Cross-Linked Chitosan-Glyoxal/Kaolin Clay Composite: Parametric Optimization for Color Removal and COD Reduction of Remazol Brilliant Blue R Dye

Kaolin clay (KN) was employed as an inorganic filler to modify a cross-linked chitosan-glyoxal as Schiff’s-based chitosan composite derivative (CTS-GLY). The resulting (CTS-GLY/KN) was found to be a promising composite synthetic biopolymer that can be potentially utilized for color removal as well as COD reduction of an industrial anionic dye (remazol brilliant blue R, RBBR). The surface porosity, crystallinity, morphology, functionality, charge, and amine content of the CTS-GLY/KN were studied using BET, XRD, SEM, FTIR, pHpzc and pH-potentiometric titration analyses, respectively. Response surface methodology-Box-Behnken design (RSM-BBD) was used to optimize the impact of the main input factors on the color removal and COD reduction of RBBR. The adsorptive performance CTS-GLY/KN towards RBBR was well-defined by both Langmuir and Freundlich isotherm models with highest adsorption capacity of 447.1 mg/g at 30 °C. This finding reveals that CTS-GLY/KN can be utilized as a promising, feasible, and environmentally friendly composite-biosorbent for color removal and COD reduction of textile dyes from aqueous medium.


Introduction
The considerable usage of organic dyes and pollutants in numerous industries such as textiles, pharmaceuticals, rubber, and paper generate a vast number of effluents containing dyes [1]. The release of the effluents into water bodies causes significant variations to appear in water attributes including color, pH, and chemical oxygen demand (COD), as well as detrimental effects on human health [2]. Severe dye residues are one of the significant attributes of textile wastewater, which can prevent sunlight penetration into water bodies resulting in reduction in photosynthetic activities of aquatic life [3]. One of the most frequent pollutants is COD, indicating the existence of organic compounds in wastewater [4]. To avoid risking the survival of aquatic life by exhausting the oxygen resources, it is essential to reduce COD levels and meet the allowed discharge threshold [5]. Remazol Brilliant Blue R (RBBR) dye is a reactive dye commonly used as an initial ingredient for generating polymeric dyes in textiles [6].
RBBR can be considered a good model for toxic and recalcitrant organic pollutants that pose a threat to ecosystem and humans [7]. Accordingly, the dye removal and COD reduction from wastewater prior to release into the water bodies is a necessary action to maintain both the humans and the environment healthy. Conventional wastewater treatment methods such as adsorption [8], nanofiltration [9], and ozonation [10] have been utilized for the removal of organic dyes and to attain reduction in COD levels of the wastewater. Simple operation design, non-sensitivity to pollutants, recyclability, being highly efficient, cost-effective, and lesser waste residuals has made adsorption among the most employed methods [11].
Chitosan (CTS) is a cationic polysaccharide of β-(1-4) acetyl-D-glucosamine units and derived from deacetylating chitin, found in the exoskeleton of crustaceans, insects etc. [12,13]. CTS reserves excellent properties, such as biodegradability, hydrophilicity, high chemical reactivity, biocompatibility, chirality, adsorption capacity, non-toxicity, anti-bacterial activity, and chelation [14,15]. CTS is featured in the existence of amino (-NH 2 ) and hydroxyl (-OH) groups in its polymeric chain, acting as the prospect adsorption sites for removal of metal ions and anionic dyes [16,17]. However, natural CTS is sensitive to pH changes, mechanically unstable, susceptible to swelling, and leachable in aqueous environment [18]. Hence, it is of great significance to improve the physiochemical properties of CTS.
Many approaches have been carried out to improve the physicochemical properties of CTS such as crosslinking reaction [19], grafting with amino functional groups [20], and compositing with nanomaterials [21]. Amongst the approaches, crosslinking reaction is the most convenient pathway for developing attributes of CTS such as its mechanical strength, solubility in acidic medium, chemical stability and hydrophobicity [22,23]. Many studies have been carried out on utilizing crosslinked CTS derivatives as adsorbents for removal of dyes from aqueous solutions [24,25]. Furthermore, CTS-based composite is considered as one of the most desired approaches for improving the surface and structural properties of CTS [26]. Consequently, several decent fillers have been introduced for preparation of CTS-based composite biopolymer to enhance physiochemical and adsorptive properties. These fillers include clay [27], metal oxide nanoparticles [28], semiconductor [22], montmorillonite [29,30], carbon nanotubes [31], and silica [32].
Kaolin clay (KN) is one of the most common natural inorganic clays, existing in rocks copiously in crystalline structure [33]. KN has outstanding properties (e.g., porosity, low-cost, thermal stability, abundant availability, and non-toxicity) in environmental applications when employed as adsorbent alone and/or within a CTS composite, such as good adsorption capacity, high surface area, easy to functionalize its surface, high thermal stability, and environmentally friendly [34]. Recent studies show that CTS and KN complement each other for producing multifunctional biocomposite adsorbents, which can be effectively employed for removal of numerous water contaminants including organic anionic dyes [27], biological contaminants [35], organic cationic dyes [36], and nutrients [37]. From the environmental standpoint, utilizing kaolin clay powder as a filler or surface modifier to chitosan will improve the surface area and adsorptive property of chitosan, and will reduce the cost and environmental impact of solid waste disposal. Hence, these previous outstanding applications inspire us to introduce Schiff's base-crosslinked chitosan-glyoxal/kaolin (CTS-GLY/KN) composite as a new biocomposite material with multifunctional performances for color removal (CR) and reduction of chemical oxygen demand (COD) of an industrial anionic dye (Remazol Brilliant Blue R) from contaminated water. Statistical optimization by Box-Behnken design (BBD) is gathered to obtain maximum outputs from minimum inputs. Furthermore, the adsorption mechanism, isotherms, and kinetics of RBBR were investigated as well.

Synthesis of CTS-GLY/KN
CTS-GLY/KN composite was prepared according to the published method [18]. In this preparation, the CTS-GLY/ KN was obtained by combining 1 g CTS flakes and 1 g KN powder (50% CTS: 50% KN, this ratio was predetermined to be the best) in a beaker holding 60 mL of acetic acid solution with 5% v/v concentration while being stirred strenuously for the duration of 24 h at room temperature to dissolve the CTS flakes. CTS/KN beads were obtained by injecting droplets of CTS/KN viscous solution into a beaker containing 1 L of NaOH solution (0.5 M) by a syringe needle (10 mL). Afterwards, deionized water was utilized to wash the remaining sodium hydroxide from the CTS/KN beads. The Schiff's base formation was achieved by cross-linking reaction with 2% GLY according to the published procedure [38]. After that, the composite (CTS-GLY/KN) beads were washed using deionized water before being dried at room temperature for 24 h. Then, the CTS-GLY/KN beads were ground into fine particles and left to dry inside an oven for 24 h at 80 °C. Eventually, the CTS-GLY/KN in powder form was sieved to attain a final particle size ≤ 250 µm. The steps taken to synthesize CTS-GLY/KN are illustrated in Fig. 1.

Characterization
The surface porosity, surface morphology and surface functional groups of CTS-GLY/KN before and after RBBR uptake were analyzed utilizing Micromeritics ASAP 2060 analyzer (by measuring N 2 adsorption/desorption isotherms at 77 K), a Scanning electron microscopy-energy dispersive X-ray (SEM-EDX, TM3030Plus, Tabletop Microscope, Hitachi, Japan) and a Fourier transform infrared (FTIR) spectrophotometer at range 450-4000 cm −1 (Perkin-Elmer, Spectrum RX I), respectively. Crystalline structure of CTS-GLY/KN was studied using X-ray diffractometer (XRD, Cu Kα radiation λ = 1.54 Å, X'Pert PRO, PAnalytical) and the zero point of charge (pHpzc) of the CTS-GLY/KN was arbitrated with reference to the methodology illustrated in the literature [39]. The unreacted amino (-NH 2 ) groups present in the polymeric matrix composite were estimated by pH-potentiometric titration approach [40].

Experimental design
The effects of three adsorption parameters (CTS-GLY/ KN dose, contact time and pH of treatment environment) on the CR and COD reduction of RBBR were studied and optimized using RSM-BBD method. The Stat-Ease Design Expert (version 13.0, Minneapolis, USA) software was utilized for the purpose of designing the adsorption experiments, optimizing the processes, and finally statistically analyzing the data. The levels and coded values of the independent parameters used for designing the experiments are displayed in Table 1. The following quadratic  Eq. (1) was employed to analyze the experimental data and predict the CR and COD values: where, Y represents the predicted response (RBBR color removal (%), and COD reduction (%)); X i and X j are coded the independent variables; 0 is the constant; i , ii and ij are coefficients of linear, quadratic, and interactive coefficient of input independent variables, respectively. To investigate and optimize the effect of the three parameters (CTS-GLY/ KN dose, time, and pH) on CR value and COD reduction by CTS-GLY/KN, BBD opted 17 experimental runs consisting of three factors with three value levels and five center points: A: CTS-GLY/KN dose (0.02-0.1 g), B: pH (4-10), and C: time (5-30 min). Table 2 presents the BBD matrix and obtained response results of CR and COD reduction of RBBR. Specific amount of the adsorbent was poured into Erlenmeyer flasks (250 mL) containing 100 mL of RBBR solution then placed inside a thermostat water bath shaker. Agitation of the solutions was achieved by shaking at a constant speed of 100 strokes per minute. Then, the adsorbents were separated from the solution after adsorption process using a syringe filter (0.45 lm). UV-Vis spectrophotometer (HACH DR 3900) was utilized to measure the RBBR CR and COD concentrations at the wavelength of maximum absorption (λ max ) 592 nm and 620 nm, respectively. Equations 2 and 3 were adopted to calculate the CR % and COD % of RBBR, respectively. (1) where C o (mg/L) and C e (mg/L) are RBBR concentrations at the beginning and at equilibrium, respectively while D o (mg/L) and D e (mg/L) are concentrations of COD at the beginning and at equilibrium, respectively.

Adsorption Study of RBBR on CTS-GLY/KN
The study on the RBBR uptake by CTS-GLY/KN was done in batch mode. The obtained results are presented Table 2 which shows the highest color removal (97.3%) and COD reduction (81.8%) were achieved in the following experimental conditions: time (30 min), solution pH (4), and CTS-GLY/KN dose (0.06 g). These conditions were later applied to test the adsorption values of RBBR at differing initial concentrations (50-400 mg/L) of the dye and contact time (0-360 min) using identical batch experiment process as abovementioned. Equation 4 was used for calculating the adsorption capacity of CTS-GLY/ KN toward RBBR at equilibrium q e (mg/g).

Characterization of CTS-GLY/KN
The physicochemical characteristics of CTS-GLY/KN are presented in Table 3. The free amine groups (%) of the CTS-GLY/KN, that are considered as one of the most active adsorption sites in the CTS composites for adsorption of anionic dyes, are determined through pH-potentiometric titration. The results demonstrate a satisfactory value of 27.4% free amine groups (%) in the CTS-GLY/ KN composite, suggesting a good potential for adsorbing the RBBR molecules. It was found from the BET analysis that the surface area of CTS-GLY/KN is 39.8(m 2 /g), while mean pore diameter is 6.2 nm. This result indicates a higher surface area for CTS-GLY/KN than that of the CTS-GLY (0.39 m 2 /g) as reported in a previous study [38], which can be attributed to the fact that the KN particles have a high surface area, scattered on the CTS-GLY matrix, which enlarges its surface [41]. Similar observations have been reported in our previous studies for the (4) improvement of the surface area of the crosslinked chitosan by intercalation of the KN clay in its structure [41,42]. According to the IUPAC classification, the CTS-GLY/ KN is a mesoporous material, with a mean pore diameter (2-50 nm) greater than that of the RBBR molecules (1.6 nm) [43]. This result indicates that the pores of CTS-GLY/KN can easily accommodate the RBBR molecules. The X-Ray Diffraction pattern of the CTS-GLY/KN is illustrated in Fig. 2, exhibiting sharp characteristic diffraction peaks at 2θ = 26, 38°, 44°, 64°, and 78° corresponding to the kaolinite Al 2 Si 2 O 5 (OH) 4 [44]. These peaks reveal the high crystallinity of CTS-GLY/KN caused by the strong electronic interactions between the positive charge of crosslinked biopolymer surface (CTS-GLY) and negative charge of KN clay, evidencing intercalation of KN particles into the structure of molecules of CTS-GLY/KN [45,46]. FTIR spectral analysis was performed to identify the changes in the functional groups of the CTS-GLY/ KN before and after adsorption of RBBR as shown in Fig. 3a and b, respectively. The FTIR spectrum of CTS-GLY/KN (Fig. 3a) displays the characteristic peaks at 3400 cm −1 , 1640 cm −1 , and 1049 cm −1 , that can be related to vibrations of -OH and -NH 2 , vibration of C = N, and stretching vibration of Si-O-Si [34,38]. The bands at 2900-2860 cm −1 , 1375 cm −1 , 940 cm −1 , and 500-795 cm −1 are due to the stretching vibrations of C-H symmetrical and asymmetrical, stretching vibration of C-N, bending vibration of -OH in Al-OH, and vibrations of Si-O-Al, respectively [47]. The FTIR spectrum of the CTS-GLY/KN after RBBR uptake (Fig. 3b) illustrates an outline similar to CTS-GLY/K with a slight shifting of some distinguished bands, evidencing that the adsorption of RBBR involves various functional groups of the CTS-GLY/KN.
The surface morphology of CTS-GLY/KN and CTS-GLY/KN after RBBR uptake, in addition to analysis of its chemical composition was done utilizing SEM-EDX analysis. The surface morphology of CTS-GLY/KN (Fig. 4a) shows a rugged surface with scarce porosities and crevices in existence. The EDX analysis highlights the existence of the following elements in the polymeric structure of CTS-GLY/KN: O, C, Al, N and P. After RBBR uptake onto the CTS-GLY/KN surface, the morphology of CTS-GLY/KN (Fig. 4b) looks more compact with less condensed porosities, in accordance with the RBBR uptake onto the surface of the CTS-GLY/KN. The existence of S is displayed in the EDX analysis of CTS-GLY/KN after RBBR uptake that further confirms RBBR adsorption onto the surface of the CTS-GLY/KN.

BBD Model Analysis
BBD-RSM was utilized to study the interactive and individual impacts of the main adsorption variables on the CR and reduction of COD of RBBR. In this study, the main parameters used were CTS-GLY/KN dose, solution pH, and time. Furthermore, analysis of variance (ANOVA) was used for further investigating the statistical analysis of the experimental results collected regarding CR and COD reduction of RBBR (Table 4). The F-value for CR was found to be 43.07 (p value < 0.0001), whereas the F value for COD was found to be 28.45 (p value of 0.0001) according to Table 4, revealing that the BBD model is statistically significant for CR and COD reduction [48].
Furthermore, the values of the coefficient of determination (R 2 ) values are found to be 0.98 and 0.97 for CR and COD reduction, respectively, showing a high coherence between experimental and predicted values of CR and reduction of COD. Moreover, according to BBD, the statistical values for p-value are ought to be < 0.05 to categorize a model term in BBD as statistically significant, hence A, B, C, AB, BC, and C 2 model terms are considered statistically significant. The other model terms with p value < 0.05 were disregarded in the second order polynomial equation; that is to obtain the best fit results for the model. Therefore, the correlation between the significant parameters towards CR and reduction of COD are shown in Eq. 5 and Eq. 6, respectively.   Another method that can also be implemented to validate the results of the BBD model is the graphical method. It works on investigating the nature of residuals distribution and the fitting of actual versus predicted values of CR and reduction of COD. Figure 5a and b show the normal probability of residuals in the BBD model for CR and reduction of COD of RBBR, respectively. In the abovementioned figures, the close alignment of points into a linear form indicates the optimum normal distributions as well as the independence of residuals. Whereas Fig. 5c and d illustrate the correspondence between the predicted versus actual results of CR and COD reduction values. From the abovementioned figures, the actual and predicted values are relatively similar which can validate the BBD model statistically [49].

The Effect of Adsorption Operating Parameters
The significance of interaction amongst every set of two parameters on the CR and reduction of COD of RBBR was investigated. It was found that the effects of interaction between the CTS-GLY/KN dose (A), and pH (B) on the CR and reduction of COD of RBBR is statistically significant while the other parameter (time = 17.5 min) is kept unchanged. The three-dimensional response surface plots for the interaction between the CTS-GLY/KN dose and pH for the CR and reduction of COD are shown in Fig. 6a and b, respectively. From the figures, it can be deduced that decreasing the pH of the solution from 10 to 4 causes an increase in the CR (%) and COD reduction (%). In the same context, the pH pzc of the CTS-GLY/KN is determined to be 7.0 as presented in Fig. 6c. This result shows that negative charge can be acquired on the surface of CTS-GLY/KN at pH > pH pzc . Consequently, when solution pH reaches 4 the charge on the surface of CTS-GLY/KN can be converted to positive charge, signifying the ability of CTS-GLY/KN on adsorption of anionic RBBR dye. Thereby, an intense electrostatic attraction can take place between the surface functional groups of CTS-GLY/KN that are positively charged and the RBBR's anionic group as seen in Eq. 7: The interaction between pH (B) and time (C) on the CR and reduction of COD of RBBR is statistically significant while the other parameter (CTS-GLY/KN dose = 0.06 g) is kept unchanged. Figure 6d and e illustrate the three-dimensional response surface plots for the impact of the interaction between the two parameters of pH and time on the CR and reduction of COD, respectively. As shown in the abovementioned figures, the increase in the contact time from 5 to 30 min led to an increase in the CR (%) and reduction of COD (%) of RBBR from 12.5% to 97.3% and 5.6% to 81.8%, respectively. This observation can be assigned to the extension of time for the molecules of RBBR dye to penetrate deeper into internal pores in the polymeric structure of the CTS-GLY/KN.

Adsorption Study
In this study, the impact of the contact time on RBBR dye uptake onto the surface of CTS-GLY/KN was investigated. Different initial RBBR concentrations (50, 100, 150, 200, 300, and 400 mg/L) in the solution were used for testing and validation of the results while the CTS-GLY/ KN dose (0.06 g/100 mL), solution pH (4) and temperature (30 °C) were kept unchanged. Figure 7a illustrates the breakthrough curves of adsorption capacities of CTS-GLY/KN, q t (mg⁄g) versus time (min) at various RBBR initial concentrations. As shown in Fig. 7a, the increase in concentration of RBBR starting from 50 gradually to 400 mg/L led to an increment in adsorption capacity of CTS-GLY/KN towards RBBR molecules from 58.6 to 284.9 mg/g. This result corresponds with the greater concentration gradient which in return increments the diffusion of RBBR molecules in the internal pores of the CTS-GLY/KN leading to the RBBR molecules moving to active adsorption sites [50].

Kinetic Modeling
To investigate the RBBR adsorption behavior on the CTS-GLY/KN surface, in addition to specifying the adsorption mechanism, kinetics of the RBBR adsorption on the surface of CTS-GLY/KN were examined by applying two kinetic models: pseudo-first order (PFO) and pseudosecond order (PSO) [51,52]. The non-linear formulas of pseudo-first order and pseudo-second order kinetic models are shown by the following Eq. 8 and Eq. 9, respectively.
where, q t and q e (mg/g) represent the amount of RBBR adsorbed by CTS-GLY/KN at various time (t), and at equilibrium, respectively. k 1 (min −1 ) represents the PFO rate constant while k 2 (g/mg min) represents the PSO rate constant.

Isotherm Modeling
One of the indispensable parameters to explain the interactions between RBBR molecules and the surface of CTS-GLY/KN is the equilibrium isotherm. Three widely used isotherms (Langmuir [54], Freundlich [55], and Temkin [56]) were selected to explain the RBBR-CTS-GLY/KN interactions, affinity of CTS-GLY/KN towards RBBR molecules, and determine the uptake capacity of CTS-GLY/ KN for RBBR. The non-linear formulas of Langmuir, Freundlich, and Temkin isotherms are given in Eqs. 10-12, respectively.
(10) q e = q max K a C e 1 + K a C e (11)    Bone char 20.6 [63] While C e (mg/L) represents the concentration of RBBR in solution at equilibrium, q max (mg/g) indicates the monolayer adsorption capacity of the CTS-GLY/KN, and q e (mg/g) is the uptake value of RBBR at equilibrium concentration. K a (L/mg), K f (mg/g) (L/mg) 1/n , and K T (L/mg) are Langmuir, Freundlich, and Temkin constants, respectively. R (8.314 J/mol K) represents the gas constant, T (K) is temperature, b T (J/mol) is the heat of adsorption and n represents adsorption intensity. Figure 7b illustrates the non-linear isotherms' plots, and Table 6 presents the values of isotherm parameters. From the R 2 values of the isotherm models (Table 6), it can be derived that the adsorption of RBBR onto CTS-GLY/KN surface is more accurately explained by both Langmuir and Freundlich models, demonstrating monolayer and multilayer coverage of the RBBR dye on the surface of CTS-GLY/KN are occurred simultaneously [57]. The q max of RBBR on the CTS-GLY/KN surface is determined to be 447.1 mg/g at 30 °C. The q max of RBBR in multiple different adsorbents including CTS-GLY/KN is compared in Table 7.  Table 7, it could be derived that CTS-GLY/ KN is an effective adsorbent to remove the reactive dyes as well as to reduce COD from aquatic media.

Adsorption Mechanism of RBBR
The RBBR adsorption mechanism onto CTS-GLY/KN surface is suggested according to the existing surface functional groups on the CTS-GLY/KN surface, e.g. -NH 2 , -OH, and ≡SiOH. Figure 8 summarizes the interactions possible between RBBR and CTS-GLY/KN. The highest RBBR dye removal was observed at pH 4 where the surface of the CTS-GLY/KN is having positive charge. The cationic amino groups (NH 3 + ) of CTS and positive charges of KN result into this positive charge on the surface of the CTS-GLY/KN, resulting into electrostatic attractions occurring between the positively charged CTS-GLY/KN surface and negative sulfonate groups (-SO 3 − ) of RBBR. Hydrogen bonding occurring between the free hydrogens of the CTS-GLY/KN with nitrogen and oxygen elements available in the structure of RBBR dye is the next significant adsorption mechanism of RBBR onto the surface of CTS-GLY/KN. Moreover, Yoshida H-bonding interaction can occur between hydrogen of hydroxyl groups of CTS-GLY/KN and the aromatic rings of RBBR dye as well [59]. Finally, n-π interaction takes place between electron donating groups of oxygen and nitrogen on CTS-GLY/KN surface and π-system (electron acceptor) in the aromatic rings of RBBR [64].

Conclusions
CTS-GLY/KN composite is successfully fabricated and employed for CR and reduction of COD of RBBR from aqueous solutions. The optimum conditions for CR and reduction of COD of RBBR were found to be pH 4, dose (0.06 g), and time (30 min). The results found that the highest values of CR (97.3%) and COD reduction (81.8%) of RBBR can be achieved by the following significant interactions: AB and BC. The adsorption capacity obtained from Langmuir model reflected a high adsorption capacity (447.1 mg/g) of CTS-GLY/KN composite. The adsorption mechanism of RBBR dye by CTS-GLY/KN is attributed to electrostatic, n-π, Hydrogen bonding, and Yoshida H-bonding interactions. The results also reveal that CTS-GLY/KN could be utilized as an adequate adsorbent for color removal and reduction of COD of textile dyes from aqueous environment.