Low-Temperature Synthesis of Magnetic Carbonaceous Materials Coated with Nanosilica for Rapid Adsorption of Methylene Blue.

This work reports the synthesis of nanosilica-coated magnetic carbonaceous adsorbents (MCA@SiO2) using low-temperature hydrothermal carbonization technique (HCT) and the feasibility to utilize it for methylene blue (MB) adsorption. Initially, a carbon precursor (CP) was synthesized from corn starch under saline conditions at 453 K via HCT followed by the magnetization of CP again via HCT at 453 K. Subsequently, MCA was coated with silica nanoparticles. MCA and MCA@SiO2 were characterized using X-ray diffraction, Fourier transform infrared, scanning electron microscopy/energy-dispersive spectroscopy, transmission electron microscopy, and Brunauer-Emmett-Teller (BET) N2 adsorption-desorption isotherms. The BET surface area of MCA and MCA@SiO2 were found to be 118 and 276 m2 g-1, respectively. Adsorption of MB onto MCA@SiO2 was performed using batch adsorption studies and in the optimum condition, MCA@SiO2 showed 99% adsorption efficiency with 0.5 g L-1 of MCA@SiO2 at pH 7. Adsorption isotherm studies predicted that MB adsorption onto MCA@SiO2 was homogeneous monolayer adsorption, which was best described using a Langmuir model with the maximum adsorption capacity of 516.9 mg g-1 at 25 °C. During adsorption kinetics, a rapid dye removal was observed which followed pseudo-first- as well as pseudo-second-order models, which suggested that MB dye molecules were adsorbed onto MCA@SiO2 via both ion exchange as well as the chemisorption process. The endothermic and spontaneous nature of the adsorption of MB onto MCA@SiO2 was established by thermodynamics studies. Mechanism of dye diffusion was collectively governed by intraparticle diffusion and film diffusion processes. Furthermore, MB was also selectively adsorbed from its mixture with an anionic dye, that is, methyl orange. Column adsorption studies showed that approximately 500 mL of MB having 50 mg L-1 concentration can be treated with 0.5 g L-1 of MCA@SiO2. Furthermore, MCA@SiO2 was repeatedly used for 20 cycles of adsorption-desorption of MB. Therefore, MCA@SiO2 can be effectively utilized in cationic dye-contaminated wastewater remediation applications.

Introduction

Textile industry being the main consumer of synthetic dyes, among several other industries like leather, paper, food, rubber, and pharmaceuticals, discharges approximately 1000 tons of dyes annually into different water bodies.[1,2] Dye waste from these industries not only increases the content of suspended solids, biological and chemical oxygen demands, and toxicity of wastewater but also makes water aesthetically unpleasant.[3] This also interferes with the penetration of light, hence affecting aquatic photosynthesis.[4] Methylene blue (MB) dye is a cationic thiazine dye having uses in different industries, especially in the textile and dyeing industries.[5] MB dye is also used in many pharmaceutical applications.[6] The chemical structure of MB is shown under Figure S1 in the Supporting Information. Short-term exposure to MB is known to cause eye and skin irritation, whereas prolonged exposure can result in nausea, breathing difficulty, increased heart rate, gastritis, and mental confusion.[5,7] Thus, the complete removal of MB from the industrial waste prior to their discharge is very important.

Various techniques have been employed to remove dye contaminants from industrial wastewater. Among them, the most popularly discussed methods include degradation either photocatalytic or sonochemical, adsorption, membrane filtration, biosorption, Fenton-biological treatment, oxidation, and coagulation–flocculation.[8−10] The most advantageous method is found to be adsorption, which is an equilibrium separation technique. Its low cost, ease and flexibility of design, reduced sensitivity to toxic contaminants makes adsorption the best technique for dye removal as compared to other techniques.[11−13] Adsorption using economical adsorbents for efficient decolorization has been studied in recent times.[14−16] The high cost of the most commonly used adsorbent, that is, activated carbon,[17] has led to the use of several cheap and nonconventional adsorbents like sawdust, sugar cane bagasse, rice husk, fruit shells, and different gum polysaccharide-based hydrogels.[1,18−21] However, their low mechanical strength does not allow them to be used on an industrial scale. In recent years, substantial progress has been made for using different nanomaterials especially metal oxide nanoparticles such as ZnO, TiO2, Fe3O4, and SiO2 in different wastewater treatment applications.[4,22−24] Therefore, researchers are focusing on developing nanoadsorbents because of high porosity and increased surface area for adsorption.[25] Nanosilica-based adsorbents have attained special attention because of low cost and easy synthesis process.[26,27] Separation of these nanoadsorbents after wastewater treatment is very difficult but the addition of nanomagnets to these adsorbents is found to have a positive impact on their separation from solution after adsorption. Nanoadsorbents with a magnetic medium can be easily removed from solutions using an external magnet.[28] However, the high cost of the synthesis of magnetic nanoadsorbents can be overcome by using carbon precursors prepared from low-cost biopolymers.[29]

Biopolymers can be converted to carbonaceous materials with water as the reaction medium via an ecofriendly thermal conversion process known as hydrothermal carbonization technique (HCT) under mild conditions and a temperature up to or below 523 K.[30] In HCT reactions, wet biopolymers can be processed, which reduces the extra cost of drying wet biopolymers; also, it has almost the same conversion efficiency as pyrolysis at comparatively higher temperatures.[31] Low-temperature HCT is advantageous as it involves one-step synthesis with mild and environment-friendly conditions.[32] However, in general, biopolymers experience difficulty to undergo conversion to carbonaceous materials at low temperatures except for particular biopolymers like corn starch and corncob even at temperatures of approximately 453 K under saline conditions (common salts such as ZnCl2) as the hydrophilic ions present in these salts reduce the partial pressure of water.[22,29,33]

Therefore, keeping in view the advantages of the low-temperature HCT method for the preparation of carbonaceous materials at comparatively lower temperatures than other existing techniques and the benefits of using various nanomaterials in water purification applications, in this research work, we initially synthesized carbonaceous material; that is, the carbon precursor (CP), from low-cost starch using low-temperature HCT at 453 K under saline conditions CP was further magnetized to produce an MCA (magnetic carbonaceous adsorbent) again via low-temperature HCT. Subsequently, MCA was successfully coated with nanosilica by treatment with tetraethyl orthosilicate (TEOS) and ethanol. The adsorption efficiency of MCA@SiO2 for effectively decolorizing MB in aqueous solution was tested and also its chemical, physical, and structural characteristics were investigated. Different adsorption parameters were optimized, and adsorption kinetics and isotherm experiments conducted to best understand the adsorption mechanism of MCA@SiO2. Additionally, its applicability for industrial applications was examined for 20 successive cycles of adsorption–desorption. The dynamic and selective adsorption of cationic dye, that is, MB, was also tested using column studies and from its mixture with methyl orange (MO) (an anionic dye). The innovative approach of this work is the use of low-temperature HCT technique for the preparation of magnetic carbonaceous material and further their coating with nanosilica and subsequently their use in water purification applications as an adsorbent.

Experimental Section

Materials and Methods

FeCl3·6H2O (reagent grade, purity ≥97%), ZnCl2 (ACS reagent, purity ≥97%), corn starch, MB (dye content, ≥82%), MO (ACS reagent, dye content 85%), acetone (HPLC grade, purity ≥99.9%), TEOS (GC grade, purity ≥99%), ammonium hydroxide (ACS reagent), ethanol (purity ≥97%) were procured from Sigma-Aldrich, UAE.

Adsorbent Synthesis

CP and MCA Synthesis

CP as well as MCA were synthesized using corn starch as the starting material via low-temperature HCT. The detailed procedure for the synthesis of CP from corn starch and its magnetization using FeCl3·6H2O is reported in our previous publication.[22] Briefly, 6.0 g of corn starch and 9.0 g of ZnCl2 were mixed thoroughly with 30 mL of deionized water in a Teflon container using a magnetic stirrer. The Teflon container was kept in an autoclave, tightly closed, and heated at a temperature of 453 K for 6 h. The solid residue was collected by vacuum filtration and excess ZnCl2 was removed by excessive washings with deionized water. In the next step, CP (4 g) synthesized in the first step was mixed and stirred with FeCl3·6H2O (2 g) in 50 mL of deionized water for 30 min. During stirring, 25 mL of NaOH (2.5 N) was added dropwise. Thereafter, the reaction mixture was heated in a tightly sealed Teflon container at 453 K for 6 h. Finally, the prepared MCA was dried in a vacuum oven after repeatedly washing with deionized water.

Synthesis of MCA@SiO2

In the second step, silica nanoparticles were coated on MCA via the Stöber method using TEOS as the silica source and ammonia as a catalyst in homogeneous aqueous-alcoholic medium. Initially, MCA particles (0.3 g) were dispersed in an ethanol (42 mL) and water (4.5 mL) mixture with constant stirring for 10 min followed by the addition of a silica source, that is, TEOS (2.5 mL), to the suspension. Subsequently, aqueous ammonia solution (2 M, 7.5 mL) was added dropwise under constant stirring to hydrolyze the silica precursor. Hydrolysis reaction was continued at room temperature for 24 h under stirring condition. Then, MCA@SiO2 was collected via centrifugation and given successive washings with ethanol/water mixture. Finally, MCA@SiO2 was dried in an oven at 80 °C for 24 h. The detailed process for MCA@SiO2 synthesis is presented in Scheme .

Scheme 1

Detailed Procedure for MCA@SiO2 Synthesis by Taking Corn Starch as the Starting Material

Characterization

The KBr pellet method was utilized to record the Fourier transform infrared (FTIR) spectra using a Bruker Vertex 70 FTIR spectrophotometer. An X’PERT Powder PANalytical powder diffractometer (The Netherlands) was used to obtain the (XRD) X-ray diffraction patterns. Scanning electron microscopy (SEM) analysis using a Quanta FEG 250 scanning electron microscope, FEI (USA), was done to capture SEM images of the carbon-coated dried samples for examining the surface morphology. Transmission electron microscopy (TEM) micrographs were obtained using FEI, Tecnai T20, The Netherlands , transmission electron microscopy with energy-dispersive spectroscopy (TEM–EDX). Surface properties of MCA@SiO2 were evaluated via N2 adsorption–desorption using Micromeritics ASAP 2020. For calculating pzc, that is, point of zero charge, 0.05 g of MCA@SiO2 was added in aqueous solution having pH in the range of 2–11 and agitated till constant pH was obtained. A graph between the starting pH and the difference between the starting and final pH was plotted to calculate the pzc.[4] The concentration of free hydroxyl (−OH) and carboxylic acid (−COOH) groups in MCA and after the coating with silica nanoparticles were determined through Boehm titration.[34] For these experiments, a fixed amount of both the materials, that is, MCA and MCA@SiO2, were suspended in 0.1 N solutions of NaOH and NaHCO3 in separate containers. Thereafter, the supernatant of both the solutions (5 mL) were back-titrated using 0.1 N HCl and the concentration of free −OH and −COOH groups present in MCA and MCA@SiO2 were calculated based on residual bases after back-titration.

MB Adsorption

To study the adsorption performance of MCA@SiO2 for MB dye removal, adsorption experiments were conducted using the batch equilibrium method. MB dye solutions (50 mL, 100 mg.L–1, Co) and varied quantities of adsorbent were agitated at 400 rpm in glass bottles for 24 h. Subsequently, the solutions were filtered and the residual dye concentration (Ce) was analyzed at λmax of MB (668 nm) using UV–vis spectrophotometry (Varian Cary 5000, USA). Adsorption efficiency was calculated using eq

Isotherm experiments were conducted with dye solutions of 50–500 mg L–1 concentration, whereas kinetics experiments were performed with 50, 100, and 150 mg L–1 concentration solutions of MB. Equilibrium adsorption capacity (qe) was calculated aswhere V and m denote the solution volume and adsorbent mass, respectively.

The dynamic adsorption behavior of MCA@SiO2 was also studied to examine its ability for the treatment of industrial dye waste stuff and to act as a column filler. For these experiments, 0.5 g L–1 MCA@SiO2 was filled in a syringe to form a filter bed and MB dye solution (50 mg L–1) was passed through it, effluent was collected, and analyzed using a UV–vis spectrophotometer.

Desorption of already adsorbed dye molecules and regeneration of adsorption sites of the adsorbent for adsorbing dye molecules again were performed successfully for 20 consecutive sets of adsorption–desorption. Acetone was used for desorption of MB dye molecules as well as for regenerating the adsorption sites of MCA@SiO2. Before using the adsorbent particles for the next cycle of adsorption–desorption, adsorbent particles were dried at 60 °C after washing with deionized water.

Results and Discussion

To understand the successful synthesis of MCA via HCT and coating of nanosilica on its surface, the FT-IR of MCA, SiO2, and MCA@SiO2 were studied as shown in Figure a. In the case of MCA, the FT-IR spectrum depicted peaks of the C–O bond (1247 and 1056 cm–1), C=C and C–H stretching related to the aromatic group (1576 and 2943 cm–1, respectively), −OH group stretching (3357 cm–1), and the peak representing the presence of magnetic Fe3O4 particles deposited on the sample surface (590 cm–1).[22,35] However, the FT-IR spectrum of SiO2 showed peaks of the surface −OH group at 1630 and 2800–3700 cm–1, respectively, whereas the silicon dioxide structure was shown by the presence of characteristic peaks at 952 (Si–OH stretching), 792 (Si–O bending), and 1050 cm–1(Si–O–Si stretching).[36] Nanosilica was successfully coated onto the MCA surface, which was shown by the characteristic peaks obtained in the FTIR spectrum of MCA@SiO2 at 1630, 3357, 1057, 966, and 784 cm–1.[37] However, a peak at 590 cm–1 represented the presence of magnetic Fe3O4 particles.[38]

Figure 1

(a) FTIR and (b) XRD patterns of MCA, SiO2, and MCA@SiO2, (c) N2 adsorption–desorption BET, (d) pore size distribution of CP, MCA, and MCA@SiO2.

XRD patterns of MCA, SiO2, and MCA@SiO2 are shown in Figure b. Characteristic XRD peaks of MCA observed at 2θ = 30.01, 31.7, 35.4, 43.01, 45.42, 56.93, and 62.47° indicated the formation of Fe3O4 magnetic nanoparticles on the CP surface.[22,29] In the XRD spectrum of SiO2 nanoparticles, a broad peak at 22.83° reflected its amorphous nature.[37] The presence of the peak at 2θ = 22.32° in the XRD pattern of MCA@SiO2 confirmed the successful coating of nanosilica on MCA.[37] Intensities of MCA peaks in the XRD spectrum of MCA@SiO2 were relatively lower than pure MCA, which could be because of the coating of amorphous SiO2 on its surface.

N2 adsorption–desorption isotherm of CP, MCA, and MCA@SiO2 are presented in Figure c,d and different surface parameters are compiled in Table S1, Supporting Information. Brunauer–Emmett–Teller (BET) surface area and pore volume of CP improved significantly after the incorporation of Fe3O4 MNPs, which further improved after coating with nanosilica. Therefore, MCA@SiO2 displayed a higher surface area and pore volume than MCA and CP (Table S1, Supporting Information). According to Figure c, both MCA and MCA@SiO2 showed a type IV isotherm with a hysteresis loop. However, the hysteresis loop of MCA@SiO2 is more profound and distinct compared to MCA. The hysteresis loop of MCA@SiO2 can be classified as H2 type, indicating the existence of mesopores that probably arise from carbon, iron, and silica interparticle vacancies. The pore size distribution of the MCA@SiO2 sample exhibited pores ranging from 2 to 25 nm with an average pore size centered at 10.8 nm (Figure d). Overall, the assembling of carbon and iron-silica nanoparticles clearly showed enhancement in the surface area and porosity improvement in MCA@SiO2.

Free −OH and −COOH groups’ concentrations in MCA and MCA@SiO2 were determined using Boehm titration and in MCA the concentration of free −OH and −COOH groups was found to be 0.045 and 0.012 mequiv g–1, respectively. After coating with silica nanoparticles, the concentration of −OH groups increased to 0.062 mequiv g–1, whereas the concentration of −COOH increased from to a value of 0.015 mequiv g–1.

SEM images of MCA before and after coating with nanosilica were investigated as shown in Figure a–c. The presence of Fe3O4 magnetic nanoparticles and carbon spheres were observed in the SEM image of MCA and before nanosilica coating, the MCA surface appeared to be smooth and homogeneous (Figure a). Figure b,c shows monodispersed clusters of nanosilica on the MCA surface. Furthermore, the MCA@SiO2 surface was heterogeneous in nature and trapped by SiO2 particles, thereby confirming the successful coating of nanosilica onto the MCA surface. The average particle seize of the composite materials, that is, MCA@SiO2, was found to be 9.20 μm. Figure d shows elemental mapping of MCA@SiO2 and the presence of different elements such as silicon, oxygen, carbon, and iron onto the MCA@SiO2 surface also confirmed the coating of MCA with nanosilica.

Figure 2

SEM images of (a) MCA; (b,c) MCA@SiO2 at different magnifications (d) SEM mapping of MCA@SiO2; (e–g) TEM images of (e) MCA; (f,g) MCA@SiO2 at different magnifications, and (h) energy-dispersive system spectrum of MCA@SiO2.

TEM images of MCA and MCA@SiO2 are represented in Figure e–g. The TEM image of MCA (Figure e) clearly revealed the deposition of magnetic nanoparticles on the CP surface, which have spheres like morphology.[22] Furthermore, TEM images of MCA@SiO2 also showed that the MCA surface was covered by silica nanoparticles, which predicted the successful coating of nanosilica onto the MCA surface (Figure f,g), which was further supported by the Si peak present in the EDX pattern of MCA@SiO2 (Figure h).

Adsorption of MB by MCA@SiO2

Figure a shows the influence of MB concentration on MCA@SiO2 performance, which was studied by varying the MB concentration (50–500 mg L–1). Adsorption efficiency decreased progressively with increasing MB concentration. The higher adsorption efficiency at comparatively lesser MB concentration was because of the lower ratio of adsorption sites of adsorbent to the number of dye molecules to be adsorbed; on the other side, the lower adsorption efficiency at a comparatively higher initial dye concentration was because of the less availability of adsorption sites.[39]

Figure 3

Effects of (a) initial MB concentration with dye solution volume = 50 mL, solution pH = 7 and adsorbent dose = 0.5 mg L–1; (b) MCA@SiO2 dosage with MB concentration = 100 mg L–1, dye solution volume = 50 mL and (c) solution pH on adsorption efficiency with MB concentration = 100 mg L–1, dye solution volume = 50 mL, adsorbent dose = 0.5 g L–1; (d) plot of pH vs ΔpH for pzc determinations and (e) effect of different cations on adsorption efficiency with MB concentration = 100 mg L–1, dye solution volume = 50 mL and adsorbent dose = 0.5 g L–1.

Effect of MCA@SiO2 Dose

For practical applicability of MCA@SiO2, it becomes very important to adsorb a higher amount of target pollutant at a comparatively lower dose. Therefore, the effect of MCA@SiO2 mass on its adsorption performance was examined with varied adsorbent dose (0.1–0.7 g L–1) for MB adsorption. The results reported in Figure b clearly demonstrate that the adsorption efficiency for MB dye removal increased gradually with increased dose of MCA@SiO2 and almost 80% dye was removed from bulk solution with 0.5 g L–1 of MCA@SiO2. This increment in the adsorption efficiency was assigned to the combined effect of a larger surface area along with a higher concentration of adsorption sites for dye molecules’ attachment.[18,40,41] However, after attaining equilibrium adsorption, further increase in MCA@SiO2 mass did not alter the adsorption efficiency, which could be ascribed to the agglomeration of MCA@SiO2 particles at high concentrations which blocked some adsorption sites, reducing the total available adsorption surface area.[18,41]

Effect of Solution pH

MCA@SiO2 performance in the dye solutions having varied pH was examined by altering solution pH from 2 (strongly acidic) to 11 (strongly basic) Figure c. The performance of MCA@SiO2 for MB removal or the adsorption efficiency was shown to be extremely low in highly acidic pH conditions, that is, in the dye solutions having pH 2–4. As the pH increased to 6, the adsorption efficiency increased to 90% and achieved the equilibrium value of 99% at pH 7. This poor performance of MCA@SiO2 in the highly acidic dye solutions was due to the strong interference of free H+ ions, present in a high concentration in the strong acidic solutions with the adsorption of cationic dye molecules.[4,22,40] These H+ ions reduced the number of available adsorption sites of MCA@SiO2 for the attachment of dye molecules, which ultimately affected the adsorption efficiency or the performance of MCA@SiO2. But with changing pH of dye solution from strong acidic to weak acidic and moving toward alkaline pH, the concentration of H+ ions keeps on decreasing, which reduced competition among cationic dye molecules and H+ ions and subsequently increased available adsorption sites for dye adsorption and ultimately increased the adsorption efficiency. Moreover, while changing solution pH towards alkaline, protonated groups started becoming deprotonated, which also increased the chances of the presence of electrostatic interactions between MCA@SiO2 and MB.[4,40] Equilibrium adsorption efficiency was observed in neutral pH dye solution; however, further increasing the solution toward strongly alkaline did not change the adsorption efficiency and it remained almost constant. Adsorption efficiency of any adsorbent also depends upon its pzc and for cationic adsorbates higher adsorption is observed in solutions where pH > pzc; however, for anionic adsorbates, it is reverse, that is, adsorption is higher if solution pH < pzc.[18] Here,@@@ in this case pzc of MCA@SiO2 was found to be 4.41 (Figure d), therefore, the adsorption efficiency increased with increasing solution pH above pzc value. For further adsorption experiments, solution pH was kept neutral because approximately 99% dye adsorption was observed at this pH.

Interference of other Metal Cations of the Performance of MCA@SiO2

Further, to understand the adsorption mechanism of MB on MCA@SiO2 adsorption sites, the interaction between MB molecules and binding sites of adsorbent were studied. MB solutions with different metal cation strengths of NaCl and CaCl2 were used. As the metal cation strength increased in the respective solution, adsorption efficiency showed a gradual decrease (Figure e). Also, the performance of adsorbent for the dye removal reduced to a higher extent in the solution having Ca2+ ions (Figure e). These finding strongly support the possibilities of the presence of electrostatic interactions between MCA@SiO2 and MB.[40]

Adsorption Kinetics

The adsorption time for removing pollutant is very critical especially in case of dyes or heavy metal ions removal, therefore, adsorption kinetics were conducted using MB solutions of 50, 100, and 150 mg L–1. A rapid adsorption of dye was observed for all the three concentrations studied and the dye solutions became colorless very fast, 50 mg L–1 solution, became colorless within 3.5 min, whereas, 100 mg L–1 solution became colorless within 5 min and in 150 mg L–1 solution became colorless within 7 min (Figure S2a, Supporting Information). This very fast and rapid dye adsorption was because of large concentration easy availability of adsorption sites.

Furthermore, in starting, the rate of adsorption was higher which progressively decreased while moving towards the equilibrium. High adsorption rate in the starting was because of easy availability of binding sites of MCA@SiO2, whereas, the slow and gradual decrease in the adsorption rate with time was due to attaining saturation during the final steps of adsorption.[4] Moreover, equilibrium adsorption reached much faster in the dye solutions of lower concentration because stronger adsorption driving forces were present in lower concentration solutions which helped attaining equilibrium.[27]

To investigate adsorption kinetics and diffusion rate controlling steps, experimental kinetics data was fitted using Elovich (Figure S2b, Supporting Information), pseudo-first and pseudo-second-order models (Figure a,b).[41,42] Equations of these models and various parameters related to them are discussed in Table S2, Supporting Information. Most suitably fitted kinetics model was selected depending on the values of correlation coefficient (R2), akaike (AIC) and bayesian information criterion (BIC) (Table ) which showed that the kinetics of adsorption for MB removal using MCA@SiO2 followed pseudo-second-order model (Table ) predicting the binding of dye molecules onto MCA@SiO2 adsorption sites via chemisorption process involving the electrons sharing.[43] Value of initial sorption rate (ho) obtained using pseudo-second-order model was also reasonably higher predicting the presence of surface exchange reactions between MB dye molecules and MCA@SiO2.[44] Furthermore, kinetics data was also satisfactorily fitted to pseudo-first-order model and showed reasonally higher value of R2 which was close to unity as well as qe,cal values of all the three dye concentrations were in very much close proximity of qe,exp values which further predicted applicability of pseudo-first-order model. Therefore, the kinetics of MB adsorption onto MCA@SiO2 followed both pseudo-first and pseudo-second-order models. Similar trends for adsorption kinetics were also observed previously for remazol golden yellow dye adsorption using orange peel[45] and MB using activated carbon.[46] Furthermore, the value of R2 for Elovich model for all the dye concentrations studied were also quite high (Table S1) suggesting that the dye molecules attached to MCA@SiO2 particles via both ion exchange as well as chemisorption.[22]

Figure 4

Plots of (a) pseudo-first order; (b) pseudo-second-order; (c) intraparticle diffusion and (d) liquid-film diffusion models for the adsorption kinetics where MB concentration = 50, 100, and 150 mg L–1, adsorbent dose = 0.5 g L–1 and solution pH = 7.0.

Table 1

Kinetics Parameters for MB Adsorption on MCA@SiO2.

		dye concentration (mg L–1)
models	parameters	50	100	150
pseudo-first-order	k1 (min–1)	2.07	0.861	0.538
	qe,cal (mg g–1)	98.25	200.33	302.83
	qe,exp (mg g–1)	99.8	199.6	298.2
	AIC	22.05	36.23	49.85
	BIC	16.29	34.00	49.79
	R2	0.995	0.996	0.997
pseudo-second-order	K2 (g mg–1 min–1)	2.74 × 10–2	3.74 × 10–3	1.39 × 10–3
	qe,cal (mg g–1)	110.66	247.16	384.66
	qe,exp (mg g–1)	99.8	199.6	298.2
	ho	335.53	228.46	205.66
	AIC	7.32	20.95	31.33
	BIC	1.55	18.71	31.27
	R2	0.999	0.999	0.999
Elovich	B (g mg–1)	5.71 × 10–2	1.8 × 10–2	1.14 × 10–2
	kE (mg g–1 min)	1833.61	482.26	429.06
	AIC	25.22	37.57	52.64
	BIC	19.46	35.34	51.99
	R2	0.993	0.995	0.997

Dye transportation from bulk solution to the internal porous structure of adsorbents especially in case of hydrogels is a combination of different processes including: (a) passage of adsorbate from solution to outer phase of adsorbent surface (b) transportation of adsorbate from boundary layer to interior adsorbent surface and (c) relocation of adsorbate molecules to adsorption sites.[14] The slowest step out of the above mentioned steps controls overall diffusion rate of adsorbate molecules. Therefore, to understand the diffusion mechanism, adsorption kinetics data was further fitted using intraparticle diffusion (Figure c) and liquid film diffusion models (Figure d).[41] The plot of intraparticle diffusion model had two regions with their individual slopes which ruled-out the possibility of dye diffusion solely governed by intraparticle diffusion mechanism. These results also suggested the possible contribution of film diffusion mechanism but the plot of liquid-film diffusion model (Figure d) also did not fulfilled both the conditions of not having slope and passing through origin, therefore, it can be concluded from these studies that the dye diffusion collectively followed intraparticle as well as liquid film diffusion mechanisms.[23] Similar dye diffusion mechanism for MB dye was also reported for its adsorption using hazelnut shell powder.[47]

Adsorption Isotherm

Applicability of different isotherm models like Halsey, Jovanovich, Temkin, Dubinin–Kaganer–Radushkevich (DKR), Freundlich, and Langmuir models was checked for the experimental data at 25 °C (Figure a), 35 °C (Figure b) and 45 °C (Figure c).[48,49]

Figure 5

Plots of different two-parameter adsorption isotherm models at (a) 25; (b) 35 and (c) 45 °C where MB concentration = 50–500 mg L–1, adsorbent dose = 0.5 g L–1, solution pH = 7.0 and (d) vant’s Hoff plot for the determination of different thermodynamics parameters for the adsorption of MB onto MCA@SiO2.

Values of different isotherm parameters are tabulated in Table , however, the equations of these models are compiled in Table S4, Supporting Information. Experimental data agreed well with Langmuir monolayer model and exhibited maximum adsorption capacities of 516.9 (25 °C), 544.3 (35 °C) and 556.9 mg g–1 (45 °C) (Table ). Applicability of Langmuir model predicted that the single layer of adsorbate (dye molecules) uniformly covered MCA@SiO2 surface which have identical adsorption sites to attach dye molecules.[41] Moreover, the value of RL (between 0 and 1 for all three temperatures) further strengthened the Langmuir model applicability (Table ). In Temkin isotherm, increased value of β also suggested the existence of evenly distributed binding energies sites on MCA@SiO2 surface.[41] In Freundlich adsorption isotherm, 1/n values were less than 1.0, which suggested that the MCA@SiO2 was applicable over complete range of MB dye concentration studied.[23,41]

Table 2

Different Adsorption Isotherm Parameters Derived Using Non-Linear Equations of Various Two-Parameter Adsorption Isotherm Models for the Adsorption of MB Onto MCA@SiO2.

		temperature (°C)
isotherm models	parameters	25	35	45
Langmuir	qm (mg g–1)	516.9	544.3	556.9
	KL (L mg–1)	0.223	0.295	0.516
	RL	0.008–0.082	0.006–0.062	0.003–0.037
	AIC	37.16	37.47	52.84
	BIC	34.07	34.38	49.75
	R2	0.999	0.999	0.999
Freundlich	KF (L mg–1)1/n(mg g–1)	198.29	222.0	257.92
	1/n	0.188	0.180	0.160
	AIC	89.16	91.31	92.81
	BIC	86.07	88.21	89.72
	R2	0.843	0.833	0.823
Temkin	β (mg g–1)	33.156	34.79	35.77
	KT (L mg–1)	6.680	10.224	28.18
	AIC	88.41	83.18	85.34
	BIC	85.32	80.09	82.25
	R2	0.931	0.926	0.916
DKR	qm (mg g–1)	471.90	502.73	522.54
	KDR (mol2J–2)	1.56 × 10–4	1.123 × 10–6	5.04 × 10–7
	Es (kJ mol–1)	0.0566	0.667	0.996
	AIC	80.88	88.54	88.34
	BIC	77.79	85.45	85.35
	R2	0.854	0.873	0.886
Jovanovic	qm (mg g–1)	488.11	515.74	533.86
	KJ (L mg–1)	0.154	0.204	0.340
	AIC	73.90	75.95	77.57
	BIC	70.81	72.86	74.48
	R2	0.965	0.964	0.961
Halsey	KH (mg g–1)	1.453 × 1012	1.037 × 1013	1.20 × 1015
	n	5.294	5.544	6.25
	AIC	89.16	91.31	92.81
	BIC	86.07	88.21	89.72
	R2	0.843	0.833	0.823

Experimental isotherm data was further fitted to three-parameter isotherm models namely Redlich–Peterson, Radke–Prausnitz, BET, Koble–Corrigen, Hill, Toth, Sips, Brouers–Sotolongo and Vieth–Sladek models.[48,49] Mathematical expressions of these models are given in Table S4, Supporting Information. Plots of all these models at different temperatures are represented in Figure S3, Supporting Information and various parameters obtained are tabulated in Table S5, Supporting Information. On comparing the values of different isotherm parameters obtained it was observed that the dye adsorption on MCA@SiO2 followed Toth, Radke–Prausnitz and Redlich–Peterson models among all three-parameter isotherm models studied which further supported possibility of homogeneous monolayer adsorption and applicability of Langmuir model.

Adsorption Thermodynamics

Study of the thermodynamics for an adsorption process is very important to find out either adsorption occurs spontaneously or not. Different thermodynamics parameters were calculated using the following equationswhere, KL represents distribution coefficient; ΔS° denotes standard entropy change, ΔH° denotes standard enthalpy and ΔG° denotes standard free energy change. Plot of ln KL versus 1/T was plotted (Figure d) to calculate values of ΔH° and ΔS°. Value of different thermodynamic parameters were tabulated in Table . Negative and decreasing value of ΔG° with increasing temperature predicted spontaneity of MB adsorption onto MCA@SiO2 which was more favorable at comparatively lower temperatures and have high affinity of adsorbate on adsorbent surface.[50] Furthermore, positive value of ΔH° also suggested endothermic nature of adsorption. Nature of adsorption process that is chemisorption or physical adsorption also depends upon value of ΔH°. For chemisorption it should be 20.9–418.4 kJ mol–1 but for physical adsorption it should be 2.1–20.9 kJ mol–1.[51] In this case, the value of ΔH° is 51.81 kJ mol–1, therefore, MB was adsorbed onto MCA@SiO2 through chemisorption process, which was also predicted by adsorption kinetics. Moreover, positive value of ΔS° indicated increased entropy and disorders at solid–liquid interface (Table ).

Table 3

Various Thermodynamics Parameters for the Adsorption of MB onto MCA@SiO2.

ΔG° (kJ mol–1)
25 °C	35 °C	45 °C	ΔH° (kJ mol–1)	ΔS° (kJ mol–1 K–1)
–185.49	–207.50	–232.73	51.81	235.84

Adsorption capacity of MCA@SiO2 was compared with other adsorbents which are already reported (Table ). MCA@SiO2 showed good adsorption capacity as compared to these adsorbents suggesting very good chances of its future applicability in large scale industrial applications.

Table 4

Comparison of the Adsorption Capacity of MCA@SiO2 with Other Adsorbents for MB Adsorption

adsorbent	pH	adsorption capacity (mg g–1)	refs
activated carbon prepared from apricot	5–10	36.68	(5)
hollow (PCPP) microspheres	10.0	50.7	(52)
Co0.3Ni0.7Fe2O4@SiO2 membrane	7.0	107.5	(53)
Pine tree leaves biomass	7.06	126.58	(7)
polypyrrole/CNTs–CoFe2O4	3–10	137	(54)
mesoporous silica from coal gasification	7.30	140.57	(55)
MCA from corncob	7.0	163.93	(29)
diatomite	8.0	198	(56)
commercial activated carbon	5–10	199.60	(5)
jute fiber carbon	4.0	225.64	(57)
halloysite nanotubes cyclodextrin nanosponges		226.0	(58)
mGO/PVA composite Gel	10.0	270.94	(59)
activated boron nitride fibers	8.0	392.2	(60)
gelatin–CNT–MNPs	7.0	465.5	(61)
MCA@SiO2	7.0	516.9	present work

The ability of MCA@SiO2 for decolorization of MB dye via adsorption process was also compared with some other processes like enzymatic treatment, coagulation coupled adsorption and combined chemical coagulation, electrocoagulation, and adsorption processes (Table ) for the decolorization of different dyes. It was observed that the MB dye decolorization ability of MCA@SiO2 via adsorption process was quite good and gave good results when compared with other methods of dyes decolorization.

Table 5

Comparison of the Dyes Decolorization Process of MB Using MCA@SiO2 via Adsorption Process with Other Processes Dyes Decolorization Process

material	process	dyes	adsorption conditions	% decolorization	references
purified laccase of paraconiothyrium variabile immobilized on porous silica beads	enzymatic treatment	acid blue 25	solution pH = 5; dye concentration = 400 mg L–1; enzyme concentration = 2 U mL–1	76	(12)
purified laccase of Paraconiothyrium variabile immobilized on porous silica beads	enzymatic treatment	acid orange 7	solution pH = 5; dye concentration = 400 mg L–1; enzyme concentration = 2 U mL–1	64	(12)
low-purity Horseradish peroxidase extracted horseradish roots	enzymatic treatment	acid orange 7	solution pH = 7; enzyme concentration = 1 U mL–1	73	(62)
chemical coagulation (poly ammonium chloride), electrocoagulation (aluminum electrodes) and adsorption (pistachio nut shell ash)	combined chemical coagulation, electrocoagulation, and adsorption processes	mixture of dyes	solution pH = 7.9; dye concentration = 600 mg L–1; adsorbent dose = 0.6 g L–1	99	(10)
l-arginine-functionalized Fe3O4 nanoparticles	adsorption	reactive blue 19	solution pH = 3; dye concentration = 50 mg L–1; adsorbent dose = 0.74 g L–1	96.34	(24)
oak acorn peel	coagulation coupled adsorption	MB	solution pH = 10; dye concentration = 600 mg L–1; adsorbent dose = 0.5 g L–1	97	(63)
Melaleuca diosmifolia	coagulation coupled adsorption	MB	solution pH = 7; dye concentration = 600 mg L–1; adsorbent dose = 0.5 g L–1	95	(64)
Melaleuca diosmifolia	coagulation coupled adsorption	Eriochrome Black T	solution pH = 2; dye concentration = 600 mg L–1; adsorbent dose = 0.5 g L–1	93.5	(64)
MCA@SiO2	adsorption	MB	solution pH = 7; dye concentration = 50 mg L–1; adsorbent dose = 0.5 g L–1	99	this work

Selective Adsorption of MB

Selective adsorption of MB onto MCA@SiO2 particles were also examined by studying the adsorption behaviour of MCA@SiO2 for a mixture of cationic that is MB and anionic that is MO dye in different concentration ratios (Figure ). It was observed that the color of mixed dye solution became green for both the mixtures that is 1:1 ratio of MB and MO (Figure a) and 5:1 ratio of MB and MO (Figure b) but after adsorption color of both the solutions again changed back to orange which was the original color of MO dye solution. Furthermore, UV–vis spectra of dye mixture solutions before adsorption showed peaks of both the individual dyes that is MB (660 nm) and MO (464 nm) whereas, after adsorption peak of MB disappeared but the peak of MO was still present at the same position having almost the same intensity. Therefore, from these findings it can be concluded that MCA@SiO2 have ability to adsorb cationic dyes selectively from a mixture of different dye solutions.[65]

Figure 6

Plots of the selective adsorption of MB onto MCA@SiO2 from the mixture of MB and MO dyes having concentrations (a) MB/MO = 1:1 and (b) MB/MO = 5:1.

Column Tests

The dynamic adsorption performance of MCA@SiO2 to adsorb MB was also tested using the column studies. MB solution of 50 mg L–1 concentration became almost colorless after passing through the column (Figure ). Figure a shows the images of column packed with MCA@SiO2 before, during and after the filtration of MB. Almost 430 mL dye solution of 50 mg L–1 concentration became colorless and its concentration became zero, thereafter, the concentration of effluent passed through the column started increasing slowly and after passing 500 mL dye solution from column the concentration of MB filtered became 1.9 mg L–1 (Figure b). Furthermore, the adsorption efficiency of column increased continuously as the volume of MB increased which is clearly seen in the Figure c. This increased adsorption efficiency with continuously increasing effluent was attributed to the high adsorption capacity of MCA@SiO2 along with the fast adsorption rate.[65] The process of filtration of MB through column packed with MCA@SiO2 is shown in Video S1, Supporting Information. Figure d also shows relationship between the concentration of MB effluent after filtration and volume of effluent passed or filtered through column packed with MCA@SiO2. In the figure red line shows the initial concentration that is Co of MB passed through the column and it was observed that even after passing 500 mL of MB the concentration of effluent after filtration was as low as 1.9 mg L–1, therefore, from these studies it can be predicted that MCA@SiO2 have potential to treat larger quantities of MB contaminated wastewater in industrial applications.

Figure 7

(a) Images of column of MCA@SiO2 for the adsorption of MB before, during and after adsorption where MB concentration = 50 mg L–1; (b) UV–vis spectra of every 10 mL (up to 500 mL) of MB before the filtration through the column packed with MCA@SiO2; (c) plot of the adsorption efficiency of MCA@SiO2 vs dye effluent volume and (d) relationship between volume of dye treated and concentration of dye after filtration through column.

Furthermore, different concentration solutions MB (50, 100, and 150 mg L–1) were also passed through the column by applying an external force (see Video S2, Supporting Information). For these experiments, initially, 10 mL solutions of each 50, 100, and 150 mg L–1 were quickly passed through column under the influence of an external force, respectively (Figure ). 10 mL of each MB solution became colorless and their absorbance reduced nearly to zero which conforms the dynamic adsorption behavior of MCA@SiO2.

Figure 8

(a) Images of MB with different initial concentration of 50, 100, and 150 mg L–1 before filtration; (b) images of column of MCA@SiO2 packed in syringe before, during and after filtration under pressure and picture of effluent of MB after filtration and (c) UV–vis plot of MB (10 mg L–1) before filtration and 50, 100 and 150 mg L–1 after filtration through syringe column under pressure.

Regeneration Studies

For the practical utility of any adsorbent, its multiple time use is very crucial. Therefore, the reusability experiments for the multiple time adsorption of MB using MCA@SiO2 were performed for consecutive 20 cycles of adsorption–desorption. Initially, MB dye was adsorbed onto MCA@SiO2 particles then dye loaded adsorbent particles were separated and dried in hot air over. Thereafter, these particles were added in 50 mL acetone and agitated at 120 rpm for 60 min followed by the separation. Adsorbent particles were washed repeatedly using deionized water, dried and again used for adsorption–desorption. It was observed that the performance of MCA@SiO2 particles for MB adsorption remained constant that is almost 100% for the first seventeen cycles, however, in the last three cycles the adsorption capacity decreased little much which was also very less (Figure S4, Supporting Information), therefore, it can be used again and again for the treatment of MB contaminated wastewater.

Adsorption Mechanism

Adsorption capacity or the performance of any adsorbent not solely depends upon the pore structure and specific surface area but also influenced by many other factors like the nature of binding sites of adsorbent, adsorbate–adsorbent interactions, charge of the adsorbate surface at a particular solution pH. Mostly, adsorbate molecules attach to adsorbent surface through interactions such as hydrogen bonding, ion-exchange, dipole–dipole interactions, hydrophobic interaction, coordination by surface metal cations and electrostatic interactions.[66] In case of dyes adsorption, dye molecules mostly attach to the adsorbent surface through electrostatic interactions, here in this case also we believe that MB dye molecules attached to MCA@SiO2 surface through electrostatic interactions. H+ ions ionized in the dye solution and MB+ ions possible attach to (SiO)− ions present on adsorbent surface using electrostatic interactions.[55] In strongly acidic solutions, the adsorption was much less which kept on increasing while changing pH from strongly acidic to weak acidic and neutral (Section ), which might be due to comparatively higher negatively charged surface of adsorbent at higher pH solutions. This observation also supports the electrostatic interactions between MB and MCA@SiO2. On the other hand, performance of MCA@SiO2 was also good in weakly acidic pH solutions which also predicted the presence of hydrogen bonding between the nitrogen atom of MB and silanol group of MCA@SiO2.[55,56] Furthermore, the presence of Na+ and Ca2+ ions also reduced MCA@SiO2 performance for MB removal to a large extent (Section ), which further supported the idea of the presence of electrostatic interactions. Therefore, these observations predicted that MB dye molecules were attached to MCA@SiO2 surface via both electrostatic interactions as well as hydrogen bonding. The possible adsorption mechanism showing the presence of both electrostatic interactions as well as hydrogen bonding is shown in Scheme .

Scheme 2

Proposed Mechanism for the Adsorption of MB onto MCA@SiO2 and the Presence of Possible Interactions between MB Dye Molecules and Binding Sites of MCA@SiO2

Figure S5a, Supporting Information shows the FTIR of MCA@SiO2 after dye adsorption in which intensity of some of the peaks reduced while some other peaks were either shifted to some other positions or completely disappeared as compared to the FTIR of MCA@SiO2 before adsorption. The peak representing Si–O–H stretching, initially present at 966 cm–1 reduced considerably suggested that H+ ions ionized in dye solution and interactions of MB+ ions and (SiO)− through electrostatic interaction mechanism.[55] Moreover, peak intensities representing Si–O–Si vibrations at 1057 and 784 cm–1 reduced considerably due to the reduced transmittance of Si–O–Si bond because dye molecules covered MCA@SiO2 particles.[55] These changes in the FTIR spectrum of MCA@SiO2 suggested that MB molecules attached to MCA@SiO2 via electrostatic interactions as well as hydrogel bonding. In the SEM images of MCA@SiO2 after the adsorption of MB dye molecules (Figures S5b,c, Supporting Information), the presence some particles covering the adsorbent surface was observed which might be possibly due to the presence of MB dye molecules. Therefore, above studies concluded that MB molecules attached to MCA@SiO2 via both electrostatic interactions as well as hydrogen-bonding as shown in Scheme .

Conclusions

MCA were successfully synthesized from starch using the low temperature hydrothermal technique and nanosilica was successfully coated on its surface. Properties of MCA especially surface properties were significantly improved after coating with nanosilica. MCA@SiO2 was utilized successfully as adsorbent for removing MB and the adsorption process followed homogeneous monolayer Langmuir isotherm with 516.9 mg g–1 adsorption capacity, whereas, very fast and rapid adsorption of dye was attributed to the applicability of both pseudo-first and pseudo-second-order rate equations. Adsorption mechanism mainly involved electrostatic interactions and hydrogen bonding. Furthermore, diffusion of dye molecules was collectively governed by intraparticle diffusion and liquid film diffusion mechanism. MCA@SiO2 particles selectively adsorbed MB from a mixture of cationic and anionic dyes. Column studies predicted that MCA@SiO2 can used in large or industrial scale treatment of MB contaminated wastewater. It was further used for continuous 20 cycles of adsorption–desorption. Therefore, from the aforementioned studies it can be conclude that MCA@SiO2 have all the properties to be used as modern adsorbents to effectively treat cationic dyes contaminated wastewater.