the MKNs composite.
The SEM images of kaolinite ( Figure 5(a)) shows a smooth-edged surface with irregular platelets. Meanwhile, for IO, a uniform particle distribution was observed with the average particle size measured is between 30 to 35 nm ( Figure 5(b)). In MKN 1:1 ( Figure 5(c)), the IO particles were distributed homogeneously in clusters on the kaolinite's surface, but the accumulation did not entirely cover the clay's surface. These observations were similarly reported by other researchers, suggesting the deposition of iron oxide only occurs on kaolinite surface only [3,32]. The estimated particle size of IO embedded on the kaolinite surface is 39±2 nm. The presence of IO on kaolinite could also be confirmed from the EDX analysis of Fe element as shown in Figure 5(d).

Adsorption Studies
At initial concentration of 5 to 50 ppm, a 100% MB removal was recorded for all adsorbents ( Figure 6). However, as the initial concentration increased, the dye removal efficiencies were decreasing. The MKN 1:1 showed the highest percent removal among all the composites. The primary adsorption mechanism of MB towards MKN are postulated to occur via electrostatic interaction [1,3,33] between the negative surface charge of MKN (Figure 4) towards the cationic MB molecules. The individual performance of kaolinite and IO against the MKN 1:1 was examined to deduce which component contribute the most towards the composites' dye removal efficiencies. Despite the kaolinite has a much lower specific surface area (Table 2), it recorded a higher percent removal (from 60 to 100%) as compared to the poorly performed IO (< 50%) ( Figure 6). Thus, Figure 6 Percent removal of MB by all adsorbents at pH 6.  it is concluded that in terms of surface reactivity of MKN, the active sites in kaolinite could play the most significant role rather than those contributed by the iron oxide. Similar conclusion was also proposed by previous work involving magnetic bentonite [33]. The detailed molecular mechanism of dye degradation by iron oxide is still unclear [34] due to the complex interfacial adsorption and surface properties of the oxide shell. Further works on kinetics, thermodynamics, and atomic-level spectroscopic investigations of MKN are necessary to assist in elucidating the precise mechanism. However, at this stage it is still noteworthy that the presence of iron oxide in MKN has effectively facilitated the recoveries of spent clays which overcomes the current separation issues for largescale operation. The MKN:1 demonstrates the highest maximum adsorption capacity value (18.1 mg/g) among all composites ( Figure 7 and Table 3). A high shoulder on the adsorption isotherm also indicates a strong interaction between the adsorbents and the Methylene Blue ( Figure 7). Experimental data of all adsorbents were best fitted to the Langmuir model attributed to the high regression coefficient value, r 2 > 0.98 (Table 3). The fitness to Langmuir model indicates the monolayer adsorption characteristics and homogenous active sites [32], as opposed to the Freundlich model (multilayer adsorption and heterogonous surface sites). Meanwhile, the RL value was found to be in the range of 0 to 1, which implies a favourable MB dye adsorption (Table 3).

Isotherm
Although kaolinite has the lowest BET surface area (Table 2), its adsorption capacity is almost at par (11.7 mg/g) with other MKNs especially MKN 1:1 (18.1 mg/g) ( Table 3). It is hence proposed that the surface active sites and surface porosity are both important in improving the dye removal capability of MKN. This finding is in parallel with the influence of a negative surface charge in enhancing the efficiencies of kaolinite and MKN 1:1 ( Figure 4) as discussed earlier. Previous work also suggested that the surface porosity facilitates towards a higher maximum adsorption capacity for modified kaolinite [32]. However, for IO, particles aggregation and less abundance active sites may responsible for its low performance despite having the highest BET surface area ( Table 2).
In terms of adsorption capacities, the performance of Malaysia's kaolinite and MKN were satisfactory considering the wide range of dyes' initial concentration used (Table 4). Optimisation of experimental conditions, (i.e. adsorbent loading and pH influence), are necessary to further evaluate the overall performance of MKN. Ascribed to its high availability, simple synthesis procedure and feasible recovery through the external magnetic field, this magnetic composite shall have great potential as a   promising low-cost adsorbent for dye remediation process.

Conclusions
A magnetic kaolinite nanocomposite (MKN) with different kaolinite: iron oxide mass ratio was successfully prepared through a simple coprecipitation method. A higher iron oxide content in MKN has enhanced the BET surface area and its magnetism susceptibility, but less significant to boost the adsorption efficiencies towards Methylene Blue. Thus, for future synthesis of magnetic composite, an optimum clay: iron oxide composition should be considered for sufficient availability of surface active sites, without sacrificing its magnetic properties for a feasible magnetic separation. All MKNs have a 100% MB removal in which MKN 1:1 showed the highest maximum adsorption capacity (18.1 mg g −1 ). Further spectroscopic, thermodynamic, and kinetic modelling studies are needed to elucidate the MB removal mechanism by MKN. This study provides new knowledge on the potential of magnetic clay nanocomposite derived from the natural local clay minerals for the treatment of dye contaminated water.