2.1. Materials

The details of chemicals used in present work are given in Table 2. All the chemicals were used without any treatment and purification. The pH of the solution was adjusted with 0.1 M NaOH. All the aqueous solutions were prepared using double distilled water.

2.2. Instruments

A digital pH meter (Spectral Lab Instrumental Pvt. Ltd., India) was use to measure pH. X-ray powder diffraction (XRD) analysis was performed by X-ray diffractometer (PAN analytical X’pert PRO) Using Cu X-ray tube (λ = 1.5406 Å). Morphologies of samples were observed with a high resolution field emission scanning electron microscope (HR-FESEM) from Zeiss, model name ULTRA Plus. It comes with a GeminiÒ column that proposes a theoretical resolution of 1.0 nm at 15 kV. Energy dispersive X-ray analysis (EDX) spectrometer was carried out using X Flash 6130 Bruker. Transmission Electron Microscopy (TEM) analysis of particles was performed in PHILIPS-CM 200, operated at 20-200 kV. High Resolution transmission electron microscopy (HR-TEM) of particles was carried out using JEOL JEM-2100.

2.3. Synthesis of calcium peroxide(CaO₂) nanoparticles

The CaO₂ nanoparticles were prepared by the existing method from Khodaveisi et al. with slight modification (^{Khodaveisi et al., 2011}). Briefly, 9 g CaCl₂ was first dissolved in 90 mL water, followed by the addition of 45 mL NH₃. H₂O (1 M) and 360 mL PEG-200 solution. The mixture was allowed to stir at 400 rpm, and 45 mL H₂O₂ was immediately added to it at the rate of three drops per minute for about 2 h. The stirring was continued, and pH of the mixture was adjusted to 11.5. The change in color of suspension from yellowish to white indicated the formation of CaO₂ nanoparticles. The suspension was centrifuged at 10,000 rpm for 5 min,and CaO₂ nanoparticles were collected. The particles were initially washed thrice using NaOH solution, and then twice with distilled water, and dried at 80°C for 120 min in vacuum oven. The dried CaO₂ nanoparticles were later used for adsorption experiments. A schematic outline of the synthesis of CaO₂ nanoparticles is shown in Figure 1.

Fig. 1 Experimental procedure for the synthesis of CaO₂ nanoparticles. 

2.4. Batch adsorption studies

Adsorption batch runs were performed for optimization process according to an orthogonal array L₉ (Table 3) and results obtained from each set as are (%E) given in Table 4. In each experimental run, 10 mL of aqueous benzeneacetic acid solution of known concentration and the known amount of nanoparticle of CaO₂ were taken in 100 mL Erlenmeyer flask. The flasks were agitated at a constant shaking rate at 22 ± 2°C in a water bath controlled shaker (REMI, RSB-12, and India). At the end of shaking, the samples were centrifuged and the supernatant used for determination of the benzeneacetic acid concentration.

Removal of benzeneacetic acid is given by Eq. (1) as follows:

where C₀ is the initial concentration of benzeneacetic acid (g/L) and C_e is the equilibrium concentration of benzeneacetic acid (g/L).

2.5. Taguchi methodology

The Taguchi method is a simple and vigorous method involves the different experimental conditions through orthogonal arrays to reduce experimental errors and process variation, enhance the efficiency, optimizing the process parameters and reproducibility of experiments. So, the method reduces work time and cost in the processes(^{Asiltürk & Neseli, 2012}). The important parameters affecting adsorption are initial concentration of the adsorbate (g/L), adsorbent dosage (W) and contact time (t) and each factor at three levels on the adsorption capacity and removal efficiency was studied. The used level setting values of the main factors (A-C). Taguchi’s L₉ orthogonal array matrix was used which incorporates three parameters and three levels.

In Taguchi methodology, the quality characteristics are employed into three different options: “larger is the better”, “nominal the-best”, and “smaller-the-better. The objective of this study was to remove benzeneacetic acid by CaO nanoparticles, the quality characteristic go for “larger is the better” of benzeneacetic acid removal defined by Eq. (2).

where the subscript LB denoted “larger is the better” and n was the number of repetitions under the same experimental conditions and y_i showed the measurement results. Each experiment was repeated twice (1st run and 2nd run) and the S/N ratio was determined using Minitab software (version 14). L and 9 mean Latin square and the number of experiments, and also, 3 and 3 indicate the numbers of factors and their levels, respectively.

3.1. CaO₂ nanoparticles characterization

X-ray diffraction patterns of CaO₂ are shown in Figure 2. The diffraction peaks at 2θ = 30.47°, 35.75°, 47.46°, 53.28°, 60.82° and 87.1° can be respectively indexed to (0 0 2), (1 1 0), (1 1 2), (1 0 3), (2 0 2) and (3 1 0) reflections of CaO₂ nanoparticles and match the reference patterns of standard file of CaO₂ (Joint Committee for Powder Diffraction Studies (JCPDS) File No. 03-0865). The morphology of prepared CaO₂ nanoparticles was studied by the HR-FESEM analysis (Fig. 3a and b). It can be clearly indicated that nanoparticles was looked like the aggregated round shape and are mostly spherical in shape. EDS showed the trace spectrum of CaO₂ nanoparticles as shown in Figure 4. The atomic compositions for calcium (Ca) and oxygen (O) were 42.88% and 57.12%, respectively. Figure 5a displays the TEM image of large number of CaO₂ nanoparticles with approximately uniform shape and size. It can be clearly seen is near spherical in shape with average particle size of about 10-40 nm. Figure 5b shows HR-TEM images of synthesized CaO₂ nanoparticles. The HR-TEM images clearly confirms, CaO₂ nanoparticles have a lattice structure with an interplanar spacing about 0.18 and 0.12 nm, which corresponds to (1 1 2) and (3 1 0) plane of CaO₂ respectively. These results are reliable with the XRD pattern.

Fig. 2 XRD pattern of synthesized CaO₂ nanoparticles. 

Fig. 3 (a, b) HR-FESEM images of CaO₂ nanoparticles. 

Fig. 4 EDX analysis of CaO₂ nanoparticles showed characteristics peaks. 

Fig. 5 (a) TEM and (b) HR-TEM images of CaO₂ nanoparticles. 

3.2. OA and the analysis of S/N ratio (Taguchi method)

According to the Taguchi L₉ (3³) OA, nine experiments were performed, and each experiment was repeated twice which were denoted by R₁ and R₂. The value of the response (removal efficiency) and S/N ratio are illustrated in Table 5. The mean of response and the mean of S/N ratio variable for each factor at a certain level can be determined. The boldfaces refer to the maximum value of the mean of response and the mean of the S/N ratios of a certain factor among four levels as shown in Table 5. The mean response plot for the removal of benzeneacetic acid using CaO₂ nanoparticles is shown in the Figure 6. The plot is used to show the relationship between the variables and output response. Adsorbent dosage ‘A’ and contact time ‘C’ increases with increases the removal efficiency of benzeneacetic acid. The effect of initial benzeneacetic acid concentration for removal is shown by factor ‘B’. Removal efficiency increases with decreases in benzeneacetic acid concentration. The optimum level of operational variables is determined from the maximum value. The mean of removal efficiency level ‘3’ is 86.18, level ‘1’ is 65.30, and level ‘2’ is 60.00 and its shows this level gives maximum efficiency. The S/N ratio plot as shown in Figure 7. According to these figures, increasing the adsorbent dose with low concentration of benzeneacetic acid concentration and increase the contact time the removal efficiency is increases and S/N ratio increases. The terms ‘signal’ to ‘noise’ ratio signifies the desirable and undesirable value for the output response, respectively. The S/N ratio of benzeneacetic acid removal was achieved. The process parameters were optimized based on S/N ratio. Lager the S/N ratio higher the benzeneacetic acid removal. The value of the S/N ratio, illustrate in Table 5. The mean of S/N ratios of level ‘3’ is 38.68, level ‘1’ is 35.74 and level ‘2’ is 35.17 and its shows this level gives higher efficiency. The optimum parameters were adsorbent dose (A) of 0.03 g, initial benzeneacetic acid concentration (B) of 6.8 g/L, and time (C) of 30 min. The optimum combination was found to be A3-B1C2 with corresponding mean of response value 86.18, 65.30, and 60.00and mean of S/N ratio value 38.68, 35.74, and 35.17 respectively.

Fig. 6 Mean removal efficiency plotted against different factor levels. 

Fig. 7 Mean SN ratios plotted against different factor levels. 

3.3. Statistical analysis of variance (ANOVA)

Analysis of variance (ANOVA) was performed to investigate which process parameters significantly affect the process responses and to determine the percent contribution of each operational variable to the response (^{Engin et al., 2008}; ^{Sadrzadeh & Mohammadi, 2008}). Table 6 shows the result of ANOVA test for mean response. ANOVA also used for estimate the error variance. The most significant factors that affect benzeneacetic acid removal was found to be adsorbent dose (93.78%) >initial concentration (4.72%) >contact time (0.16%). Percent contribution (P-ratio) is defined as a relation between the parameters’ sum of square to the total sum of square, which indicates the contributions of these parameters (^{Nik, Sadrzadeh, & Kaliaguine, 2012}):

3.4. Development of regression model

Regression analysis was developed to investigate the relationship between the variables (^{Deniz, 2013}). The mathematical model for benzeneacetic acid removal through the statistical analysis is shown as

Regression equation provides predicted data from different experimental condition and it’s compared with experimental data shown in Figure 8. Table 7 shows the regression analysis. Experimental and predicted values of the removal efficiency of benzeneacetic acid are almost equal. So as to the model tells the interaction of process parameters. ANOVA was derived to observe the null hypothesis for the regression shown in Table 8. A p value less than 0.05 was chosen which show the estimated model is considerable. The predicted and the experimental response (benzeneacetic acid removal) show a very good agreement with each other, and the coefficient determination (R²) of 0.91 implicates a good fitted of the model (Fig. 8).

Fig. 8 Comparison of predicted and experimental adsorption data for removal of benzeneacetic acid. 

3.5. Confirmation experiment

In Taguchi method, confirmation testis a necessary for the optimization study (^{Santra et al., 2014}). The confirmation test is performed with optimum levels and factors. From optimum conditions, to validate the predicted optimal conditions, new experiments were carried out and 94.49% benzeneacetic acid removal efficiency was achieved. Table 9 shows the predicted percent removal and the experimental percent removal (response) at the optimal conditions.

3.6. Adsorption Isotherm

Adsorption isotherm models such as ^{Langmuir isotherm (Langmuir, 1918}), Freundlich isotherm (^{Freundlich, 1906}), and Temkin isotherm (^{Tempkin & Pyzhev, 1940}) were tested to explain the equilibrium adsorption of benzeneacetic acid from aqueous solution (Table 10).

The initial benzeneacetic acid concentration (C₀) was varied from 4.8 to 13.47 g/L, at a constant contact time of 60 min, with 1 g/L CaO₂ nanoparticles at 25°C. The q_e values were calculated using

where q_e is the equilibrium adsorption capacity of benzeneacetic acid on the adsorbent (mg g−¹), C is the initial benzeneacetic acid concentration (g/L), C_e is the equilibrium concentration of benzeneacetic acid in solution (g/L), V is the volume of the solution (L), W is the mass of CaO₂ nanoparticles (in g).

Langmuir isotherm assumes that monolayer adsorption takes place over the homogeneous CaO₂nanoparticles adsorbent surface. The Freundlich isotherm considers the heterogeneous surface of an adsorbent and is used to describe the adsorption data. Temkin isotherm based on heat of adsorption of all the molecules in the layer would decrease linearly with coverage.

The best fit of data was evaluated from the correlation coefficient (R²), sum of the squares of errors (SSE), and Marquardt’s percent standard deviation (MPSD) values. The value of R² is higher and the value of SSE and MPSD shows the lower for the most favored situation. From the Table 10, it clearly shows that the Langmuir isotherm has high R² value and low SSE and MPSD as compared to Freundlich and Temkin isotherm. Thus, this study confirms that adsorption of benzeneacetic acid takes place on the CaO₂ nanoparticles, and follows the Langmuir model (^{Madan, Ravikumar, & Wasewar, 2016a}).