Novel Cellulose/Silica Microspheres as Sustainable and Efficient Adsorbents for Methyl Orange Removal

The effective removal of organic dyes from aqueous solutions is essential for environmental protection and remediation. Hence, methods for fabricating efficient and sustainable adsorbents for removing toxic dyes are urgently desired. Here, new cellulose/silica microspheres containing amino groups were prepared and applied as functional materials for capturing anionic pollutants. Tosyl cellulose, which has a high degree of substitution, was prepared from cellulose using an ionic liquid as the solvent. Further, (3-aminopropyl)trimethoxysilane was utilized as a silica precursor to prepare the microhybrids via nucleophilic substitution, followed by the sol–gel process. The fabricated microhybrids exhibited an excellent capacity for methyl orange (MO) adsorption. Furthermore, the effects of different pH values, contact times, and initial dye concentrations on the MO adsorption capacity were evaluated. The Langmuir isotherm and pseudo-second-order kinetic models were effective for modeling the adsorption of MO on the cellulose/silica microspheres. Under the established optimal conditions, the cellulose/silica microspheres exhibited a higher adsorption capacity (588 mg/g) than those in previous reports. Therefore, the proposed cellulose/silica microspheres offer a promising platform for the sustainable development of water-purification systems.


Introduction
In industrial processes, such as textile, leather, and paint manufacturing, synthetic dyes are often generated and released in large amounts, causing various environmental problems, particularly in aquatic biosystems [1][2][3]. Many synthetic dyes used in industrial processes are toxic, mutagenic, and occasionally carcinogenic. Most of these dyes, 1 3 unfortunately, are discharged into wastewater, causing serious health issues to humans and animals [4,5]. Approximately, 0.2 million tons of dye effluents are generated annually by the textile industry. Synthetic dyes contain aromatic structures stabilized by resonance, which account for their low biodegradation and consequent adverse environmental effects. To date, new biomaterials are required for removing dyes from wastewater. Recently, eco-friendly and cost effective techniques for removing toxic dyes have been receiving extensive research interest [6,7]. Among the various methods for removing organic and inorganic contaminants from wastewater, the use of bioadsorbents has recently been established as an interesting approach [8][9][10]. In particular, natural polymers exhibit numerous advantages for bioadsorption, including low cost, nontoxicity, high biodegradability, and high eco-compatibility [5,11,12]. As bioadsorbents, polysaccharides exhibit low adsorption capacities, owing to their relatively low specific surface area, solubility, and functionality. Recently, extensive efforts have been undertaken to improve the sorption properties of natural materials via the development of organic-inorganic composites [13,14].
Recent studies have demonstrated the effectiveness of polysaccharides for preparing cost effective, environmentally friendly adsorbents that are readily available in tonnage quantities [15]. Moreover, efficient adsorbents have been produced from polysaccharides containing charged surfaces, such as carboxymethyl cellulose (CMC) and chitosan [16]. This is made possible by their electrostatic interactions with adsorbates. Furthermore, researchers have employed many techniques to increase the adsorption capacities of polysaccharides, including crosslinking (chemical or physical), composite fabrication, blending, grafting, and nanofabrication. For example, N-guanidinium chitosan acetate was synthesized by the guanylation of chitosan using cyanamide in the presence of scandium(ӀӀӀ) triflate. Second, N-guanidinium chitosan/silica microhybrids were prepared using 3-glycidoxypropyl trimethoxysilane via the sol-gel process. The results demonstrated the excellent adsorption properties of the N-guanidinium chitosan/silica microhybrids, with capacities as high as 917 mg/g [17]. A CMC/Fe 3 O 4 nanocomposite was fabricated using the CMC prepared from cellulose extracted from mesquite trees. The results of the tests conducted confirmed the formation of homogenous and spherical magnetic nanoparticles (diameter: ∼25 nm). Finally, the nanocomposite was employed for the removal of methylene blue, and the maximum adsorbed amount was estimated to be 1597 mg/g [18].
Cellulose, the most abundant biopolymer from forest or agricultural biomass, is a linear polysaccharide comprising D-glucopyranose rings joined by a β-1,4-glycosidic bond [19,20]. Generally, membranes and filters are produced from cellulose because of its intrinsic properties. As a result, they are suitable for use in industrial and drinking-water-treatment systems. In addition, the high stiffness and water insolubility of cellulose make it a good material for high-pressure applications. Furthermore, neat cellulose exhibits a low adsorption capacity because of the small surface area and large pore size of cellulose fiber webs. Thus, for treating complex effluents and to meet the increasing demand for clean water by local communities, the production of functional materials from cellulose remains essential.
This paper describes the preparation of a novel, functionalized cellulose/silica microhybrid for the sequestration of anionic methyl orange (MO). In the first step, cellulose was tosylated using an ionic liquid (IL) in the presence of pyridine. Next, the obtained tosyl cellulose was used to develop the cellulose/silica microhybrids by grafting (3-aminopropyl) trimethoxysilane. The hybrid was constructed with uniform microspheres and applied for the efficient sequestration of anionic MO. Finally, the effects of pH, contact time, and the initial dye concentration on the MO-adsorption capacity were systematically studied.

Preparation of Tosyl Cellulose
Tosyl cellulose was prepared from cellulose using ILs as the solvent [21]. Briefly, 2.0 g of cellulose was dissolved in 18 g of [BMIM] Cl at 80 °C for 16 h. Thereafter, the solution temperature was decreased to 25 °C, after which 10 ml of pyridine was added under continuous stirring. Subsequently, 7.06 g of p-toluenesulfonyl chloride was dissolved in 10 ml of pyridine and added under regular stirring at room temperature. After 12 h, the reaction mixture was poured into a sufficient amount of ethanol to form a white precipitate. Afterward, the precipitate was separated, washed with ethanol, and dried under vacuum at 60 °C (yield: 3.2 g). Elemental analysis afforded the following results for cellulose (%) and tosyl cellulose (%), respectively: C 42.9, H 6.4, N 0.0, S 0.0 and C 45.3, H 5.2, N 0.03, S 10.2.

Preparation of the Cellulose/Silica Hybrid
The cellulose/silica hybrid containing amino groups was synthesized in two steps. First, 0.5 g of tosyl cellulose was dissolved in 5 mL of dimethyl sulfoxide (DMSO) and stirred for 3 h at 100 °C with 2 ml of (3-aminopropyl)trimethoxysilane. Next, a few drops of 0.01 M HCl were added to the solution with continuous stirring for 24 h. The resulting gel was carefully washed with distilled water and methanol, after which it was lyophilized for 24 h. 13 C-NMR spectroscopy was performed on a Bruker Avance 400 spectrometer at room temperature. Deuterated DMSO(DMSO-d6) was utilized as the solvent for the liquidstate NMR experiments.

Characterization Methods
The attenuated total reflectance Fourier-transform infrared (ATR-FTIR) spectra of the cellulose, tosyl cellulose, and cellulose/silica hybrid were recorded on a Thermo Nicolet Avatar 320 FT-IR spectrometer with a diamond crystal. Thermogravimetric analysis (TGA) was conducted on a NETZSCH STA 409 PC instrument. All the materials were burned in an oxygen atmosphere between 25 °C and 900 °C, at a heating rate of 5 °C/min. The surface morphology of the cellulose/silica microhybrid was examined using a Hitachi S-4800 scanning electron microscope.

Batch-Adsorption Study
A stock solution of the MO dye (1000 mg/l) was prepared using distilled water and diluted to the required experimental concentrations. The MO adsorption properties of the cellulose/silica microhybrid were evaluated under different operating conditions. The effects of pH, contact time, and the initial concentration were also considered. The adsorption studies were performed at pH 3-8, and the pH was adjusted using HCl/NaOH (0.1 M) by mixing 25 mg of the microhybrid in the MO solution (100 mg/l) using a shaker operated for 120 min. Kinetic studies were performed at pH 4, using 25 mg of the microhybrid and 100 mg/l of the MO solution, with a contact-time range of 10-160 min. The effects of the initial concentration and adsorption isotherms of MO were evaluated using initial concentrations ranging from 50 to 1200 mg/l, 25 mg of the microhybrid, and an equilibrium time of 120 min at pH 4. Afterward, a spectrophotometer (JASCO V-650) was employed to measure the ultraviolet-visible absorption spectrum of the dye solution.
To calculate the dye concentration in the solution, the optical density at λ max = 464 nm and the calibration curve were employed. The amount of adsorbed dye at a given time was calculated using Eq. (1): where C 0 and C e represent the initial and equilibrium dye concentrations, respectively (mg/l); V is the volume of the dye solution used in the adsorption experiment (l); W is the weight of the microhybrid (g).

Structural Study
The formation of tosyl cellulose by the reaction of neat cellulose with tosyl chloride has been reported. However, this reaction requires specific conditions to dissolve the cellulose and enhance the degree of substitution (DS). Gericke et al. prepared tosyl cellulose using an IL as the solvent in the presence of cosolvents. The reaction was performed at 25 °C to produce tosyl cellulose with a DS of ≤ 1.14 [21]. Here, this protocol was applied to produce tosyl cellulose, which was further subjected to nucleophilic substitution employing a silica precursor containing amine groups. Scheme 1 illustrates the synthesis of the cellulose/silica microsphere containing amine groups (C) from tosyl cellulose (A). The nucleophilic substitution of the tosyl group with amine is illustrated in Scheme 1(B).

Characterization of Tosyl Cellulose
Tosyl cellulose was subjected to elemental analysis to estimate the degree of tosyl substitution (DS) based on the following equation: where M AGU is the molar mass of the anhydroglucose unit (AGU) = 162.05 g/mol, M S is the molar mass of sulfur = 32.065 g/mol, M TS is molar mass of tosyl group = 155.02 g/mol, and Ws is the weight (%) of sulfur determined by elemental analysis = 0.8788. Thus, the calculated DS for tosyl cellulose was 1.02. Figure 1 shows the 13 C-NMR spectra of tosyl cellulose. The anhydroglucose unit signals appeared between 80.7 and 73.5 ppm, representing anhydroglucose carbons C4, C3, C2, and C5. Moreover, the signals at 103.3 ppm (C1) and 60.7 ppm (nontosylated C6) were also assigned to the anhydroglucose unit. However, the aromatic carbon signals at 145.6, 132.5, 130.7, and 128.1 ppm confirmed the tosylation of cellulose. In addition, the peak at 68.6 ppm (C6S tosylated) and the signal at 21.1 ppm (C11) further confirmed the formation of the tosyl group.   Fig. 2A, the cellulose spectrum exhibited a broad band centered at 3390 cm −1 , associated with the stretching of the O-H bonds, and a band at 2898 cm −1 , assigned to the C-H stretching vibration [22]. The bands in the 1160-1055 cm −1 region were attributed to the C-O stretching vibrations. The cellulose spectrum displayed an O-H bending band at 1635 cm −1 , which was assigned to the absorbed water. Moreover, the sharp bands at 898 cm −1 were attributed to the C-H rock vibrations, and the intense bands in the 1165-890 cm −1 region were assigned to the motion of the C-O-C and C-O linkages [6,23]. Through the tosylation process, the symmetric and asymmetric S-O stretching vibrations of the SO 2 group were detected [24,25]. In Fig. 2B, the characteristic bands of the tosyl groups are at 820 (ν S-O-C), 1176 (ν s SO 2 ), 1363 (ν as SO 2 ), and 1595 (ν C = C aromatic ) cm −1 . The presence of tosyl cellulose was further confirmed by the significant decrease in the absorption-band intensity at 3450 cm −1 . Moreover, the cellulose/silica hybrid spectrum (Fig. 2C) displayed an additional strong absorption peak between 1124 and 1030 cm −1 , assigned to the siloxane (Si-O-Si) bonds [7]. In addition, the cellulose/silica hybrid exhibited a distinct broad band at ~ 3400 cm −1 , indicating the presence of the silanol groups generated by incomplete condensation. Since the silanol groups can improve intramolecular interactions via hydrogen bonding, they may improve the adsorption properties of the material.

Scanning Electron Microscopy (SEM) and Energy-Dispersive X-Ray (EDX) Spectroscopy
SEM combined with EDX spectroscopy enables the mapping of the elements that shape the microspheres (Fig. 3). The SEM images of the cellulose/silica hybrid produced from the reaction between tosyl cellulose and (3-aminopropyl)trimethoxysilane are also shown. The morphology of the material indicates that it has a considerably rough surface under low magnification. However, under high magnification, the SEM images indicate that the hybrid has a considerably fine substructure and contains particles with narrowsize distribution and homogenous shapes. The surfaces of these particles are smooth and without holes, reflecting the homogeneity of the organic cellulose and inorganic silica. The EDX pattern of the derived cellulose/silica microspheres is shown in Fig. 3C. The hybrid consists essentially of silicon, carbon, oxygen, nitrogen, sulfur, and chlorine. The Si peak in the EDX spectrum indicates that the hybrid mainly contains the silica-gel network. The homogenous distribution of silicon and other elements reflects the properties of the actual organic/inorganic hybrid.

TGA
The TG thermograms of the cellulose, tosyl cellulose, and cellulose/silica hybrid are shown in Fig. 4. The weight losses occurred at three main temperature regions, which were designated as I, II, and III. The weight losses ranging from 2.8 to 8.1% from room temperature to 120 °C (Region I) are likely related to the elimination of the physically adsorbed water. Compared with cellulose and tosyl cellulose, the cellulose/silica hybrid exhibited a higher amount of physically adsorbed water. In Region II (120 °C-320 °C), the hydrocarbon backbone of cellulose suffered a dramatic weight loss due to thermal degradation. Moreover, the cellulose/silica hybrid started degrading at a lower temperature (~ 120 °C) compared with the degradation temperatures of tosyl cellulose (~ 175 °C) and cellulose (~ 260 °C). After 320 °C, the weight losses in Region II were 61.5%, 53.6%, and 26% for cellulose, tosyl cellulose, and the cellulose/silica hybrid, respectively. In Region III, complete mass losses were observed at 500 and 600 °C for neat cellulose and tosyl cellulose, respectively. In addition, the cellulose/silica composite exhibited higher thermal stability than the neat cellulose and tosyl cellulose. The recorded residual mass of cellulose/silica was 26.8%, indicating the high silica content of the designed cellulose/silica hybrid.

Potential Applications of the Cellulose/Silica Microspheres for MO Adsorption
In this study, the removal of MO from the water was carried out with the as-synthesized cellulose/silica containing amine microspheres. Using adsorbents, toxic dyes can be removed from aqueous solutions. However, this depends on several factors, including the adsorbent mass, pH of the solution, equilibrium time for the adsorption, and initial concentration of the dye.

The Effect of Adsorbent
Dose was investigated in the range of 0.75-2.75 g/l (0.015-0.055 g of dried hybrid). The pH of the dye solution was adjusted at 4, and the initial concentration of the dye was 100 mg/l. It was estimated from Fig. 5A that 1.25 g/l of adsorbent exhibited high adsorption capacity. Moreover, the adsorption capacity decreased as the amount of adsorbent increased because more active adsorptive sites remained unsaturated. Further experiments were carried out using a fixed dose of 1.2 g/l of adsorbent (0.025 g of dried hybrid).

Effect of pH
The adsorption process was affected by the pH of the MO solution. It influenced the surface charge of the adsorbent, dissociation of the functional groups on the active sites, and chemistry of the solution. The pH of the dilution water for the dye was varied from 4 to 8 using 0.1 M NaOH and 0.1 M HCl solutions. Figure 5B shows that the ideal pH for MO adsorption is ~ 4. At this pH, all the amino groups on the adsorbent are fully protonated, inducing strong electrostatic interactions between the anionic MO molecules and the cellulose/silica microspheres. The adsorption process is as follows: MO is dispersed in the aqueous solution, after which the sulfonate group in the MO dye (R-SO 3 Na) dissociates to form anionic dye ions. The adsorption process is induced by the electrostatic attraction between the MO anion and the adsorbent surface. With an increase in the pH of the MO solution, the adsorption capacity decreases. In a basic solution, the dye-uptake capacity decreases because of the deprotonation of the functional groups in the cellulose/silica microspheres, which decreases the interaction between the cellulose/silica microspheres and MO. The abundant OH ions compete with the MO anions for adsorption sites, which may explain the decrease in the MO adsorption capacity. Furthermore, MO exists in acidic media as an ammonium sulfonate zwitterion, and the cellulose/silica microspheres can interact effectively with neutral species via hydrogen bonds. Thus, increasing the pH values (> 7) significantly decreases the adsorption capacity toward MO. In this study, the cellulose/silica hybrid exhibited a relatively stable MO removal efficiency over a wide pH range. Studies have shown that the optimum pH for amine-containing adsorbents, such as chitosan, is between 3 and 6 [26,27].

Effect of the Contact Time
The dye uptake depends on the contact time of the solid adsorbent and the liquid dye. The cellulose/silica microspheres exhibited high adsorption selectivity and capacity because of their surface structure. Initially, the adsorptioncapacity curve displayed a smooth, continuous pattern, before ultimately exhibiting a monotonous pattern (Fig. 6). The required equilibrium time was ~ 60 min. During adsorption, solutes are transferred from the liquid phase to the surface of the solid phase. For a better understanding of the kinetics, pseudo-first-order (Eq. (2)) and pseudo-second-order (Eq. (3)) models were adopted: where q t and q e are the amounts (mg/g) of the adsorbed dye at a given time t (min) and equilibrium, respectively; K 1 and K 2 are the equilibrium rate constants of both models, respectively. The pseudo-second-order kinetic model could describe the adsorption of MO onto the cellulose/silica hybrids, and its fitting coefficient (R 2 = 0.99) was higher than that of the pseudo-first-order kinetic model (R 2 = 0.83). Thus, the pseudo-second-order model can be considered to represent the adsorption kinetics and to be dependent on both the amount of solute adsorbed on the adsorbent surface and the amount adsorbed at equilibrium. In addition, the q e (100 mg/g) calculated using the pseudo-second-order kinetic model was close to the experimental q e value (94 mg/g), further confirming the suitability of the pseudo-second-order kinetic model. Table 1 lists the calculated results for the firstand second-order kinetic models.

Effect of the Initial Concentration of the Dye
MO solutions with different initial concentrations (50-1000 mg/l) were employed to evaluate the adsorption capacity of the cellulose/silica hybrid. Evidently, the dyeremoval efficiency increased from 45 to 480 mg/g with an increase in the dye concentration from 50 to 600 ppm and, thereafter, stabilized at higher concentrations (Fig. 7). At a low initial concentration of MO, the adsorption sites were unsaturated, resulting in the adsorption of a low amount of dye per unit mass of the adsorbent. The MO molecules penetrated the inner surfaces of the adsorption sites through diffusion at relatively high initial concentrations. Thus, by increasing the MO concentration, the driving force for the MO transfer from the aqueous to the solid phase increased, enhancing the MO microhybrid interaction. Moreover, the adsorption efficiency stabilizes as the active adsorption sites become saturated after 600 ppm.
The adsorption isotherm describes the adsorption mechanism of the dye molecules on the sorbent surface. Therefore, correlating the equilibrium data with either a theoretical or an empirical equation is essential for interpreting and predicting the adsorption capacity [28]. In this study, two isotherm equations, namely the Langmuir and Freundlich equations, were tested with the experimentally obtained equilibrium data. The Langmuir adsorption isotherm was employed to evaluate the relationship between the adsorbed amounts of MO onto the cellulose/silica microhybrid. The Langmuir isotherms assume that the surface is homogeneous, the sites are identical, and the adsorption occurs in the monolayers. In addition, it assumes that no interaction occurs between adjacent adsorbate molecules. This isotherm equation can be written in a linearized form, as follows [29,30]: where q e is the amount of MO adsorbed on the cellulose/silica hybrid at equilibrium (mg/g), C e is the concentration of MO at equilibrium (mg/l), q max is the maximum amount of MO adsorbed on the hybrid (mg/g), and K s is Langmuir's constant (mg/l). The plot of the experimental C e ∕q e against C e for the experimental data displays the high correlation coefficient (R 2 > 0.98) of the linearized Langmuir equation, indicating that the Langmuir model can describe the adsorption of MO onto the cellulose/silica hybrid (see Table 2). The values of q max and K s , as estimated from the slope and intercept of the straight line, are 588 mg/g and 54.7 mg/l, respectively.
The Freundlich equation is given by where P is a constant representing the adsorption capacity (mg/g), and n represents the adsorption intensity (dimensionless). The Freundlich isotherm model describes the adsorption on heterogeneous surfaces with molecules interacting with each other. According to the results, the linear coefficient is 0.91. These results suggest that this model does not sufficiently explain the adsorption processes of MO by the cellulose/silica hybrid. The values of the constants, P and n , of the Freundlich model are 23 and 1.9, respectively ( Table 2).
The feasibility of the adsorption process was calculated employing the separation factor ( R L ), defined by the following equation [31]: where C 0 is the initial concentration of MO (ppm). Based on this factor, an adsorption system can be predicted to be favorable or unfavorable. The calculated R L values for the adsorption of MO at different initial concentrations were in the 0-1 range. Thus, the adsorption of MO on cellulose/ silica containing amine groups has been established to be favorable in this study.

Adsorption Mechanism of the Cellulose/Silica Microspheres
MO was adsorbed by the cellulose/silica microspheres via a monolayer chemical process, which was favorable for adsorption. The protonation of the amine groups generated positive charges on the cellulose/silica microsphere surfaces. Contrarily, MO exhibited a negative charge in the aqueous solution owing to the presence of SO 3 − . Thus, electrostatic attraction between the cellulose/silica microspheres and MO was evident throughout the adsorption process. This type of harmonious adsorption improved the adsorption capacity for MO. For MO adsorption, electrostatic attraction was the dominant interaction force [32]. Further, the surface of the cellulose/silica microspheres contains abundant OH groups that can combine with the N atom on MO to facilitate dipole-dipole H-bonding interactions. It may form a chemical bond with MO called Yoshida H-bonding, which facilitates adsorption. It is also necessary to consider the possibility of n-π stacking interactions. The n-π stacking interaction frequently occurs between the lone pair of electrons on an oxygen atom and the π orbital of the aromatic rings of dyes [33]. This indicates that we may also observe n-π stacking interactions. Scheme1 shows a schematic diagram of the interaction between the cellulose/silica microspheres and MO. Notably, several factors account for the high adsorption speed of the microadsorbent, including its hydrophilic nature. However, the polymeric microparticles are entangled, which promotes the diffusion process since the hydrophilic amine groups swell quickly in aqueous environments, facilitating dye diffusion. Consequently, intraparticle diffusion could also occur during the adsorption process of anionic dye and could control the adsorption mechanism [34]. In addition, the Langmuir isotherm suggests homogeneity in the cellulose/silica hybrid microspheres with equivalent binding sites and monolayer coverage for MO.   Table 3 presents the comparison between the performances of the cellulose/silica microspheres in this study and other polysaccharide-based materials for the adsorption of different organic dyes. The cellulose/silica microspheres exhibited a high adsorption capacity compared with the other organic-based adsorbents.

Conclusion
Tosyl cellulose was reacted with (3-aminopropyl)trimethoxysilane via nucleophilic substitution, followed by a sol-gel reaction to form cellulose/silica microspheres. These microspheres comprised regular particles with narrow-size distribution and a high degree of homogeneity between cellulose and inorganic silica. Furthermore, the microsphere exhibited higher thermal stability than neat cellulose and recorded a residual mass of 26.8%, demonstrating its high silica content. The maximum adsorption capacity of the microspheres for MO was determined to be 588 mg/g. The adsorption process followed the pseudo-second-order kinetics and Langmuir isotherm. The mechanism of the adsorption could be explained via hydrogen bonding and electrostatic interactions. The synthetic simplicity and high adsorption capacity highlight the significant application potential of cellulose/ silica microspheres as environmentally friendly and economical bioadsorbents.