Evaluation of Imidacloprid adsorption on a new activated carbon based nanocomposite consisting of magnetite and copper oxide nanoparticles: Adsorption modelling, optimisation, and characterisation

ABSTRACT Standard purification systems have mainly demonstrated the inefficient removal of neonicotinoid insecticides with high persistence in the environment. In the present work, a new magnetic adsorbent composed of Fe3O4 and CuO nanoparticles impregnated with activated carbon (AC) has been used to remove imidacloprid (IMCP) as the most common neonicotinoid insecticide from the water solutions. Various techniques such as Fourier Transform infrared spectroscopy, x-ray diffraction, field emission scanning electron microscopy, energy-dispersive x-ray diffraction, and vibrating sample magnetometer were employed to characterise the fabricated composite. The main affecting parameters, including adsorbent amount, solution pH, agitating time, initial pesticide concentration, and temperature, were examined to obtain the optimised conditions. According to the obtained results from kinetic and isotherm modelling, the adsorption mechanism of IMCP on Fe3O4/CuO@AC nanocomposite (NC) is a combination of chemisorption and physisorption phenomena. The presence of heterogeneous sites for IMCP adsorption was shown by fitting the experimental data to the Freundlich isotherm model. In addition, the adsorption process was well consistent with the pseudo-second-order kinetic model, which indicates that chemisorption is the determining step in the adsorption process. Thermodynamic data showed the exothermic and spontaneous nature of the adsorption process. Under optimal conditions (initial concentration of pesticide 10 mg L−1, pH = 7, and contact time of 10 min), 99.6% of IMCP was effectively removed, indicating the excellent ability of the adsorbent to adsorb this insecticide from aqueous solutions.


Introduction
Imidacloprid (IMCP) with the chemical name of [1-(6-chloro-3-pyridylmethyl)-N-nitroimidazolidin-2-ylideneamine] (Figure 1) is a member of neonicotinoids as a relatively new class of insecticides which are widely used for the control of sucking and chewing insects including aphids, thrips, planthoppers, diamondback moths, and whiteflies.They act via binding to postsynaptic nicotinic receptors in the central nervous system of the insects [1,2].IMCP is widely used for the control of insects in various crops such as cereals, maize, vegetables, citrus fruit, and pome fruit because of its high effectiveness and low toxicity to mammals.However, due to the extensive consumption of the IMCP in agricultural production, its high solubility in water (500 mg L -1 ), and half-life in soil is (48-190 days), occurring of its residues in surface waters and plant foods can pose a serious risk to human health because of their interactions with acetylcholinesterase receptors and proteins [1,[3][4][5][6].In addition, recent studies showed that IMCP and other neonicotinoids represent a significant risk to honeybees, aquatic beings, and birds.They are categorised as persistent contaminants, and therefore, they can transport into aquatic ecosystems through the drainage of agricultural areas [7][8][9].The levels of neonicotinoids, especially IMCP, have been monitored in some countries, including the Netherlands, with a highest reported concentration of 0.32 mg L −1 [10][11][12][13].The maximum residual levels of these insecticides in honey and other related products have been announced by the European Union in the range of 50 to 200 μg kg −1 [14].
During the past decade, several methods such as membrane process [15], ion-exchange [16], chemical precipitation [17], coagulation-flocculation [18], photocatalytic reduction [19], biological [20], and adsorption [21] methodologies have been used for the removal of contaminants from the polluted water bodies.However, most of these methods have high operating costs associated with the consumption of large amounts of chemicals and high sludge production as well as low removal efficiency [22][23][24].Among these methodologies, the adsorption strategy has been widely exploited to treat various contaminated food and environmental samples.This process has been proven as a suitable method to remove many pollutants such as pesticides, heavy metals, drugs, etc. from wastewater [25][26][27].
So far some strategies have been developed for the removal of neonicotinoids, including bio purification [28], microbial degrading [20], oxidation processes, and adsorption technique [5].The use of the adsorption process in order to remove these pesticides from polluted water solutions has been rarely reported [5,6,29].
Activated carbon has been known as a standard and effective adsorbent for the purification of municipal, industrial wastewaters, and aqueous solutions for more than decades [21,27,30,31].However, its turbidity creation and difficult separation from effluents is a challenging issue in the removal process.In this regard, cost-effective, simple, and rapid separation of Fe 3 O 4 (magnetite) nanoparticles as a biocompatible adsorbent introduces it as a suitable candidate for composite formation with activated carbon to solve the turbidity problem [32,33].The easy separation of the magnetic adsorbent from an aqueous solution is performed using an external magnetic field [34,35].In addition, the presence of CuO nanoparticles in the composition of Fe 3 O 4 /CuO@ACNC, and the interaction between copper and the amine groups of IMCP molecule leads to the production of an effective adsorbent to remove this contaminant from aqueous solutions [36,37].
So far, few studies have been reported to remove IMCP from aqueous solutions via an adsorption mechanism.The scope of the current work is to focus on developing a new effective sorbent for the adsorption of IMCP due to the vital role of the adsorbent material in this process.To the best of our knowledge, there are no scientific reports on the application of Fe 3 O 4 /CuO@ACNC to remove IMCP from water solutions.Copper oxide nanoparticles, activated carbon, and magnetite nanoparticles were selected to increase the selectivity, adsorption power, and the easy separation of the adsorbent from the solution, respectively.This new activated carbon-based NC was used to effectively remove IMCP from water solutions.

Reagents and standards
All the used chemicals were of analytical grade.All the solutions were prepared using double-distilled water (DDW) throughout the experiments.Standard of IMCP and copper (II) oxide (nanopowder) were obtained from Merck Company (Germany).Iron (III) chloride anhydrous, iron (II) chloride tetra-hydrate, hydrochloric acid solution (HCl), sodium hydroxide (NaOH), and methanol were obtained from Samchun chemicals company (Korea).Activated carbon powder was obtained from Sigma-Aldrich. 100 mg L −1 stock solution of IMCP was used to prepare the standard solutions with different concentrations.Double distilled water (DDW) was used in all experiments.

Apparatus
A Perkin Elmer UV-VIS double beam spectrophotometer was applied to record the UV-Vis absorbance measurements and spectra.A Bruker spectrometer (model Tensor 27, Germany) was used to obtain Fourier transform infrared (FT-IR) spectra.The magnetic properties of the synthesised Fe 3 O 4 nanoparticles and Fe 3 O 4 /CuO@ACNC were recorded using an MDK-VSM instrument (Meghnatis Daghigh Daneshpajouh Co., Iran).The morphology and elemental composition of nanocomposite were evaluated by Field Emission Scanning Electron Microscope (FESEM) and Energy Dispersive X-ray Diffraction (EDX) using a Tescan model Mira3-XMU instrument (Czech Republic).The pattern of the NC x-ray diffraction (XRD) was recorded using a X'Pert Pro MPD instrument (PANalytical, Netherlands).The pH values were adjusted using a Metrohm pH-metre model 623 (Switzerland).

Synthesis of Fe 3 O 4 nanoparticles
The co-precipitation of Fe(II) and Fe(III) chloride as the iron oxides (FeO and Fe 2 O 3 ) in the presence of ammonia solution was exploited for the preparation of Fe 3 O 4 nanoparticles at room temperature as described previously [24].In brief, 5 mL of a solution composed of Fe(II) and Fe(III) with a 1:2 molar ratio-was vigorously stirred, while 50 mL of ammonium hydroxide solution (1.5 M) was added drop-wisely.A black suspension identified the formation of magnetite nanoparticles.The mixture was stirred at 500 rpm for another 10 min at this condition.An external magnetic field was applied to collect Fe 3 O 4 nanoparticles.Then, the collected nanoparticles were thoroughly washed with DDW to eliminate excess ammonia and dried in the oven at 80°C.

Preparation of Fe 3 O 4 /CuO/ACNC
In order to prepare the nanocomposite, 1 g of each CuO and Fe 3 O 4 nanoparticles were mixed with 0.5 g of AC powder and suspended in 50 ml of DDW, homogeneously.After adjusting the suspension pH at 8 using HCl and NaOH solutionsit was heated at 80°C for 10 min.Then the prepared nanocomposite was separated from the supernatant using an external magnetic field and dried at 80°C in the oven.

Design of IMCP adsorption experiments
For the adsorption studies, the batch procedure was used.A beaker containing 10 mL of IMCP solution was used in all the removal experiments.The effects of some influencing experimental parameters, including pH of the solution, the adsorbent amount, contact time, temperature, and initial pesticide concentration, were examined on the adsorption performance.

The effect of pH on adsorption
5 mg of the adsorbent was mixed with 10 mL of IMCP solutions (10 mg L −1 ) in the pH range of 4-10.By adding dilute NaOH or HCl solutions, the pH value of the suspension remained constant throughout the removal process.The mixture was stirred at room temperature (20 °C) at a rate of 500 rpm for 5 min.

The effect of adsorbent amount on adsorption
The influence of the adsorbent amount on the adsorption was examined with 10 mg L −1 IMCP solutions (10 mL) containing 2-30 mg of the adsorbent at pH 7 and agitating at room temperature (20 °C) with the rate of 500 rpm for 5 min.

The effect of contact time on adsorption
For kinetic studies, 20 mg of the adsorbent with 10 mg L −1 IMCP solutions (10 mL) at pH 7 were mixed.The suspensions were shaken at room temperature (20 °C) in fixed intervals, including 1, 2, 5, 10, 20, 30, 40, 50, 60, and 90 min with the rate of 500 rpm.

The effect of initial concentration of IMCP on adsorption
The effect of the initial concentration of the pesticide on adsorption equilibrium was studied by mixing 20 mg of the adsorbent and IMCP solutions (10 mL) with varying initial concentrations in the range of 5 to 30 mg L −1 .These prepared suspensions were stirred for 10 min at pH = 7 and 20°C with the rate of 500 rpm.

Analysis of IMCP
In all adsorption experiments, after separating the adsorbent by the external magnetic field, IMCP concentration in solution was measured by spectrophotometry at the maximum wavelength of the analyte (270 nm).The UV-Vis.spectrum of IMCP was obtained within the wavelength region from 200 to 350 nm to select the maximum wavelength.The spectra related to IMCP solution before and after agitating with Fe 3 O 4 /CuO@ACNC are depicted in Figure 2. In each adsorption experiment, the removal efficiency and the equilibrium adsorption capacity were obtained by the following equations, respectively: where C 0 and C e (mg L −1 ) are IMCP initial and equilibrium concentrations, respectively.In Equation ( 2), q e (mg g −1 ), m (g), and V (L) are the adsorption capacity, adsorbent amount, and solution volume, respectively.All experiments were conducted in triplicate, and the mean values of the measurements were reported as results.

Modelling of the adsorption process: isotherms and kinetics
To design a proper adsorption unit, kinetic studies showing the removal rate of poison from the aqueous solutions are necessary.For modelling the adsorption process of IMCP on Fe 3 O 4 /CuO@ACNC, four common and widely used kinetic models including pseudofirst-order, pseudo-second-order, intra-particle diffusion, and Elovich, were exploited.The linear equations of these models are given in Table 1.
Generally, the adsorption behaviour of molecules onto the surface of adsorbent is revealed by the isotherm studies.In this way, to fit the experimental data of the adsorption process, four common theoretical isotherm models, including Freundlich, Langmuir, Temkin, and Dubinin-Radushkevich (D-R), were examined.The parameters and equations related to these models are presented in Table 2.

The effect of temperature on adsorption
In order to examine the influence of temperature on the adsorption process, a series of experiments were performed by mixing 20 mg of the adsorbent with 10 mL of IMCP solutions (10 mg L −1 ) at pH 7. The suspensions were shaken for 10 min at a rate of 500 rpm at the temperature range of 20-60°C.

Characteristics of Fe 3 O 4 /CuO/ACNC
As shown in Figure 3(a), the FESEM image of the prepared Fe 3 O 4 nanoparticles shows the diameter distribution of around 25 nm with a spherical shape.Moreover, Figure 3(b) shows the morphology of NC with a large number of cavities as well as metal nanoparticles and layered carbon particles.The EDX spectrum of Fe 3 O 4 /CuO@ACNC (Figure 3(c)) exhibits the peaks for iron, copper, carbon, and oxygen, indicating the successful synthesis of the adsorbent.
As shown in Figure 4(a), the FTIR spectra of Fe 3 O 4 nanoparticles (Figure 4(a)), Fe 3 O 4 /CuO@ACNC (Figure 4(a-)), and IMCP adsorbed on Fe 3 O 4 /CuO@ACNC (Figure 4(a-)) are obtained in the wavenumber range of 400-4000 cm −1 .The strong and sharp peak at 573.82 cm −1 can be related to the Fe-O vibration in Fe 3 O 4 nanoparticles (Figure 4(a)) [38].The strong absorption peaks in Figure 4(a-) at 576.12 cm −1 (with some shoulders) and 3433.99 cm −1 can be ascribed to the vibrations of metal oxide bands (Fe-O and Cu-O) and O-H band of H 2 O molecules that adsorb on the NC surface.As shown in Figure 4(a-), after IMCP adsorption, there are almost all characteristic peaks of the components with minor changes due to the physicochemical mechanism of the adsorption process.These shifts can be seen in metal oxide (579.13 cm −1 ) and N-H or O-H vibrations (3412.53cm −1 ).
XRD pattern of Fe 3 O 4 /CuO@ACNC in position at 40 kV and 30 mA with Cu anode over the 2θ: 4-70° is presented in Figure 4(b).The diffraction peaks related to Fe 3 O 4 , CuO, and AC in this XRD spectrum indicated the successful synthesis of the adsorbent without any phase change.
Table 1.The equations and parameters related to the various kinetic models.
Isotherm models Linear equation Plot Parameters Langmuir C e /q e = 1/K L q m +C e /q m (C e /q e ) vs. C e q m (mg/g), K L (L/mg) Freundlich logq e = logK f +1/n logC e logq e vs. ln C e K F (L/mg), n Dubinin-Radushkevich (D-R) lnq e = lnq m -Dε 2 lnq e vs. ε 2 q m (mg/g), D(mol 2 /kJ 2 ) Temkin q e = B lnK T +B lnC e q e vs. lnC e B, K T A vibrating sample magnetometer (VSM) was used to study the magnetic properties of Fe 3 O 4 and Fe 3 O 4 /CuO@ACNC.As shown in Figure 4(c), the absence of magnetic residual loops in Fe 3 O 4 and Fe 3 O 4 /CuO@ACNC saturation magnetic curves indicates their supramagnetic properties [39].Furthermore, the saturation magnetisation of Fe 3 O 4 (69.1 emu g −1 ) was higher than that of Fe 3 O 4 /CuO@ACNC (42.8 emu g −1 ), revealing the existence of non-magnetic CuO and AC in the NC structure.Despite the reduction in the magnetic property, this adsorbent was efficiently and rapidly eliminated from aqueous solutions using a magnet due to its relatively high saturation magnetisation value.The separation of CuO, as well as graphite from the supernatant together with Fe 3 O 4 can be attributed to the possible hydrogen bonding or electrostatic forces between the nanocomposite components.This phenomenon cannot be observed in the mixture of magnetite and some other metal oxide nanoparticles, for example, in the case of zirconia [40].

The pH effect
In adsorption studies, the solution pH is vital in the ionisation and surface charge of adsorbent functional groups.The influence of solution pH on the removal efficiency of IMCP (Section 2.5.1) is presented in Figure 5(a).As seen, the adsorption efficiency (removal percentage) of IMCP increases substantially with an increase in the solution pH from 4.0 to 7.0 and then decreases in higher pH values.According to these results, the maximum percent of IMCP removal (56.0%) was achieved at a pH value of 7.0, and the adsorbent exhibits the best function in the neutral media.This observation may be due to the neutral molecular structure of IMCP at this pH.In addition, the obtained result is consistent with the literature showing the maximum IMCP adsorption at pH 7 [4,14,32,41].At acidic media, the lower uptake of the adsorbent may be related to the protonation of the amine groups of IMCP molecules and the surface of magnetite and CuO nanoparticles of the adsorbent.This phenomenon, on the one hand, reduces the tendency between IMCP and the adsorbent because of the electrostatic repulsion between the moieties with positive charges.On the other hand, vigorous competition between proton ions and the poison takes place for the adsorption on the active sites of the adsorbent [23,24].However, the observed decrease in the removal rate at alkaline environments (pH > 7) might be correlated to the existence of a large number of hydroxide ions and their competition with IMCP for the active sites of the adsorbent.Considering the maximum removal of IMCP, pH 7 was chosen for the subsequent experiments.

The effect of the adsorbent amount
The influence of adsorbent amount on the adsorption extent of IMCP on Fe3O4/ CuO@ACNC was tested in the range of 2-30 mg under conditions are described in section 2.5.2.As depicted in Figure 5(b), the uptake of IMCP enhances by increasing the adsorbent quantity, and the removal percentage reaches 96.7% using 20 mg of the adsorbent.Moreover, the removal trend shows a direct relationship with the adsorbent amount up to 20 mg, which means that the greater surface area and more active sites provide the adsorption with more efficiency at fixed IMCP initial concentration.After that, increasing the adsorbent dosage cannot cause further adsorption, and thus, the removal rate levels off.The aggregation of Fe 3 O 4 /CuO@ACNC particles at higher amounts and, therefore, the lower usage of the adsorbent active sites might be the reasons for this observed phenomenon [24].Based on these results, 20 mg of Fe 3 O 4 /CuO@ACNC was chosen as the optimal adsorbent amount for further tests.

Kinetic modelling: contact time effect
The effect of stirring time on the adsorption efficiency over a 90-minute period (under optimal pH and adsorbent conditions) was evaluated, as described in Section 2.5.3.As shown in Figure 6(a), equilibrium and almost complete removal of IMCP (99.6%) from the test solution occurs within the first 10 minutes.This rapid increase in the removal percentage in the initial contact times and reaching to an equilibrium state in a concise time can be explained in the context of fast kinetics of adsorption reaction (immediate occupation) and saturation of reactive sites of the adsorbent.The previous studies have reported similar trends [25,42].According to the findings, the subsequent experiments were carried out in the optimum contact time of 10 min.Besides, based on these experiments, the plots and parameters of four widely used kinetic models, including pseudo-first-order, pseudo-second -order, intra-particle diffusion, and Elovich, were obtained and are given in Figure 6(b) and Table 3.In Table 3, q t , is the adsorption capacity at a given time t, and q 1 , and q 2 are the adsorption capacity at equilibrium state at pseudo-first-order and pseudo-second-order kinetic models, respectively.The rate constants of pseudo-first-order, pseudo-second-order, and intra-particle diffusion kinetic models are K 1 , K 2 , and K ip , respectively.Firstly, the correlation coefficients (R 2 ) obtained from the linear form of equations related to the kinetic models (Figs S1-S4) were compared.Then, the highest value of R 2 was used to choose the most proper kinetic model fitting the IMCP adsorption on Fe3O4/CuO@ACNC.As can be seen from Table 3, the R 2 obtained for the pseudo-second-order kinetic model (R 2 = 0.9991) is the highest compared to the other models, indicating that this model fits better by absorbing IMCP on the adsorbent.The plots of various kinetic models (Figure 6(b)) also confirm that the experimental data of IMCP adsorption onto Fe 3 O 4 /CuO@ACNC agrees with the pseudo-second-order model, implying that the determining step is chemisorption.Furthermore, the calculated q 2, cal.value (5.076 mg g −1 ) is closer to the experimental q e, exp.(4.688 mg g −1 ) than those of the other models, which is another confirmation for chemisorption mechanism of IMCP in the adsorption process [26,43,44].The adsorption process probably involves the electron sharing between IMCP molecules (via free electrons of nitrogen atoms) and the binding sites of the adsorbent, especially CuO nanoparticles [22].
The other models showed lower R 2 values revealing their lesser role in the removal process.Besides, for the intra-particle diffusion model, the value of C i >0 indicates that linear parts of the curve (q t vs. t 0.5 ) do not pass through the origin (Table 3).This observation reveals that although this model can control the adsorption process to some extent by diffusion in the boundary layer, but is not the only rate-controlling step [23,44,45].

Isotherm modelling: IMCP initial concentration effect
The effect of initial concentrations of IMCP ranging from 5-30 mg L −1 on its adsorption by Fe 3 O 4 /CuO@ACNC (Section 2.5.4) is depicted in Figure 7(a).As seen, the removal efficiency decreases by increasing IMCP concentration from 99.7% to 67.1% mg L −1 for 5 and 30 mg L −1 , respectively, probably owing to the fixed number of active sites on the adsorbent as Table 3.The obtained parameters for various kinetic models.

Kinetic Models
Parameters Values Pseudo-first order q 1, cal.(mg g −1 ) 6.057 K 1 (min −1 ) 3.561 R 2 0.9360 Pseudo-second order q 2, cal.(mg g −1 ) 5.076 K 2 (g mg −1 min −1 ) 0.172 R 2 0.9991 Elovich α (mg g −1 min −1 ) 21.18 β (g mg −1 ) 0.798 R 2 0.7637 Intra-particle diffusion K ip (mg g −1 min 0.5 ) 0.3964 C i (mg g −1 ) 2.553 R 2 0.5417 Experimental q e, Exp.4.688 reported in previous studies [23].However, the adsorption capacity increases with increasing initial IMCP concentration due to its increased adsorption per unit mass of adsorbent (from 2.47 to 14.61 mg g −1 ).This observation might be owing to the increase of the driving force to overcome the mass transfer resistance of IMCP molecules between the solid and aqueous phases [42].Under the optimal conditions, the linear equations of four isotherm models, including Freundlich, Langmuir, Temkin, and D-R, were used to explain the adsorption of IMCP in the concentration range of 5-30 mg L −1 (Figs S5-S8).The obtained graphs for these isotherm models and also values for the equilibrium isotherm parameters are presented in Figure 7 (b) and Table 4.In this Table, K L , K F , and K T are Langmuir, Freundlich, and Temkin constants, respectively.Besides, in the D-R model, ε (kJ mol −1 ) is the Polanyi potential and can be determined via ε = RT ln (1 + (1/Ce) equation.The q m , n, R L , and E parameters are the maximum adsorption capacity, Freundlich adsorption heterogeneity factor, the separation parameter, and the mean free energy of adsorption, respectively [45].
As shown in Figure 7(b) and Table 4, based on the R 2 values, the isotherm models are fitted to the experimental results in the following order: Freundlich > Langmuir > Temkin > D-R.The correlation coefficient of 0.9924 for the Freundlich model suggests the presence of heterogamous conditions in the adsorption process.The Langmuir and Freundlich isotherms show the single-layer (homogeneous) and multi-layer (heterogeneous) mechanisms of the adsorption process, respectively.In addition, the indirect interactions between the adsorbent and the adsorbate are considered in the Temkin model, and the D-R isotherm model shows a surface area with the homogeneous adsorption energy [24,46].
In the D-R model, the E value is obtained by use of the following equation: The desirability of IMCP adsorption onto the adsorbent was studied using R L (1/ (1 + C 0 K L ) ) as a dimensionless parameter which is derived from the Langmuir model (Table 4).Generally, the adsorption process will be favourable if R L lies within 0 and 1 (0 < R L < 1).
For R L = 0, R L > 1, and R L = 1, the adsorption is irreversible, unfavourable, and linear, respectively [46].A favourable behaviour of IMCP onto the adsorbent surface was confirmed by the R L value (0.060), as shown in Table 4.The 1/n value in the Freundlich model shows the favourability of IMCP adsorption on Fe 3 O 4 /CuO/AC.The small value of 1/n (0.318), between 0 and 1, and the considerable value of K F (5.974 mg g −1 ) confirm the effective adsorption of IMCP molecules on the adsorbent.The E value in the D-R model gives an idea about either chemical or physical adsorption according to its magnitude.With the E values of 8-16 kJ mol −1 the mechanism of adsorption can be a chemical process, while the values of E < 8 kJ mol −1 show that the physical adsorption mechanism is predominant [23].As can be seen from Table 4, the E value for IMCP adsorption on Fe 3 O 4 /CuO@ACNC (4.784 kJ mol − 1) is lower than 8 kJ mol −1 , indicating that the dominant mechanism is the physical adsorption.Generally, in the adsorption process, the chemisorption mechanism is the slowest step, while the physisorption is the primary mechanism in the system, as mentioned before (Section 3.2.3)[24].

Determination of the thermodynamic parameters: the effect of temperature
The effect of solution temperature was studied in the range of 20-60 °C under conditions described in Section 2.7, and the optimal temperature for IMCP adsorption and the related thermodynamic parameters was obtained.The results revealed that in the temperatures higher than 20°C, the removal efficiencies of IMCP slow down gradually, showing that the lower temperatures are favourable in the adsorption process (Fig. S9).The thermodynamic parameters, including standard enthalpy change (ΔH 0 , kJ mol −1 ), standard entropy change (ΔS 0 , kJ mol −1 K −1 ), and Gibbs energy (ΔG 0 , kJmol −1 ) were calculated from Van't Hoff equation [47]: In Equation ( 4), R is the gas constant (8.3145Jmol −1 K −1 ).K d is the distribution coefficient -(Equation 5) which is defined as the ratio of the adsorbed IMCP concentration (C s ) to the aqueous phase IMCP concentration (C e ) at a given temperature (T) at equilibrium state [48].
The ΔG value was calculated using Equation ( 4).In addition, ΔH 0 and ΔS 0 values were determined using slope and intercept of ln K d versus 1/T plot from another form of Van't Hoff equation according to Equation (6), respectively (Fig. S10).
In Table 5, the calculated parameters and their values are presented.As observed, all three thermodynamic parameters (ΔH 0 , ΔS 0 , and ΔG 0 ) have a negative value under the standard conditions (25°C and 1 atm pressure).The negative ΔH 0 value indicates the exothermic behaviour of the adsorption process and describes the decrease in the sorption ability of the adsorbent with an increase in the solution temperature.Furthermore, the negative value of ΔS 0 confirms the affinity of the adsorbent for IMCP and also reveals that the randomness decreases at the solid-solution interface during the sorption phenomenon.As seen from Table 5, the ΔG value for all the examined temperatures is negative, demonstrating the spontaneous nature of the adsorption process.Moreover, with an increase in the temperature, the value of ΔG 0 increases, indicating that lower temperatures are favourable for the IMCP adsorption and the enhancement in temperature reduces the spontaneity of the removal reaction.According to ΔG 0 value, the adsorption process is categorised as chemisorption (−80 to −400 kJ mol −1 ) or physisorption (0 to −20 kJ mol −1 ).From Table 5, it can be concluded that the sorption process has a physical nature (ΔG 0 = −7.756kJ mol −1 ), and physical adsorption is the dominant mechanism in this system.This result is consistent with the data documented in Table 4, as the E value refers to the physisorption phenomenon.

Comparison of the performance of Fe 3 O 4 /CuO@ACNC with other sorbents
Some features of the adsorption process on the removal of IMCP by Fe 3 O 4 /CuO@ACNC, including q m , pH, temperature, contact time, and R%, are compared with the results reported in the literature using other adsorbents, as presented in Table 6.As can be seen, the characteristics obtained from the present method in some cases are better (contact time and removal efficiency) or comparable (q m , temperature, and pH) to other studies.
Although some adsorbents provide q m values for IMCP higher than Fe 3 O 4 /CuO@ACNC, they have various operating limitations such as the lengthy removal process and the difficult separation from the sample solution.

Application of the proposed adsorption method in real samples
Three real water samples, including mineral water, tap water, and Karun River, were tested to evaluate the adsorption performance of Fe 3 O 4 /CuO@ACNC.Real water samples were spiked by 10.0 and 20.0 mg L −1 of IMCP and subjected to the optimal adsorption method to investigate the effects of real matrices on removal efficiency.As shown in Table 7, the removal efficiencies of 85.3% to 97.9% were obtained, indicating that the adsorbent has an excellent ability to remove IMCP from real water samples without a significant matrix effect.

Conclusion
Herein, a selective and effective approach for IMCP elimination from water solutions has been developed with the removal efficiency of about 99% using Fe 3 O 4 /CuO@ACNC through the adsorption process.The experimental results showed that operating factors  such as pH, adsorbent amount, temperature, and initial concentration of IMCP remarkably influence adsorption efficiency.Fe 3 O 4 /CuO@ACNC exhibited better performance at pH 7 and 25°C with an initial concentration level of 10 mg L −1 .The optimum contact time of 10 min implied that the adsorption process reaches the equilibrium state in a very short time, which can be related to the high tendency of the adsorbent to IMCP.The adsorption process was consistent with the pseudo-second-order kinetic model and the Freundlich isotherm model.According to these results, it was proved that both physisorption and chemisorption mechanisms were involved in the adsorption process.The synthesised adsorbent presented the high saturation magnetisation (42.8 emu/g), demonstrating its rapid and easy isolation from water solution.This inexpensive adsorbent composed of metal oxides nanoparticles and activated carbon showed a rapid removal process in the soft conditions in terms of pH and temperature with the high adsorption performance and the excellent capability for water treatment.

Figure 2 .
Figure 2. UV-Vis spectra of 20 mg L −1 IMCP solution before (a) and after (b) exposure to the adsorbent.

Figure 5 .
Figure 5.The influence of a) pH and b) the adsorbent amount on the removal efficiency of IMCP.

Figure 6 .
Figure 6.The effect of contact time on the removal efficiency (a), and the kinetic models and experimental data (b) for the IMCP adsorption on Fe 3 O 4 /CuO/ACNC.

Figure 7 .
Figure 7. effect of initial IMCP concentration on the removal efficiency (a), and the isotherm models and experimental data (b) for the IMCP adsorption on Fe 3 O 4 /CuO/ACNC.

Table 4 .
The obtained parameters for various isotherm models.

Table 5 .
The thermodynamic parameters related to the IMCP adsorption on the adsorbent.

Table 6 .
Comparison of adsorption process information of several adsorbents reported in the literature.