Abstract

Zirconium oxide/activated carbon (Zr3O/AC) composite was synthesized to remove methylene blue (MB) and crystal violet (CV) from the aqueous medium. The Zr3O/AC sample was characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), energy dispersive X-ray analyses (EDS), Raman spectroscopy (RS), BET surface area, and Fourier transform infrared spectroscopy (FTIR). XRD profiles confirmed the successful synthesis of the zirconium oxide/activated carbon composite. SEM images showed multideveloped walls with irregular particle size with channel arrays. The nitrogen physisorption combines I and IV types with a calculated BET surface area of 1095 m2/g. Raman spectrum illustrated a disorder of both crystalline structure and the graphitic structure. The adsorption was better fitted to the pseudo-second-order (PSO) kinetic model. Langmuir model fitted better the experimental results of MB adsorption, whereas the CV was better consistent with the Freundlich model. The obtained results suggested that the MB and CV adsorption might be influenced by the mass transfer that involves multiple diffusion steps. The maximum adsorption capacities are 208.33 and 204.12 mg/g for MB and CV, respectively. The MB and CV removal mechanisms were proposed, and statistical optimization was performed using central composite design combined with the response surface methodology.

1. Introduction

Nowadays, many industrial activities such as coloring, textile, cosmetic, paper, food, paint, printing, and pharmaceuticals are considered a big concern from the environmental point of view as many hazardous synthetic dyes are generally produced [1]. These pollutants are regarded as reactive, acid, and basic and are very harmful to environment including vegetation, humans, and animals [2]. Methylene blue (MB) is a common basic dye that is used for dyeing clothes in textile industries [3]. MB can cause the adverse effects such as cyanosis, vomiting, diarrhea, tachycardia, and jaundice on human beings. [3, 4]. Crystal violet is another type of toxic pollutant with a poor degradability that is widely used in textile industries because of its low cost and high solubility in water [5, 6].

Various processes are effectively investigated for the removal and elimination of pollutants like chemical precipitation, adsorption, membrane filtration, biological treatment, and photocatalytic degradation [7, 8]. Among these techniques, adsorption is the most effective and broadly utilized approach due to its many advantages being not only the relatively low cost and ease implementation of the operation process [9] but also the effectiveness of the adsorption process that primarily depends on the performance of the adsorbent mass transfer and thermodynamics. Nevertheless, traditional adsorbents including carbonaceous materials, metal nanoparticle oxides, and polymer resins suffer from either low adsorption capacities or low efficiency [1016]. Therefore, in the general framework of environmental and waste management, there is a need to synthesize environmental recyclable and friendly materials with better uptake and thermodynamic and kinetic properties. Recently, ZrO2 or zirconium-based composites was used as an adsorbent in the gas and liquid phase. More precisely, zirconia ZrO2 was used for CO2 adsorption and hydrogen adsorption to understand the relation to the dehydrogenation of hexane [17, 18]. While in recent works, zirconium and zirconia was used with activated carbon (AC) as composites for the removal of pollutants in water. For instance, Mullick and Neogi [19] have impregnated zirconium on powdered activated carbon (AC) using ultrasound as the tool for synthesis and applying it for fluoride adsorption from water. However, Xiong et al. [20] have used iron-zirconium-modified activated carbon nanofiber for the adsorption of phosphates. Similar works have used these types of composite for the removal of other pollutants [21, 22].

In the present work, commercial suboxide Zr3O nanoparticles have been fixed onto AC to improve the sorption capacity of MB and CV dyes. This suboxide is unique because of close-packed Zr layers in H stacking and O in octahedral voids, and additionally, the OZr6 octahedra share vertices to form a 3D framework. This 3D framework is very beneficial if it is fixed within the prepared adsorbent leading to high density interactions. The effect of various parameters such as mass loading of Zr3O/AC, pH, contact time, dye concentration, temperature, and regeneration was studied. In addition, the mechanism of MB and CV onto Zr3O/AC and adsorptions experiments optimization were investigated using FTIR analysis after adsorption and RSM-CCD method. It should be noted that the optimization of the adsorption system using classic adsorption and experimental protocol based on changing each single variable does not give useful information about variables. Hence, RSM coupled with the CCD method was considered as a new approach and helpful allowing changing more than a variable at one time and studying the interaction effects between operational parameters [8, 2329].

To the best of our knowledge, this work reports an effective methodology to study the adsorption of CV and MB, a straightforward study to calculate the optimum parameters through surface modeling, and a sorption mechanism based on different techniques before and after adsorption.

2. Materials and Methods

2.1. Synthesis of the Activated Carbon

The chemicals used in this study were of analytical reagent grade and used without purification. The activated carbon-based almond shell was prepared by chemical activation using KOH. For this purpose, the almond shell was mixed with KOH solution (Sigma-Aldrich, 98%), 10 g of KOH (Sigma-Aldrich, 90%), with 20 g of almond shell and 100 mL of distillated water to obtain a solution which was heated at 60°C for 12 h and then dried at 110°C. Then, the dried sample was pyrolyzed under N2 flow (200 cm3/min) at 300°C for 2 h and then at 800°C for 3 h at a heating rate of 10°C/min. The Zr3O/AC composite was prepared by impregnation of 20 wt.% of zirconium oxynitrate ZrO(NO3)2·xH2O (Sigma-Aldrich, 99%) on the prepared activated carbon after 6 hours of stirring at room temperature. Finally, the impregnated almond shell was dried at 70°C for 12 h and then pyrolyzed at 800°C for 1 hour.

2.2. Adsorption Experiments

The adsorption batch system tests were conducted in glass beakers (150 mL) containing 50 mL of CV or MB solutions with a constant stirring rate. Then, the influence of different key parameters was fully investigated. After a certain time, the pollutants concentration was determined using a UV-Vis spectrophotometer (UV 2300) at their respective maximum wavelengths of 664 nm for MB and 582 nm for CV. The adsorbed amount and removal percentage at different conditions were calculated using the following equations:where (mg/L) and (mg/L) are the initial and equilibrium liquid-phase concentrations, (g) is the mass of Zr3O/activated carbon, and (L) is the volume of the MB/CV solution.

Langmuir [30] and Freundlich [31] were used to analyze the equilibrium data, and the adsorption isotherm models were also applied to simulate information about the distribution of adsorbate molecules at the solid-liquid interface. Table 1 summarizes the mathematical parameters of these models. However, the equations of the selected kinetics models to fit the experimental data are listed in Table 2.

3. Results and Discussion

3.1. Structural and Textural Characterizations

Figure 1 shows the XRD patterns of the synthesized zirconia-AC powder, and the diffractogram shows two aspects: amorphous aspect in the range of 20–30° and amorphous aspect in the range of 40–50°, which are characteristics to carbon/two-dimensional reflections of turbostratic layer stacks [32] and to graphitic structure of activated carbon. The intense diffraction profiles are associated with the crystalline hexagonal zirconium oxide Zr3O (JCPDS card no 03-065-7450), the calculated lattice parameters for zirconia by Le Bail fit are a = b = 5.613 (2) and c = 5.192 (3), and these parameters are very close to the standard Zr3O JCPDS zirconia file.

Thermal analyses, illustrated in Figure 1(b), show the thermal decomposition of zirconia/AC composite as a function of temperature. This experiment was performed under air. Two major weight losses were recorded: the first loss of 15% below 100°C is attributed to the surface-adsorbed H2O [33], while the second weight loss of 60% in the range of 300–480°C might be originated from thermal decomposition organic matters such as lignin, cellulose, and hemicelluloses of the almond shell. An intense endothermic peak was recorded for this loss showing the remarkable amount of decomposed products [34].

Figure 2 illustrates the morphology and elemental dispersive spectroscopy of Zr3O/AC composite. In the SEM micrographs, multideveloped walls were observed, and pores can be seen with irregular particles and channel arrays in the range of few micrometers to few hundred micrometers. Clearly, knowing that almond shell contains micropores, it is inferred that the AC preserved the original structure of the almond after even after carbonization. The results of elemental analysis indicate the presence of three main atoms: carbon, oxygen, and zirconium.

Raman analysis is a very important tool for the characterization of active modes in the activated carbons adsorbents [35, 36]. Clearly, all the samples have two obvious D and G bands, which are assigned to the chaos and disorder of the crystalline structure and the graphitic structure of carbon materials, respectively [37, 38]. The synthesized Zr3O/AC and AC were studied by Raman analysis, and the results are presented in Figure 3(a). The spectra displayed the typical spectra of AC with the characteristic D and G bands at 1362 and 1604 cm−1, respectively. Furthermore, the intensity of Zr3O/AC decreased indicating that the preparation process of Zr3O/AC slightly changed the structure of activated carbon. Moreover, this means that the degree of graphitization of Zr3O/AC decreases because zirconia occupies the pore channel [39].

The N2 physisorption isotherm is illustrated in Figure 3(b). It combines types I and IV characteristics, meaning the coexistence of microporous and mesoporous structures in AC and Zr3O/AC. Indeed, the distributions of pores size confirm the presence of micropores and mesopores in AC and Zr3O/AC, and pore size distributions shown in Figure 3(c) are bimodal with a peak at around 0.5 nm for microporous structures and around 2 nm for mesoporous structures. The Zr3O/AC BET surface area is 1095 m2/g with total pore volume of 0.66 cm3/g (0.16 cm3/g of microporous volume and 0.56 cm3/g of mesoporous volume).

The FTIR analysis of Zr3O/AC composite is displayed in Figure 3(d). In the spectrum, a broad and strong absorption band at approximately 3449 cm−1 attributed to the O-H stretching vibration of water molecules adsorbed on the surface Zr3O/AC [40]. The spectrum of Zr3O/AC presents a peak for Zr-O stretching vibration at 620 cm−1 [41]. However, the band centered on 1622 cm−1 indicates enhancement in the aromatic C=C groups [27, 42].

3.2. Adsorption Results
3.2.1. Influence of Operational Parameters

Binding sites and surface promoted by Zr3O/AC doses are one of the main parameters influencing the adsorption process. Consequently, several experiments of the dyes removal with different Zr3O/AC doses have been conducted to evaluate the effect of these key factors. These experiments have been performed at room temperature unless mentioned. The influence of Zr3O/AC doses on MB and CV dyes removal by Zr3O/AC is presented in Figure 4; As the Zr3O/AC composite weight increased from 5 mg to 16 mg, the percentage removal increased as well from 84.43% to 100% for MB removal and from 75.41% to 98% for CV removal. The increase in MB and CV removal at higher dosages explain (i) the increase in number of active binding sites and (ii) bulk surface area [43].

Additionally, up to 20 mg, the dyes removal was not significantly affected by further increase in the Zr3O/AC amount due to the possibility of aggregation, which limits the active surfaces for adsorption. The optimum Zr3O/AC dosage was found to be 16 mg for both dyes.

Figure 5(b) illustrates the removal of MB and CV dyes as a function of solution pH. Knowing that pH can play a substantial role in the adsorbent surface charges and interfacial solid-liquid transport [44], the pH-dependence experiment has been performed for both MB and CV. As the pH increases, adsorption of both dyes on Zr3O/AC was found to be increasing; this is mainly due to elimination of the positive proton in the aqueous media thereby promoting activities of electrostatic attraction between the positive charges of both dye and the composite Zr3O/AC. Indeed, according to the point of zero charge (Figure 5(a)), it is confirmed that the surface of the composite can be negatively charged at pH higher than 7.5. While at an acidic pH, the functional groups of activated carbon become protonated, which are mainly the carboxylic (-COOH-), phenolic (-OH-), and chromonic group. At pH 3, the surface charge of Zr3O/AC becomes more positively charged, which enhances MB and CV electrostatic repulsion: the number of positively charged sites on Zr3O/AC increased and negatively charged sites on Zr3O/AC decreased. Therefore, the electrostatic repulsion between the Zr3O/AC sites and the positively charged CV and MB dye molecules increased which inhibit the adsorption pathway.

Figure 6 describes the influence of contact time on MB and CV adsorption by Zr3O/AC. It was observed that the adsorption process of both dyes takes place in two different stages:(i)The first stage was rapid which infers strong electrostatics between cationic dyes and the Zr3O/AC positive surface.(ii)The second stage was with no significant improvement of the adsorbed dyes, and this equilibrium stage could be due to the fast saturation of the binding sites in the Zr3O/AC surface.

3.2.2. Kinetics and Isotherm Adsorption

In this section, pseudo-first-order (PFO), pseudo-second-order (PSO), and intraparticle diffusion (IPD) kinetic models were used to fit the experimental data of MB and CV adsorption on Zr3O/AC. It is well established that the pseudo-first-order kinetic model has been widely applied for the adsorption of solid/liquid systems [4547].

Table 3 lists different calculated parameters using PFO and PSO models. Figure S1 (Supporting information) shows the fitting of experimental data to the PSO model at different initial concentrations of both MB and CV dyes, and the correlation coefficient is more significant than those of the PFO model. Moreover, a large difference of Qe between the experiment (Qe,exp) and calculation (Qe,cal) is observed [48]. These data support the fact that the adsorption data are well represented by the PSO kinetics model for the entire adsorption period of MB and CV dyes onto Zr3O/AC [49].

In order to gain more insights into the kinetics of the adsorption process, the IPD model was studied. The plot (Qt vs t1/2) (Figure 7) shows an initial curved portion followed by a linear portion and a plateau.

It is clear that the adsorption processes of both MB and CV dyes can be roughly divided into three steps, as indicated by the guidelines for each portion: The first portion is attributed to the diffusion of the MB and CV molecule through solution to the external surface of the Zr3O/AC composite. This is a fast process that mainly depends on the surface area of the Zr3O/AC. The second portion is attributed to gradual equilibrium with IPD as the rate-determining step. The last step is due to an adsorption/desorption equilibrium which is established for both MB and CV dyes [8]. As can be seen from the Figure 7, the Zr3O/AC exhibits faster adsorption kinetics at the initial adsorption stage due to its much higher surface area, as inferred in physisorption measurement.

The adsorption isotherm models were used to investigate the interaction between the MB and CV dyes and the Zr3O/AC when the adsorption process reaches equilibrium. Figure S2 shows plots of adsorption data of MB and CV dyes fitted to Langmuir and Freundlich isotherm models. Table 4 summarizes the isotherm parameters for this study. It is obvious that the Langmuir model provides a better fit for MB and Freundlich models for CV. A good fit with the Langmuir model indicate monolayer adsorption. However, the Freundlich model is indicating multilayer adsorption.

We have provided an extensive comparison of Zr3O/AC to the best performing materials reported to date in Table 5. Our prepared material is one of the best adsorbents for the removal of both MB and CV.

3.2.3. Regeneration of Zr3O/AC

Reusability was carried out by regenerating Zr3O/AC with 1% sodium hydroxide at room temperature for 3 hours. The regenerated Zr3O/AC was then washed with Milli-Q water and dried in a vacuum oven.

The Zr3O/AC was separated and washed with distilled water. Recycling efficiency of Zr3O/AC was investigated for the removal of MB and CV for 3 cycles. The results show that the adsorption efficiency of Zr3O/AC was reduced to 88% for MB and to 82% for CV dye (Figure 8). We should note that, after every cycle, NaOH was used as a desorption medium to remove adsorbed MB ions from the Zr3O/AC surface.

3.2.4. Statistical Analysis and Model Fitting

The significance of the independent variables and their interactions was tested with various statistical analyses such as analysis of variance (ANOVA), correlation coefficient (R2), and adjusted correlation coefficient () [58]. The Box-Behnken matrix used in this work is presented in Table 1S. The analysis of variance results (Table 6) suggests that the developed models (Equations (2) and (3)) are highly significant with a very low value for all models:where , , and represent the adsorbent dose, initial dyes concentration, and contact time, respectively.

The R2 values of full second-order polynomial regression models were found to be 0.951 ( 0.895) and 0.972 ( 0.952) for MB and CV adsorption on Zr3O/AC, respectively [8]. Moreover, the control test was used to determine the significance of coefficients of the parameters [59]. The linear, quadratic, and interaction effects of the parameters are given in Table 7. The results confirm that most of the coefficients models were significant ( values are less than 0.05) except quadratics effects of contact time, initial concentration, and the interactions effects of CT-IC and AD-CT, which are statistically insignificant ( values are more than 0.05). The same conclusion was observed in the case of CV adsorption onto Zr3O/AC.

3.2.5. Optimization of CV and MB Using RSM

Pareto diagrams (Figure S3) indicate the percent influence of different independent variables on MB and CV adsorption onto Zr3O/AC. An observation of the Pareto diagram makes it possible to say that the Zr3O/AC dose and initial concentration factors account more than 84.81% of the variation of the response in the case of MB adsorption onto Zr3O/AC. For the adsorption of CV onto Zr3O/AC, the Zr3O/AC dose and the quadratic effects of Zr3O/AC dose, contact time, and initial concentration are the most influential parameters on the adsorption efficiency, with an influence percentage of 48.59, 18.70, 16.18, and 15.12%, respectively.

From this study, it can be clearly seen that the Zr3O/AC dose (with a positive influence) and the initial concentration (with a negative influence) are the first factors influencing the MB and CV adsorption onto Zr3O/AC or Zr3O/AC. This probably returns to the availability of active sites, when the Zr3O/AC mass of the support increases [60].

To achieve the maximum adsorption of the CV and MB dyes onto Zr3O/AC and AC, the RSM modeling was used (Figures 9(a) and 9(b)). The optimum values of different independent variables (initial dye concentration, Zr3O/AC dose, and contact time) are determined as shown in Table 8. It shows that the model-predicted values are in very good agreement with the experimentally determined values which confirm that the proposed model combined with RSM is an effective approach for modeling the adsorption process and to understand the relationships between the independent variables and response.

3.2.6. Adsorption Mechanism

From the FTIR, Raman, and EDS analyses, a large number of hydroxyl functional groups and carbonyl functional group on the surface of Zr3O/AC were observed. Knowing that the prepared material has a very high surface area, the functional groups are participating in the adsorption mechanism of CV and MB as a function of solution pH and the load of Zr3O/AC on the surface. After CV adsorption, the FTIR spectrum of the Zr3O/AC, presented in Figure 10(a), shows characteristic changes due to the adsorption of CV onto Zr3O/AC: the intensity of the peak representing the stretching vibrations of the O-H group shifted and the peak representing the stretching vibrations of the C-C group diminished. This might be due to the formation of chemical bonds between C-C of Zr3O/AC and electrophilic N+ of CV. Furthermore, the Zr-O peak shifted to higher wavenumbers suggesting a formation of chemical bonds between Zr-O and N+ of CV dye. The same behavior was observed in the case of MB adsorption on Zr3O/AC. As shown in the FTIR spectroscopy after MB (Figure 10(b)), -OH, C=C, and Zr-O peaks shifted to different frequencies, confirming the interaction between these groups and our MB dye. Figure 10(c) shows the proposed model of MB and CV adsorption onto Zr3O/AC, and similar observations were reported elsewhere [8]. From this study, the mechanism can be divided into these 4 fundamental steps:(i)The adsorption of MB and CV molecules inside the Zr3O/AC pores and the interaction between Zr-O and nitrogen of CV and MB(ii)The hydrogen bonding would be produced between hydroxyl functional groups of Zr3O/AC and the nitrogen of CV and MB(iii)The π-π interaction between the double bonds of AC and molecule and that of pollutants

4. Conclusion

Zr3O/AC was prepared by a facile chemical route and was characterized by different structural and textural techniques. The composite Zr3O/AC was used as an efficient adsorbent for the MB and CV removal. Zr3O/AC presented irregular particle size with a regular channel array and micro-mesopores. Also, numerous functional organic groups on the surface were observed by FTIR spectroscopy before and after MB and CV adsorption. Functional groups on the surface along with the porous structure, channel array, and physical interactions (π-π and hydrogen) between the Zr3O/AC and the dyes are the main factors controlling the adsorption capacity. Langmuir and Freundlich models better described the adsorption process for MB and CV removal, respectively. The adsorption process followed the PSO kinetic model. Statistical response modeling together with the regression model helped us to optimize the experimental factors which provided a good fit with the experimental values.

Data Availability

Analysis data are available upon reasonable request to the corresponding author.

Conflicts of Interest

The authors declare that there no conflicts of interest regarding the publication of this paper.

Acknowledgments

The authors acknowledge the support of Ibn Zohr University and MASCIR Foundation for the support and help during SEM analyses.

Supplementary Materials

Supplementary material includes characterizations methods, response surface modeling procedure with a table reporting independent parameters, and response variables of the used matrix. Figures of linearized form of pseudo-second-order model, linearized form of Langmuir and Freundlich, and Pareto Chart were also reported in this section. (Supplementary Materials)