Investigate the ultrasound energy assisted adsorption mechanism of nickel(II) ions onto modified magnetic cobalt ferrite nanoparticles: Multivariate optimization

Accepted Manuscript Investigate the ultrasound energy assisted adsorption mechanism of nickel(II) ions onto modified magnetic cobalt ferrite nanoparticles: Multivariate optimization Fatemeh Mehrabi, Ebrahim Alipanahpour Dil PII: DOI: Reference:

S1350-4177(16)30477-1 http://dx.doi.org/10.1016/j.ultsonch.2016.12.038 ULTSON 3493

To appear in:

Ultrasonics Sonochemistry

Received Date: Revised Date: Accepted Date:

14 October 2016 28 December 2016 29 December 2016

Please cite this article as: F. Mehrabi, E. Alipanahpour Dil, Investigate the ultrasound energy assisted adsorption mechanism of nickel(II) ions onto modified magnetic cobalt ferrite nanoparticles: Multivariate optimization, Ultrasonics Sonochemistry (2016), doi: http://dx.doi.org/10.1016/j.ultsonch.2016.12.038

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Investigate the ultrasound energy assisted adsorption mechanism of nickel(II) ions onto modified magnetic cobalt ferrite nanoparticles: Multivariate optimization Fatemeh Mehrabi *, Ebrahim Alipanahpour Dil Young Researchers and Elite Club, Gachsaran Branch, Islamic Azad University, Gachsaran, Iran

Graphical abstract

*

Corresponding author: E-mail address: [email protected]

1

Abstract In present study, magnetic cobalt ferrite nanoparticles modified with (E)-N-(2-nitrobenzylidene)2-(2-(2- nitrophenyl)imidazolidine-1-yl) ethaneamine (CoFe2O4-NPs-NBNPIEA) was synthesized and applied as novel adsorbent for ultrasound energy assisted adsorption of nickel(II) ions (Ni2+) from aqueous solution. The prepared adsorbent characterized by Fourier transforms infrared spectroscopy (FT-IR), transmission electron microscope (TEM), vibrating sample magnetometer (VSM) and X-ray diffraction (XRD). The dependency of adsorption percentage to variables such as pH, initial Ni2+ ions concentration, adsorbent mass and ultrasound time were studied with response surface methodology (RSM) by considering the desirable functions. The quadratic model between the dependent and independent variables was built. The proposed method showed good agreement between the experimental data and predictive value, and it has been successfully employed to adsorption of Ni2+ ions from aqueous solution. Subsequently, the experimental equilibrium data at different concentration of Ni2+ ions and 10 mg amount of adsorbent mass was fitted to conventional isotherm models like Langmuir, Freundlich, Tempkin, Dubinin-Radushkevich and it was revealed that the Langmuir is best model for explanation of behavior of experimental data. In addition, conventional kinetic models such as pseudo-first and second-order, Elovich and intraparticle diffusion were applied and it was seen that pseudo-second-order equation is suitable to fit the experimental data. Keyword: Multivariate optimization; Ultrasound energy; Adsorption; Ni2+ ions; Magnetic Cobalt ferrite nanoparticles; Adsorption mechanism.

1. Introduction An increasing demand for nickel in different industrial applications like mineral processing, electroplating, smelting, battery production, etc., leads to more and more discharge of nickel in wastewater, which is harmful to human health and environment. High-level exposure to nickel for human is linked to dermatitis, lung fibrosis, cardiovascular and kidney diseases, reduced lung function and also causes lung cancer [1]. As nickel is a non-biodegradable heavy metal with a high toxicity present in wastewater, it is of great need to develop effective and environmentally friendly methods to remove the nickel from the aqueous solution. Different conventional 2

physicochemical treatment methods including chemical precipitation, ion exchange, membrane filtration, and electrochemical have been widely used in the last decades for nickel removal. However, these processes have significant disadvantages such as incomplete removal, high cost and second-pollution problem, which hinder their promotion and development in the treatment of nickel-contaminated wastewater. Recently, numerous research have been carried out to explore the potential of cheaper and more efficient technologies for nickel removal in order to reduce the second-pollution produced and improve the removal efficiency of treatment [2-6]. Among methods for wastewater treatment, the adsorption methods are widely used to remove classes of chemical pollutants from wastewater. These wide application concern due to high efficiency, capacity of adsorbent usable for large scale application, while the adsorbents are regenerable, safe, non-toxic and/or lower toxicity [7]. The adsorption based on application of various metallic and semimetallic nanoparticles are designed to clean up aqueous contaminated water in short time [8]. Due to the unique properties of nanoparticles such as a high number of reactive atoms, high mechanical and thermal strength, large number of vacant reactive surface sites metallic or semi-metallic behavior and high surface area widely applied for adsorption of various toxic materials [9]. Nanoparticles as sorbents for separation of target compounds attain increasing interest owing to their effectiveness. Nanotechnology as powerful platform substantially enhances environmental quality and sustainability through pollution prevention, treatment, and remediation. In particular, application of nanoparticle technology leads to solve environmental problems and have become more important owing to its special physical and chemical properties. The size, surface structure and interparticle interaction of nanomaterial determine their unique properties and improve their performances for application in many areas [10-13]. Cobalt ferrite nanoparticles, a well-known hard magnetic spinel ferrite, show unique properties such as strong magneto-crystalline anisotropy, moderate saturation magnetization, high coercivity, good physical and chemical stability, and large magneto-strictive coefficient at a room temperature [14]. Chemical modification of nanomaterial with different chelating agents or complexation centers significantly changes its nature, increases the number of reactive functional groups as well as enhances its selectivity and sorption capacity [15]. The adsorption capacity of adsorbent depends strongly on various factors such as initial concentration, pH value of solution, contact time and adsorbent mass. Conventional methods for investigating the adsorption process are usually conducted by changing one independent variable 3

and keeping the other factors constant. However, in this case, the obtained results cannot reveal the combined effect of all the factors involved simultaneously. Additionally, such methods are often time consuming and require a number of experiments which are unreliable to determine optimum levels. Therefore, response surface methodology (RSM), based on the statistical experimental design, has been widely used to eliminate the limitations of single factor experiments by optimizing all the influencing factors collectively [16, 17]. The multivariate optimization has several advantages over the conventional “one-at-a-time”method because of providing more information with running fewer experiments. During multivariate optimization a response surface methodology (RSM) leading to an empirical mathematical model relates the adsorption percent of sorbate with effective variables and their interactions [18, 19]. Sonication causes more acceleration of adsorption percentage over short time using small amount of adsorbent. Ultrasound via secondary activity like cavitation (nucleation, growth and transient collapse of tiny gas bubbles) improves the mass transfer through convection pathway that is emerged from physical phenomena such as microstreaming, micro-turbulence, acoustic (or shock) waves and micro-jets without significant change in equilibrium characteristics of the adsorption/desorption system [20, 21]. The aim of this research: (1) synthesis of magnetic CoFe2O4 nanoparticles modified by (E)-N-(2nitrobenzylidene)-2-(2-(2- nitrophenyl)imidazolidine-1-yl) ethaneamine (NBNPIEA) and characterization by FT-IR, TEM, VSM and XRD; (2) to apply a central composite design combined with RSM and optimization modeling the adsorption of Ni2+ ions from aqueous solution; (3) determine the effects of four independent variables such as: initial concentration of Ni2+ ions, initial pH of the solution, adsorbent mass, ultrasound time and their interactions on the adsorption; and (4) equilibrium adsorption isotherms and kinetics models were analyzed.

2. Experimental 2.1.Instruments and reagents All chemicals such as Ni(NO3)2·6H2O, Fe(NO3)3.9H2O, Co(NO3)2.6H2O, ethanol, NaOH and HCl were purchased from Merck, Darmstadt, Germany. A stock solution (100 mg L-1) of Ni2+ ion was prepared and subsequently, different working solutions of required concentrations are prepared by proper dilution. The analysis of Ni2+ ion concentration was carried out by using flame atomic absorption spectrophotometer (FAAS) Varian model AA 240 (USA). The solution 4

pH measurements were carried out using pH/Ion meter model- 686 (Metrohm, Switzerland, Swiss). An ultrasonic bath with heating system (Tecno-GAZ SPA Ultra Sonic System, Italy) at 40 KHz of frequency and 130 W of power was used for the ultrasound-assisted adsorption procedure. A HERMLE bench centrifuge (2206 A, Germany) was used to accelerate the phase separation. A Fourier transform infrared (FT-IR) spectrum was recorded using a Perkin ElmerSpectrum RX-IFTIR spectrometer in the range of 400-4000 cm-1. X-ray diffraction (XRD, Philips PW 1800) was performed to characterize the phase and structure of the prepared nanoparticles using Cukα radiation (40 kV and 40 mA) at angles ranging from 10 to 80º. The morphology and the average particle size of CoFe2 O4 nanoparticles were further investigated by a transmission electron microscope (TEM, Hitachi H-800). A vibrating sample magnetometer (VSM) was used to measure magnetization versus applied magnetic field. The STATISTICA 10.0 and Design-Expert 7.0 software (Stat Soft Inc., Tulsa, USA) was used for experimental design analysis and their subsequent regression analysis.

2.2.Synthesis of CoFe2O4 nanoparticles (CoFe2O4-NPs) The reaction solution for synthesis of CoFe2O4 nanoparticles was prepared as follows: Fe(NO3)3.9H2O and Co(NO3)2.6H2O (molar ratio of Fe3+/Co2+=2) were dissolved in 100 mL of distilled water. Then 20 mL of NaOH solution (1 mol L-1) was slowly added to the solution for 30.0 min. The solution was heated at 80 °C for 2 h. The precipitate was then centrifuged and rinsed with distilled water, followed by being left in an atmosphere environment to dry.

2.3.Synthesis of (E)-N-(2-nitrobenzylidene)-2-(2-(2- nitrophenyl)imidazolidine-1-yl) ethaneamine (NBNPIEA) A solution of 2-nitrobenzaldehyde (0.302 g, 2 mmol) in methanol (20 mL) was gradually added to methanolic solution of diethylenetriamine (0.103 g, 1 mmol) (10 mL) at room temperature and then the reaction mixture was vigorously stirred. After 3 h, the ligand as a cream product was obtained in 85% yield [22].

2.4.Preparation of CoFe2O4-NPs-NBNPIEA 1 g of CoFe2O4-NPs was washed twice with distilled water and mixed with 0.1 g of NBNPIEA dissolved in ethanol. Stirring the above mention mixture for 4 h at 400 rpm leads to deposition of 5

ligand on CoFe2O4-NPs. It is noticeable that the proposed ligand due to presence of large amount of non-localized π electrons and ability to form hydrogen bands with CoFe2O4-NPs functional group at the investigated value completely loaded on adsorbent.

2.5.Ultrasound energy-assisted adsorption method The study of ultrasound energy assisted adsorption of Ni2+ ion onto the CoFe2O4-NPs-NBNPIEA is conducted as follow: 50 mL of 19 mg L-1 Ni2+ ion solution at pH 5.7 was mixed completely with 18 mg of CoFe2O4-NPs-NBNPIEA and its subsequent exposure under ultrasound energy at the frequency of 40 kHz (3.5 min) led to best dispersion of adsorbent into the solution which raising temperature simultaneously led to enhancement coefficient and progress in the mass transfer. At the end of ultrasound-assisted adsorption experiments, the supernatant and the adsorbent were separated by means of a magnet and the amount of non-adsorbed Ni2+ ions was analyzed by flame atomic absorption spectrometry. After the determination of concentration of Ni2+ ions versus time, its adsorption percentage was calculated using the following equation [23]:

A d so rp tio n ( % ) =

C -C C 0

t

(1)

×100

0

where C0 (mg L-1) and Ct (mg L-1) are the ion concentration at initial and after time t respectively. The data were used to calculate the adsorption capacity (qe (mg g-1)) of the adsorbent. The concentration of Ni2+ ions on the adsorbent surface at equilibrium was calculated by [24]:

( - )V q= C C m 0

e

(2)

e

where C0 and Ce (mg L-1) are initial and equilibrium Ni2+ ions concentrations in liquid phase, respectively, V (L) is the total volume of Ni2+ ions solution and m (g) is the mass of adsorbent.

2.6.Design of experiments or response surface methodology 6

Experiments were designed via response surface methodology (RSM). RSM is a collection of statistical and mathematical tools used to optimize the response governed by several independent variables. It has been useful for modeling and analysis of problems in which a response of interest is influenced by several variables and its objective is to optimize this response. Classically, response is optimized by varying one parameter one at a time and keeping other parameters constant. Classical method is time consuming and does not provide the correct picture of quantitative interactions between various parameters. To overcome these drawbacks, RSM is employed to know interaction among various parameters. By taking into account various independent parameters, optimization is achieved through RSM. Tow designs of RSM are: Box– Behnken design (BBD) and Central composite design (CCD). In BBD, cubic points are taken into consideration, whereas in CCD, axial points, in addition to cubic points are taken into consideration. It means that BBD has only 3 degrees of freedom (-1, 0, +1), whereas CCD system has five degrees of freedom (-α, -1, 0, +1, +α). In present studies, the four parameters studied were initial concentration of Ni2+ ions (A), initial pH of the solution (B), adsorbent mass (C) and ultrasound time (D) at five levels (-α, -1, 0, 1, +α) with a significance level of 0.05. Here, each parameter is coded by -α, -1, 0, 1, and +α. -1, 0 and +1 are the minimum, central and maximum coded cubic values respectively. Likewise +α and -α represent minimum and maximum coded axial values used in the model, respectively. The value of α depends on the number of process independent variables taken in the design [25-27]. The value of α for orthogonal design is calculated as follows: α= 2 (

number of independent variables 4

)

(3)

Here number of variables was 4. So,

4 α= 2 ( 4 ) = 2

(4)

The total number of experiments obtained by operating CCD of RSM in a STATISTICA 10.0 software was 30. In the experiments, there were 16 factorial points, 8 axial points and 6 replicates.

7

k

N = 2 ×2k +n

4

0

(5)

= 2 × 2(4) + 6 = 30

where N is the total number of experiments, k is the number of factors and n0 is the number of central runs. On the basis of the results, a second order polynomial is applied to explain the relationship between response and processed variables as follows [28]:

4

4

4

4

y = β 0 + ∑ βi xi +∑∑ βij x i x j + ∑ βii x i2 + ε i=1

i=1 j=1

(6)

i=1

Y denotes the predicted response, and i and j take values from 1 to the number of independent process variables. β0, βi, βii, and βij are the offset terms, linear, square and interaction effects predicted by the method of least squares, ε denotes the error of prediction and xi and xj are coded independent process variables.

3. Result and discussion 3.1.Characterization of adsorbent The formation of the spinel structure of cobalt ferrite in the nanocrystalline form and its cation distribution is supported by FT-IR analysis. Fig. 1(a) shows the room temperature FT-IR spectra of CoFe2O4 nanoparticles. The spectra indicate the presence of absorption bands in the range of 400-600 cm-1 which is a common feature of the spinel ferrite. The higher frequency absorption band lies around 570 cm-1 and is assigned to the intrinsic vibration of the tetrahedral metal complex which consists of a bond between the oxygen ion, the tetrahedral site metal ion, and the lower frequency absorption band lies around 490 cm-1 and is assigned to the intrinsic vibration of the octahedral metal complex which consists of a bond between the oxygen ion and the octahedral site metal ion. The appearance of higher frequency band is due to the stretching vibration, while the lower frequency band is due to the bending vibration. The splitting of the octahedral (‘B’ site) absorption band near 570 cm-1 is due to the presence of different kinds of cations, including Co2+ and Fe3+ on the ‘B’ site. The bands observed around 3460 and 1670 cm-1 frequencies are ascribed due to the stretching modes and H–O–H bending vibration of the free or 8

absorbed water molecules [29, 30]. The FT-IR of CoFe2O4 nanoparticles strongly supports loading and impregnation of ligand (NBNPIEA) on its surface shown in fig.1(b). Therefore, this novel sorbent due to presence of various reactive sites such as CoFe2O4 nanoparticles and ligand non-localized π electron, OH and other reactive centers are good choice for complexation of understudy metal ions throughout various mechanism such as ion-dipole interaction and mainly covalent binding. According to Fig. 1, bonds in spectra (b) in comparison with (a), the emergence of new peaks and increasing of peak intensity shown modification of CoFe2O4 nanoparticles surface by NBNPIEA.

Insert Fig. 1.

XRD pattern of CoFe2O4 nanoparticles is shown in Fig. 2. The XRD pattern of prepared CoFe2O4 nanoparticles is indexed as a pure cubic phase which is very close to the literature values (Joint Committee for Powder Diffraction Standards, JCPDS No. 01-1121). The average size of nanoparticles (D) was estimated by using the Scherrer equation, Dc = Kλ/β cosθ, where K usually takes a value of about 0.9, β is the width of the observed diffraction peak at its half maximum intensity (FWHM) and λ is the X-ray wavelength (CuKα radiation, equals to 0.154 nm).

Insert Fig. 2.

The study of variation of magnetization with applied magnetic field gives advantageous information about magnetic properties of the magnetic particles. The magnetic properties of synthesized particles are studied using a vibrating sample magnetometer (VSM) at room temperature. As expected, the curves show no obvious remanence or coercivity at room temperature, indicating that they possess the superparamagnetic character. As shown in Fig. S1, the saturation magnetization (magnetization) value of CoFe2O4-NPs found to be 42.5 emu/g. After modified with NBNPIEA, the magnetization value of CoFe2O4-NPs-NBNPIEA decrease to 38.3 emu/g. Nevertheless, the magnetic properties of all the particles are strong enough to ensure the modified magnetic cobalt ferrite nanoparticles to be easily separated by an external magnetic field within 30 s, which is advantageous to the application. 9

TEM image of CoFe2O4 nanoparticles is shown in Fig. S2(a). It can be seen that the mean diameter of CoFe2O4 nanoparticles is near 40 nm. The particle size with a narrow distribution is given in the inset of Fig. S2(b).The average particles size is 45 nm, which is in good agreement with the particle sizes estimated by Scherer’s formula. Also, TEM image was used to determine the morphology and the particle size of the prepared CoFe2O4-NPs modified by NBNPIEA (Fig. S2(c)) that suggest essentially non-spherical shaped particles with the approximately size of 60 nm. The particles are almost uniform in the shape and the size. The particle size frequency of the sample was measured by using particle size analysis technique and shown in Fig. S2(d). According to the obtained results the average size of particles was 65 nm. TEM micrographs and histograms (fig. S2(a-d)) confirm that crystal size increased with modified of nanoparticles by NBNPIEA.

3.2.Optimization studies by statistical experimental design The central composite design of RSM is employed in this experiment to obtain a quadratic model, consisting of 30 trials. The range and levels of the four independent variables, viz., initial concentration of Ni2+ ions (10-30 mg L-1), pH (2-10), adsorbent mass (5-25 mg) and ultrasound time (1-5 min), were chosen as shown in Table S1. By applying analysis of variance (ANOVA) to the regression model, the predicted response, Y was obtained as:

Y= +21.73951 +0.16636A +3.54461B +2.87459C +24.24371D +0.07279AB -0.02780AC +0.51125AD -0.02557BC -0.55149BD -0.17050CD -0.05341A2 -0.28700B2 -0.04161C2 3.93025D2

(7)

where Y is the predicted adsorption percentage of Ni2+ ions. A, B, C and D are the corresponding coded variables of initial concentration of Ni2+ ions, pH, adsorbent mass and ultrasound time, respectively. The significance of the model was studied by analysis of variance (ANOVA). The significance of the model equation for the adsorption percentage of Ni2+ ion was checked by the F test and is shown in Table 1. The model F-value for the adsorption percentage of Ni2+ ions was 528.1. The model probability value (<0.0001) of Ni2+ ion was also low enough to show that the model obtained was significant. Lack-of-fit test, in which an insignificant lack-of-fit is desired, was also 10

used to evaluate the model adequacy. The lack-of-fit value for the adsorption percentage of Ni2+ ions was 0.1785; the lack-of-fit value for the adsorption percentage of Ni2+ ions was statistically insignificant and showed that the constructed model was suitable to describe the observed data. The goodness-of-fit of the model was also checked by the determination coefficient R2, and the R2 value of the adsorption percentage of Ni2+ ions was 0.998, indicating that the model fitted well with the observed data. Regarding the model of the adsorption percentage of Ni2+ ions, the values of the adjusted R2 value was 0.996, and that of predicted R2 value was 0.990. The adjusted and predicted R2 values were in good agreement, and the adjusted R2 value was close to predicted R2 value, indicating a good adjustment between the observed and predicted values. Fig. S3(a) shows that the points of the predicted versus actual plot for the adsorption percentage of Ni2+ ions were clustered along the diagonal, indicating that the predicted data match well with the observed data.

Insert Table 1.

To examine the significance of the main and interaction effects of the parameters on the adsorption percentage of Ni2+ ions, the data of ANOVA, as shown in Table 1 were used. The statistical analysis showed that pH, concentration of Ni2+ ions, adsorbent mass and ultrasound time significantly influenced on the adsorption percentage of Ni2+ ions, since the value of Prob > F (P-value) less than 0.0001 indicates the significance of the model term. And the pure error was very low, indicating good reproducibility of the data obtained. The coefficient of variation (CV) indicates the degree of the precision [31]. A low value of CV (0.48%) here clearly indicated a very high degree of precision and a good reliability of the experimental values. Adequate precision measures the signal to noise ratio and a ratio greater than 4 is desirable. Therefore, in the quadratic model of the adsorption percentage of Ni2+ ions, the ratio of 90.2 indicate an adequate signal for the model to be used to navigate the design space. Fig. S3(b) is a plot of normal probability of the residuals for the adsorption percentage of Ni2+ ions. As can be seen, most of the obtained data points were consistently distributed on a straight trend line and, except two points; none of the individual residuals exceeded the residual variance, which could indicate very good adequacy of the regression model that was utilized for the adsorption percentage of Ni2+ ions. 11

Fig. S3(c) shows the diagnostic plot of internally studentized residual versus run for the adsorption percentage of Ni2+ ions. Outliers in the diagnostic plot simply indicate the magnitude of the residuals and justify if any data had particularly large residuals. As shown in the diagnostic plot, the red line was produced by the software based on the internally studentized residual to define outliers. Absence of any outlier in the plot confirm that the model is consistent with all data. Furthermore, there is no significant distribution pattern for the diagnostic graph and the internally studentized residual was randomly scattered across the graph. Thus there is no violation of the independency or constant variance assumption for all runs. The analysis of the results is visualized using standardized main effect, quadratic and interaction according to Pareto chart (P = 95%) as shown in Fig. 3, which are based on the significant parameter must have value higher than ±t. The experimental results confirm positive contribution of ultrasound time (L) and adsorbent mass (L) on the response and their value raising led to significant enhancement on response, and the initial concentration (L) and pH (L) were the most relevant parameter with negative effects for the adsorption of Ni2+ ions. It is important to infer that the absorption percentage of Ni2+ ions increased with an increase in ultrasound time and/or adsorbent mass in the analyzed samples. Also, with the increases in the initial pH and/or Ni2+ ions concentration decreases the adsorption. These results confirm that sonication following lateral formation of cavitation (nucleation, growth and transient collapse of tiny gas bubbles) led to occurrence of two phenomena like increasing high contact area and higher concentration gradient and raising the diffusion coefficient which all these process led to improvement the rate and magnitude of mass transfer via diffusion and migration.

Insert Fig. 3.

3.3.3D response surface plots In the next step of the design, response surface methodology (RSM) was developed by considering all the significant interactions in the CCD to optimize the critical factors and describe the nature of the response surface in the experiment. Fig. 4 (a, b) represents the most relevant fitted response surfaces for the design and depicts the response surface plots of the adsorption percentage of Ni2+ ions versus significant variables. These plots were obtained for a given pair of factors at fixed and optimal values of other variables. The curvatures of these plots 12

indicate the interaction between the variables. Fig. 4(a) shows the effects of interaction between pH and initial Ni2+ ions concentration on the adsorption percentage of Ni2+ ions. When the pH is fixed, the adsorption percentage of Ni2+ ions increased with the increase of initial Ni2+ ions concentration until reaching a maximum (19 mg L-1) and then decreases. The pH of the system exerts profound influence on the ability of adsorbent surface for interaction and Ni2+ ions tendency for binding to solid surface. These phenomena presumably are due to its influence on the surface properties of the adsorbent. The effect of pH on adsorption can be described on the basis of point zero charge (pHPZC), which is the point at which the net charge of the adsorbent is zero. The pHZPC value of CoFe2O4-NPs-NBNPIEA was found to be 3.2 (Fig. S4). The surface of CoFe2O4-NPs-NBNPIEA is neutral when pH of the aqueous solution is equal to pHPZC. At pH < 3.2, sorbent has positive charged that may emerged from accumulation of H3O+ ions into adsorbent functional groups. Thus a repulsion force occurs between the Ni2+ ions and the sorbent surface which causes a decrease in the adsorption percentage of Ni2+ ion. At higher pH value (pH > pHPZC) the number of negatively charged sites enhances and lead to appearance of significant strong electrostatic attraction between adsorbent and Ni2+ ions. Therefore, the adsorption of Ni2+ ion was increased (until pH=5.7) due to increased electrostatic force of attraction and then the adsorption percentage significantly decreased at higher pH values (pH > 5.7). The respective results (Fig. 4(a)) shows that the initial pH significantly affects the extent of adsorption of Ni2+ ions over the CoFe2O4-NPs-NBNPIEA. This result indicated that pH and initial Ni2+ ions concentration were important variables for adsorption percentage of Ni2+ ions. Fig. 4(b) shows the effect of interaction between ultrasound time and adsorbent mass on the adsorption percentage of Ni2+ ions. Accordingly, the adsorption percentage increased at higher value of adsorbent mass due to its high specific surface area. Increase in surface area and availability of more active adsorption sites at higher amount of adsorbent is associated with increase in adsorption rate. Also, the adsorption rate in 1-3.5 min of ultrasound time is very rapid due to the highly available adsorbent surface area and vacant sites achieved and enhanced by the dispersion of adsorbent into solution via ultrasound power. Then, at vary larger ultrasound time maybe probably due to desorption of ions on the adsorbent surface a decrease in adsorption percentage was observed. In the other words, probable mechanism for decreasing adsorption percentage maybe related to the back distribution of adsorb species to the bulk of sample.

13

Insert Fig. 4.

3.4.Optimization of adsorption process by CCD The profile for forecasted values and desirability option in the STATISTICA 10.0 software is utilized to optimize the factor impacts (Fig. 5). Profiling the desirability of responses contains characterizing the desirability function for the adsorption percentage of Ni2+ ions by allocating predicted values. Desirability is an objective function that ranges from zero outside of the limits to one at the goal. The goal seeking begins at a random starting point and proceeds up the steepest slope to maximum. There may be two or more maximums because of curvature in the response surfaces and their combination into the desirability function. The importance of each goal was changed in relation to the other goals. The desirability values of the minimum and maximum were configured as 0 and 1, respectively [32]. The maximum desirability function obtained was taken as the optimum operating condition. According to the overall results of optimization study, the following experimental conditions are chosen: 19 mg L-1 Ni2+ ions of concentration, pH 5.7, 18 mg of adsorbent mass, and ultrasound time of 3.5 min. The observed experimental of Ni2+ ions adsorption under the above conditions (N=5) was 99.3±2.41 whilst the predictive value was 99.47%. conduction of t-test at 95% confidence interval revel absence of significant difference between the experimental and theoretical predicted response, which also confirm high efficiency of the model for best explanation and representation of data.

Insert Fig. 5.

3.5.Adsorption isotherms The equilibrium adsorption isotherm studies are required to give useful information about mechanism, properties, and tendency of adsorbent toward each ion. Various isotherm models (Langmuir, Freundlich, Tempkin, and Dubinin–Radushkevich (D-R)) have been used to discuss the equilibrium characteristics of the adsorption process [33-36]. The constant parameters of the isotherm equations for this adsorption process and the determination coefficient (R2) for conventional isotherms based on known equation and requirement are summarized in Table 2. Based on the linear form of Langmuir isotherm model (according to Table 2), the values of KL (the Langmuir adsorption constant (L mg-1)) and Qm (theoretical maximum adsorption capacity 14

(mg g-1)) were obtained from the intercept and slope of the plot of Ce/qe versus Ce (figure not shown), respectively. The Langmuir isotherm assumes a monolayer adsorption onto a solid surface with a definite number of identical sites in addition to that there is no interaction between the adsorbate molecules. The fitness of experimental data was evaluated at 10 mg of adsorbent mass. The high determination coefficient (0.998) over the whole concentration range strongly support the fact that the adsorption data closely follow the Langmuir model. The Langmuir monolayer capacity Qm is being 151.5 mg of Ni2+ ions per gram of adsorbent that exhibit relevant adsorption properties. The parameters of Freundlich isotherm model such as KF ((mg g1

)/(mg L-1)1/n) and 1/n were calculated from the intercept and slope of the linear plot of ln qe

versus ln Ce (figure not shown), respectively (Table 2). KF is the Freundlich constant, which indicates the extent of adsorption, and 1/n is the heterogeneity factor (adsorption effectiveness). The 1/n value ranges between 0 and 1. When the value of 1/n is equal to unity, the adsorption is linear; when the value of 1/n is below the unity, the adsorption process is chemically driven; and when the value of 1/n is above the unity, the adsorption is a physically driven process. The value of 1/n (0.539) gives an indication of the favorability of adsorption and high tendency of Ni2+ ions for the adsorption onto CoFe2O4-NPs-NBNPIEA , while lower R2 value (0.941) show unsuitability of Freundlich model for fitting the experimental data. The heat of the adsorption and the adsorbent–adsorbate interaction were evaluated using Tempkin isotherm model. In this model, B1 is the Tempkin constant related to heat of the adsorption (J mol-1), T is the absolute temperature (K), R is the universal gas constant (8.314 J mol-1 K-1) and KT is the equilibrium binding constant (L mg-1) [37]. The values of the Tempkin constant (14.84) and the determination coefficient (0.925) are lower than the Freundlich and Langmuir values. Therefore, the Tempkin isotherm represents a worse fit of experimental data than both other isotherms. D-R model was applied to estimate the porosity apparent free energy and the characteristic of adsorption. In this model, B (mol2 (kJ2)-1) is a constant related to the adsorption energy, Qs (mg g-1) is the theoretical saturation capacity and ε is the Polanyi potential. The slope of the plot of ln qe versus ε2 gives B and the intercept yields the Qs value. In this case, the D-R equation represented poorer fit to the experimental data than Langmuir isotherm equation. The fitting results show that the determination coefficient of Langmuir isotherm was higher than that of other isotherms, indicating that the adsorption of Ni2+ ions onto CoFe2O4-NPs-NBNPIEA can be better fitted using Langmuir model and the adsorption is a monolayer adsorption. 15

Insert Table 2.

3.6.Adsorption kinetics The nature of the adsorption process will depend on physical or chemical characteristics of the adsorbent system and the system conditions. Such kinetic models (pseudo-first and second-order, Elovich and intraparticle diffusion) were investigated to study the rate and mechanism of adsorption process [38-41]. Table 3 summarized the properties of each model at initial Ni2+ ions concentration (19 mg L-1) and 10 mg of adsorbent mass. In the pseudo-first-order model (Lagergren model), by plotting the values of log (qe - qt) versus t (figure not shown), the value of k1 and qe can be determined from the slope and intercept of the obtained line, respectively. Distance of intercept from experimental qe value, indicate that this model not explain experimental data and rate of adsorption do not follow to this equation. The validity of the model is checked by the determination coefficient (R2). The calculated equilibrium adsorption capacities appeared that the pseudo-first-order model was not fit well with the experimental data. Hence, the sorption kinetic data were described with the pseudosecond-order kinetic equation. Despite pseudo-first-order model, the plot of t/qt versus t for the pseudo-second-order kinetic model gives a straight line with a high determination coefficient that k2 and equilibrium adsorption capacity (qe) were calculated from the intercept and slope of this line, respectively. The high value of R2 (0.981) and closeness of experimental and theoretical adsorption capacity (qe) value show the applicability of the pseudo-second-order model to explain and interpret the experimental data (Table 3). The Elovich equation as another rate equation based on the adsorption capacity in linear form has been successfully applied for the adsorption of solutes from a liquid solution. The plot of qt versus ln (t) (figure not shown) should yield a linear relationship with a slope of (1/ β) and an intercept of (1/β) ln (αβ) if the Elovich is applicable. The Elovich constants obtained from the slope and the intercept of the straight line are reported in Table 3. The determination coefficient is 0.863 for the adsorption of Ni2+ ions, showing the unsuitability of this model for complete evaluation of the adsorption process. The later process possibility is explored using the intraparticle diffusion model based on diffusive mass transfer that adsorption rate expressed in terms of the square root of time (t). The values of Kdiff (the intraparticle diffusion rate constant (mg g-1 min-1/2)) and C (thickness of the boundary 16

layer) were calculated from the slope and intercept of the plot of qt versus t1/2 (Table 3). Since, the intraparticle curve did not pass through the origin; one can notice that this model is poorer than another model (pseudo-second-order model) to control the kinetic of adsorption process. From Table 3 it is observed that the adsorption of Ni2+ ions follows more closely to pseudosecond-order kinetics since in this case, the theoretical equilibrium value is close to experimental value with determination coefficient of 0.981 which fit the experimental data better than the other kinetic models for the entire adsorption process. Therefore, it can be concluded that pseudo-second-order equation is better in describing the adsorption kinetics of Ni2+ ions onto CoFe2O4-NPs-NBNPIEA. Insert Table 3.

3.7.Comparative adsorption study The maximum adsorption capacity Qmax (mg g-1) is great criterion for judgment useful in scaleup considerations. Some studies have been conducted using various types of adsorbents for Ni2+ ions adsorption. The maximum monolayer adsorption capacity and contact time of CoFe2O4NPs-NBNPIEA for the adsorption of Ni2+ ion was compared with other adsorbents reported in the literature and the values are shown in Table 4. From Table 4, CoFe2O4-NPs-NBNPIEA showed the highest adsorptive and lowest contact time for adsorption of Ni2+ ions. It may be seen from Table 4 that the contact time for proposed method in comparison with all of the adsorbents are preferable and superior and shows satisfactory adsorption performance for Ni2+ ions. It can be seen from Table 4 that the adsorbent show a comparable sorption capacity with the respect to other sorbents, revealing that the CoFe2O4-NPs- NBNPIEA is suitable for adsorption of Ni2+ ions from aqueous solutions since it has a relatively high adsorption capacity. Also, the adsorption capacity of CoFe2O4-NPs and CoFe2O4-NPs modified by NBNPIEA are shown in Table 4. It may be observed that the adsorption capacity of CoFe2O4-NPs-NBNPIEA is sufficiently higher than unmodified CoFe2O4-NPs.

Insert Table 4.

4. Conclusion 17

In this study a new simple, low cost and environmentally friendship adsorption technique based on magnetic CoFe2O4 nanoparticles modified with (E)-N-(2-nitrobenzylidene)-2-(2-(2nitrophenyl)imidazolidine-1-yl) ethaneamine (CoFe2O4-NPs-NBNPIEA) has been reported for adsorption of Ni2+ ion from aqueous solution. The effects of various parameters such as initial concentration of Ni2+ ions, pH, adsorbent mass and ultrasound time on the adsorption percentage of Ni2+ ions were investigated and optimized by central composite design (CCD) and desirability function (DF). A good agreement between experimental and predicted values was observed. Maximum adsorption percentage of Ni2+ ions , which was found to be 19 mg L-1 of initial Ni2+ ions concentration, pH 5.7, 18 mg of adsorbent mass and 3.5 min ultrasound time at room temperature. The experimental data were analyzed by the Langmuir, Freundlich, Tempkin, and Dubinin-Radushkevich isotherm models. The equilibrium data fitted very well in Langmuir isotherm equation. The distinguished advantage of present research is its applicability for adsorption of large amount of Ni2+ ions (151.5 mg g-1) following using small amount of adsorbent (10 mg) in reasonable time (3.5 min). This adsorbent simply synthesized in our laboratory, while it is a safe, green and non-toxic material and its back diffusion to aqueous solution is low and reasonable. In addition, the kinetic models to describe experimental data show successful fit of to the pseudo-second-order kinetic model.

Acknowledgment The author would like to thank the Young Researchers and Elite Club, Gachsaran Branch, Islamic Azad University for the financial support of this work.

Reference [1] S. Malamis, E. Katsou, A review on zinc and nickel adsorption on natural and modified zeolite, bentonite and vermiculite: Examination of process parameters, kinetics and isotherms, Journal of hazardous materials, 252 (2013) 428-461. [2] B. Liao, W. Y. Sun, N. Guo, S. L. Ding, S. J. Su, Equilibriums and kinetics studies for adsorption of Ni (II) ion on chitosan and its triethylenetetramine derivative, Colloids and Surfaces A: Physicochemical and Engineering Aspects, 501 (2016) 32-41. [3] F. Fu, Q. Wang, Removal of heavy metal ions from wastewaters: a review, Journal of environmental management, 92 (2011) 407-418. [4] Y. H. Peng, J. N. Wang, X. Yang, C. Cheng, T. Wintgens, Preparation of a novel chelating resin for the removal of Ni2+ from water, Chinese Chemical Letters, 25 (2014) 265-268.

18

[5] J. Wang, L. Xu, C. Cheng, Y. Meng, A. Li, Preparation of new chelating fiber with waste PET as adsorbent for fast removal of Cu2+ and Ni2+ from water: kinetic and equilibrium adsorption studies, Chemical engineering journal, 193 (2012) 31-38. [6] S. Sirianuntapiboon, O. Ungkaprasatcha, Removal of Pb2+ and Ni2+ by bio-sludge in sequencing batch reactor (SBR) and granular activated carbon-SBR (GAC-SBR) systems, Bioresource Technology, 98 (2007) 2749-2757. [7] M. Ghaedi, A. Ghaedi, M. Hossainpour, A. Ansari, M. Habibi, A. Asghari, Least square-support vector (LS-SVM) method for modeling of methylene blue dye adsorption using copper oxide loaded on activated carbon: Kinetic and isotherm study, Journal of Industrial and Engineering Chemistry, 20 (2014) 1641-1649. [8] M. Ghaedi, E. Nazari, R. Sahraie, M. Purkait, Kinetic and isotherm study of Bromothymol Blue and Methylene blue removal using Au-NP loaded on activated carbon, Desalination and Water Treatment, 52 (2014) 5504-5512. [9] M. Ghaedi, A. Ghaedi, A. Ansari, F. Mohammadi, A. Vafaei, Artificial neural network and particle swarm optimization for removal of methyl orange by gold nanoparticles loaded on activated carbon and Tamarisk, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 132 (2014) 639-654. [10] S. Davoodi, F. Marahel, M. Ghaedi, M. Roosta, A. Hekmati Jah, Tin oxide nanoparticles loaded on activated carbon as adsorbent for removal of Murexide, Desalination and Water Treatment, 52 (2014) 7282-7292. [11] M. Martínez-Cabanas, M. López-García, J.L. Barriada, R. Herrero, M.E.S. de Vicente, Green synthesis of iron oxide nanoparticles. Development of magnetic hybrid materials for efficient As(V) removal, Chemical Engineering Journal, 301 (2016) 83-91. [12] A. Asfaram, M. Ghaedi, S. Hajati, A. Goudarzi, E.A. Dil, Screening and optimization of highly effective ultrasound-assisted simultaneous adsorption of cationic dyes onto Mn-doped Fe3O4-nanoparticle loaded activated carbon, Ultrasonics Sonochemistry, 34 (2017) 1-12. [13] E.A. Dil, M. Ghaedi, A. Ghaedi, A. Asfaram, M. Jamshidi, M.K. Purkait, Application of artificial neural network and response surface methodology for the removal of crystal violet by zinc oxide nanorods loaded on activate carbon: kinetics and equilibrium study, Journal of the Taiwan Institute of Chemical Engineers, 59 (2016) 210-220. [14] R. Safi, A. Ghasemi, R. Shoja-Razavi, Factors controlling magnetic properties of CoFe2O4 nanoparticles synthesized by chemical co-precipitation: Modeling and optimization using response surface methodology, Ceramics International, 42 (2016) 15818-15825. [15] F.N. Azad, M. Ghaedi, K. Dashtian, M. Montazerozohori, S. Hajati, E. Alipanahpour, Preparation and characterization of MWCNTs functionalized by N-(3-nitrobenzylidene)-N′-trimethoxysilylpropylethane-1, 2-diamine for the removal of aluminum(III) ions via complexation with eriochrome cyanine R: spectrophotometric detection and optimization, RSC Advances, 5 (2015) 61060-61069. [16] S. Sun, J. Yang, Y. Li, K. Wang, X. Li, Optimizing adsorption of Pb (II) by modified litchi pericarp using the response surface methodology, Ecotoxicology and environmental safety, 108 (2014) 29-35. [17] E.A. Dil, M. Ghaedi, A. Asfaram, S. Hajati, F. Mehrabi, A. Goudarzi, Preparation of nanomaterials for the ultrasound-enhanced removal of Pb2+ ions and malachite green dye: Chemometric optimization and modeling, Ultrasonics Sonochemistry, 34 (2017) 677-691. [18] J. Zolgharnein, M. Bagtash, N. Asanjarani, Hybrid central composite design approach for simultaneous optimization of removal of alizarin red S and indigo carmine dyes using cetyltrimethylammonium bromide-modified TiO 2 nanoparticles, Journal of Environmental Chemical Engineering, 2 (2014) 988-1000. 19

[19] E.A. Dil, M. Ghaedi, G.R. Ghezelbash, A. Asfaram, A.M. Ghaedi, F. Mehrabi, Modeling and optimization of Hg2+ ion biosorption by live yeast Yarrowia lipolytica 70562 from aqueous solutions under artificial neural network-genetic algorithm and response surface methodology: kinetic and equilibrium study, RSC Advances, 6 (2016) 54149-54161. [20] H. Mazaheri, M. Ghaedi, A. Asfaram, S. Hajati, Performance of CuS nanoparticle loaded on activated carbon in the adsorption of methylene blue and bromophenol blue dyes in binary aqueous solutions: Using ultrasound power and optimization by central composite design, Journal of Molecular Liquids, 219 (2016) 667-676. [21] M. Ghaedi, S. Hajati, M. Zaree, Y. Shajaripour, A. Asfaram, M. Purkait, Removal of methyl orange by multiwall carbon nanotube accelerated by ultrasound devise: Optimized experimental design, Advanced Powder Technology, 26 (2015) 1087-1093. [22] S. Shahamirifard, M. Ghaedi, M. Rahimi, S. Hajati, M. Montazerozohori, M. Soylak, Simultaneous extraction and preconcentration of Cu2+, Ni2+ and Zn2+ ions using Ag nanoparticle loaded activated carbon: response surface methodology, Advanced Powder Technology, 27 (2016) 426-435. [23] A. Asfaram, M. Ghaedi, S. Hajati, A. Goudarzi, Synthesis of magnetic γ-Fe2O3-based nanomaterial for ultrasonic assisted dyes adsorption: Modeling and optimization, Ultrasonics Sonochemistry, 32 (2016) 418-431. [24] A.B. Albadarin, C. Mangwandi, Mechanisms of Alizarin Red S and Methylene blue biosorption onto olive stone by-product: Isotherm study in single and binary systems, Journal of environmental management, 164 (2015) 86-93. [25] D. Gusain, F. Bux, Y.C. Sharma, Abatement of chromium by adsorption on nanocrystalline zirconia using response surface methodology, Journal of Molecular Liquids, 197 (2014) 131-141. [26] E.A. Dil, M. Ghaedi, A. Ghaedi, A. Asfaram, A. Goudarzi, S. Hajati, M. Soylak, S. Agarwal, V.K. Gupta, Modeling of quaternary dyes adsorption onto ZnO–NR–AC artificial neural network: Analysis by derivative spectrophotometry, Journal of Industrial and Engineering Chemistry, 34 (2016) 186-197. [27] D.C. Montgomery, Design and analysis of experiments, John Wiley & Sons, 2008. [28] E.A. Dil, M. Ghaedi, A. Asfaram, F. Mehrabi, A.A. Bazrafshan, A.M. Ghaedi, Trace determination of safranin O dye using ultrasound assisted dispersive solid-phase micro extraction: Artificial neural network-genetic algorithm and response surface methodology, Ultrasonics Sonochemistry, 33 (2016) 129140. [29] R. Safi, A. Ghasemi, R. Shoja-Razavi, M. Tavousi, The role of pH on the particle size and magnetic consequence of cobalt ferrite, Journal of Magnetism and Magnetic Materials, 396 (2015) 288-294. [30] T. Prabhakaran, J. Hemalatha, Combustion synthesis and characterization of cobalt ferrite nanoparticles, Ceramics International, 42 (2016) 14113–14120. [31] Y. L. Han, J. Gao, Y. Y. Yin, Z. Y. Jin, X. M. Xu, H. Q. Chen, Extraction optimization by response surface methodology of mucilage polysaccharide from the peel of Opuntia dillenii haw. fruits and their physicochemical properties, Carbohydrate Polymers, 151 (2016) 381–391. [32] M. Roosta, M. Ghaedi, A. Daneshfar, R. Sahraei, A. Asghari, Optimization of the ultrasonic assisted removal of methylene blue by gold nanoparticles loaded on activated carbon using experimental design methodology, Ultrasonics sonochemistry, 21 (2014) 242-252. [33] S. Muthusamy, S. Venkatachalam, Competitive biosorption of Cr (VI) and Zn (II) ions in single-and binary-metal systems onto a biodiesel waste residue using batch and fixed-bed column studies, RSC Advances, 5 (2015) 45817-45826.

20

[34] A. Asfaram, M. Ghaedi, A. Goudarzi, M. Rajabi, Response surface methodology approach for optimization of simultaneous dye and metal ion ultrasound-assisted adsorption onto Mn doped Fe3O4-NPs loaded on AC: kinetic and isothermal studies, Dalton Transactions, 44 (2015) 14707-14723. [35] S. Agarwal, I. Tyagi, V. Gupta, A. Bagheri, M. Ghaedi, A. Asfaram, S. Hajati, A. Bazrafshan, Rapid adsorption of ternary dye pollutants onto copper (I) oxide nanoparticle loaded on activated carbon: Experimental optimization via response surface methodology, Journal of Environmental Chemical Engineering, 4 (2016) 1769-1779. [36] F. Ansari, M. Ghaedi, M. Taghdiri, A. Asfaram, Application of ZnO nanorods loaded on activated carbon for ultrasonic assisted dyes removal: Experimental design and derivative spectrophotometry method, Ultrasonics sonochemistry, 33 (2016) 197-209. [37] A. Ghaedi, M. Ghaedi, A. Vafaei, N. Iravani, M. Keshavarz, M. Rad, I. Tyagi, S. Agarwal, V.K. Gupta, Adsorption of copper (II) using modified activated carbon prepared from Pomegranate wood: optimization by bee algorithm and response surface methodology, Journal of Molecular Liquids, 206 (2015) 195-206. [38] Z. Hajahmadi, H. Younesi, N. Bahramifar, H. Khakpour, K. Pirzadeh, Multicomponent isotherm for biosorption of Zn (II), CO (II) and Cd (II) from ternary mixture onto pretreated dried Aspergillus niger biomass, Water Resources and Industry, 11 (2015) 71-80. [39] S. Agarwal, I. Tyagi, V.K. Gupta, M. Dastkhoon, M. Ghaedi, F. Yousefi, A. Asfaram, Ultrasoundassisted adsorption of Sunset Yellow CFC dye onto Cu doped ZnS nanoparticles loaded on activated carbon using response surface methodology based on central composite design, Journal of Molecular Liquids, 219 (2016) 332-340. [40] A. Asfaram, M. Ghaedi, S. Hajati, A. Goudarzi, A.A. Bazrafshan, Simultaneous ultrasound-assisted ternary adsorption of dyes onto copper-doped zinc sulfide nanoparticles loaded on activated carbon: optimization by response surface methodology, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 145 (2015) 203-212. [41] M. Ghaedi, A. Ghaedi, F. Abdi, M. Roosta, R. Sahraei, A. Daneshfar, Principal component analysisartificial neural network and genetic algorithm optimization for removal of reactive orange 12 by copper sulfide nanoparticles-activated carbon, Journal of Industrial and Engineering Chemistry, 20 (2014) 787795. [42] S. Agarwal, I. Tyagi, V.K. Gupta, M. Dehghani, J. Jaafari, D. Balarak, M. Asif, Rapid removal of noxious nickel (II) using novel γ-alumina nanoparticles and multiwalled carbon nanotubes: Kinetic and isotherm studies, Journal of Molecular Liquids, 224 (2016) 618-623. [43] I. Lakhdhar, D. Belosinschi, P. Mangin, B. Chabot, Development of a bio-based sorbent media for the removal of nickel ions from aqueous solutions, Journal of Environmental Chemical Engineering, 4 (2016) 3159-3169. [44] L. Mangaleshwaran, A. Thirulogachandar, V. Rajasekar, C. Muthukumaran, K. Rasappan, Batch and fixed bed column studies on nickel (II) adsorption from aqueous solution by treated polyurethane foam, Journal of the Taiwan Institute of Chemical Engineers, 55 (2015) 112-118. [45] S. Yang, Y. Wu, A. Aierken, M. Zhang, P. Fang, Y. Fan, Z. Ming, Mono/competitive adsorption of Arsenic (III) and Nickel (II) using modified green tea waste, Journal of the Taiwan Institute of Chemical Engineers, 60 (2016) 213-221. [46] A. Ghaee, M. Shariaty-Niassar, J. Barzin, A. Zarghan, Adsorption of copper and nickel ions on macroporous chitosan membrane: equilibrium study, Applied Surface Science, 258 (2012) 7732-7743. [47] K.A. Krishnan, K.G. Sreejalekshmi, R.S. Baiju, Nickel (II) adsorption onto biomass based activated carbon obtained from sugarcane bagasse pith, Bioresource technology, 102 (2011) 10239-10247. 21

[48] F. Yang, S. Sun, X. Chen, Y. Chang, F. Zha, Z. Lei, Mg–Al layered double hydroxides modified clay adsorbents for efficient removal of Pb2+, Cu2+ and Ni2+ from water, Applied Clay Science, 123 (2016) 134-140. [49] R. Ahmad, S. Haseeb, Absorptive removal of Pb2+, Cu2+ and Ni2+ from the aqueous solution by using groundnut husk modified with Guar Gum (GG): Kinetic and thermodynamic studies, Groundwater for Sustainable Development, 1 (2015) 41-49. [50] L. Gonsalvesh, S. Marinov, G. Gryglewicz, R. Carleer, J. Yperman, Preparation, characterization and application of polystyrene based activated carbons for Ni (II) removal from aqueous solution, Fuel Processing Technology, 149 (2016) 75-85. [51] C. Lu, C. Liu, G.P. Rao, Comparisons of sorbent cost for the removal of Ni2+ from aqueous solution by carbon nanotubes and granular activated carbon, Journal of hazardous materials, 151 (2008) 239-246. [52] S. Pap, J. Radonić, S. Trifunović, D. Adamović, I. Mihajlović, M.V. Miloradov, M.T. Sekulić, Evaluation of the adsorption potential of eco-friendly activated carbon prepared from cherry kernels for the removal of Pb2+, Cd2+ and Ni2+ from aqueous wastes, Journal of Environmental Management, 184 (2016) 297-306.

22

Table caption Table 1. Analysis of variance (ANOVA) for the adsorption percentage of Ni2+ ions. Table 2. Isotherm constant parameters and determination coefficients calculated for the adsorption percentage of Ni2+ ions. Table 3. Kinetic parameters for the adsorption percentage of Ni2+ ions. Table 4. Comparison for the adsorption of Ni2+ ions onto the CoFe2O4-NPs-NBNPIEA by different adsorbents.

23

Figure caption Fig.1. FT-IR spectra of CoFe2O4-NPs (a) and CoFe2O4-NPs-NBNPIEA (b). Fig. 2. XRD pattern of CoFe2O4 nanoparticles. Fig. 3. Pareto chart (P=0.05) showing the standardized effect of independent variables and their interaction on the adsorption of Ni2+ ions obtained from the CCD. Fig. 4. 3D response surfaces plot for the adsorption of Ni2+ ions. Fig. 5. Profiles for predicated values and desirability function for the adsorption of Ni2+ ions.

24

Table 1. Analysis of variance (ANOVA) for the adsorption percentage of Ni2+ ions. Factor Model A B C D AB AC AD BC BD CD A2 B2 C2 D2 Residual Lack of Fit Pure Error Cor Total Std. Dev. Mean C.V. % PRESS

Sums of squares

Df*

1402 102.4 20.55 97.02 595.4 7.54 7.55 102.1 0.93 17.31 11.35 48.71 36.07 29.56 422 2.85 2.35 0.5 1405 0.44 91.02 0.48 14.57

14 1 1 1 1 1 1 1 1 1 1 1 1 1 1 15 10 5 29

* Degrees of freedom.

25

Mean squares 100.2 102.4 20.55 97.02 595.4 7.54 7.55 102.1 0.93 17.31 11.35 48.71 36.07 29.56 422 0.19 0.23 0.1

F-value

P-value

528.1 539.9 108.4 511.4 3138 39.8 39.8 538.2 4.91 91.3 59.8 256.8 190.2 155.8 2224

< 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 0.0426 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001

2.35

0.1785

R-Squared Adj R-Squared Pred R-Squared Adeq Precision

0.9980 0.9961 0.9896 90.18

Table 2. Isotherm constant parameters and determination coefficients calculated for the adsorption percentage of Ni2+ ions. Isotherm Langmuir

Freundlich

Tempkin

DubininRadushkevich

Equation 1/qe = 1/(KL Qm Ce ) +1/Qm A plot Ce/qe versus Ce should indicate a straight line of slope 1/Qm and an intercept of 1/(KL Qm). ln qe = ln KF + (1/n) ln Ce The values of KF and 1/n were determined from the intercept and slope of linear plot of ln qe versus ln Ce, respectively. qe = B1 ln KT + B1 ln Ce Values of B1 and KT were calculated from the plot of qe against ln Ce. ln qe = ln Qs -Bε2 The slope of the plot of ln qe versus e2 gives B (mol2(kJ2)-1) and the intercept yields the adsorption capacity, Qs (mg g-1).

26

Parameters Qm (mg.g-1) KL (L mg-1) R2

10 mg 151.5 13.20 0.9984

1/n KF (L mg-1)

0.5389 7.925 0.9407

R

2

B1 KT (L mg-1) R2 Qs (mg g-1) B×10-8 R2

31.28 14.84 0.9246 133.5 3.0 0.9120

Table 3. Kinetic parameters for the adsorption percentage of Ni2+ ions.

log (qe-qt) = log (qe) – (k1/2.303)t Plot the values of log (qe-qt) versus t to give a linear relationship from which k1 and qe can be determined from the slope and intercept, respectively.

k1 (min-1)

Ni2+ ions 19 mg L-1 1.891

qe (calc) (mg g-1)

19.33

(t/qt) = 1/(k2qe2) + 1/qe(t) Plot the values of (t/qt) versus t to give a linear relationship from which k1 and qe can be determined from the slope and intercept, respectively.

k2 (min-1)

Kdiff (mg g-1 min-1/2)

Intraparticle diffusion

qt = Kdif t1/2 + C The values of Kdif and C were calculated from the slope and intercept of the plot of qt versus t1/2, respectively.

β (g mg-1)

Elovich

qt =1/β ln(αβ) + 1/β ln(t) Plot the values of (qt) versus ln (t) to give a linear relationship from which α and β can be determined from the slope and intercept, respectively.

Model

Pseudo-firstorder kinetic

Pseudo-secondorder kinetic

plot

Parameters

27

R

2

0.8755 3.625 -1

qe (calc) (mg g )

140.8

R2

0.9810

-1

C (mg g ) R2

-1

5.798 9.369 0.9630 0.1620

-1

α (mg g min )

1328.3

R2

0.8630

qe (exp) (mg g-1)

99.96

Table. 4. Comparison for the adsorption of Ni2+ ions onto the CoFe2O4-NPs-NBNPIEA by different adsorbents. Adsorbent MWCNTs γ-Al2O3 nanoparticle Chitosan/Polyethylene oxide (Cs/PEO) nanofibers PUC Modified green tea waste (MGTW) Macroporous chitosan membrane Activated carbon obtained from sugarcane bagasse pith (SBP-AC) Pal/MgAl–LDH Groundnut husk modified with Guar Gum Polystyrene based activated carbons Single-walled Carbon nanotubes (SWCNTs) Multiwalled Carbon nanotubes (MWCNTs) Granular activated carbon (GAC) Activated carbon prepared from sour cherry/sweet cherry kernels CoFe2O4-NPs CoFe2O4-NPs- NBNPIEA

Adsorption capacity (mg g-1) 94.25 99.14 91.74

Contact time References 30 min 30 min 3h

[42] [42] [43]

24.39 0.3116 5.21 140.85

120 min 3h 24 h 4h

[44] [45] [46] [47]

23.9 6.74 40.82 47.85 38.46 26.39 77.707

20 min 120 min 24 h 12 h 12 h 12 h 30 min

[48] [49] [50] [51] [51] [51] [52]

27.496 151.515

3.5 min 3.5 min

This work This work

28

Fig.1. FT-IR spectra of CoFe2O4-NPs (a) and CoFe2O4-NPs-NBNPIEA (b).

29

Fig. 2. XRD pattern of CoFe2O4 nanoparticles.

30

Fig. 3. Pareto chart (P=0.05) showing the standardized effect of independent variables and their interaction on the adsorption of Ni2+ ions obtained from the CCD.

31

Fig. 4. 3D response surfaces plot for the adsorption of Ni2+ ions.

32

Fig. 5. Profiles for predicated values and desirability function for the adsorption of Ni2+ ions.

33

Highlights
Cobalt ferrite nanoparticles modified with (E)-N-(2-nitrobenzylidene)-2-(2-(2nitrophenyl)imidazolidine-1-yl) ethaneamine.
Ultrasound energy assisted adsorption of Ni2+ ions onto CoFe2O4-NPs-NBNPIEA.
Response surface methodology help in to optimized the adsorption process.
Ultrasound energy assisted adsorption of Ni2+ ions was well described by Langmuir isotherm and second-order kinetic model.
Maximum adsorption capacity was found to be 151.5 mg g-1 adsorption of Ni2+ ions from aqueous solution.

34