Effect of pH on the Absorption Process

The characterization of the prepared nanocomposite was required to confirm the formation of desired nanomaterials. Different techniques such as FT-IR spectroscopy, TGA/DTA, X-ray diffraction and FE-SEM and TEM were employed to survey the morphology and structure of the nanocomposite involved in this study. The morphology of the polyaniline, magnetic nanoparticles and polyaniline-magnetic nanocomposite were investigated by FE-SEM technique. As it can be seen, the FE-SEM images of the polyaniline consist of nanofibrous structure with the diameter in the range of 50 to 100 nm with high level of agglomeration (Figure 2a). TEM image obtained for polyaniline (inserted image at Figure 2a) confirm the nanofibrous structure of polyaniline. The magnetic nanoparticle consists of poly dispersed sphere-like particles with an average diameter of 20 nm (Figure 2b). The FE-SEM image of polyaniline-magnetic nanocomposite presents some different morphology compared with the magnetic nanoparticle. According to the TEM image of PANI/Fe₃O₄ (inserted image at Figure 2c), it can be seen that diameter of the nanofibers is larger than PANI nanofibers and the magnetic nanoparticles adsorbed on PANI nanofibers. These observations might be due to aggregation of PANI nanofibers in pure PANI through the van der Waals interactions and embedding the MNPs inside the PANI nanofibers leading to the construction of PANI/Fe₃O₄ with more porous structure.

The FT-IR technique was employed to identify the characteristics peaks of prepared materials and elucidate the changes in chemical structures. The FT-IR spectrum of the PANI/Fe₃O₄ nanocomposite and PANI are displayed in Figure 4, indicating that most of the characteristic absorption peaks are common between the PANI/Fe₃O₄ nanocomposite and PANI. The broad peak with high intensity at 3450 cm⁻¹is related to the N–H stretching vibration. The peak at 1628 cm^-1, 1488 cm^-1 and 832 cm⁻¹were corresponded to the C=N, C=C and C-H stretching of PANI rings, respectively. The peak at 583 cm^-1 in PANI/Fe₃O₄ nanocomposite spectrum related to the characteristic band of Fe–O vibration, which is involved as the characteristic adsorption peak in the FT-IR spectrum of magnetic particles. These observations demonstrated the successfully prepared polyaniline-magnetite nanocomposite.

Thermal stability of PANI/Fe₃O₄ nanocomposite verified by thermal gravimetric analysis (TGA) and differential thermal gravimetric (DTG). The PANI/Fe₃O₄ nanocomposite exhibits two steps of weight loss (Figure 5). The first step, which occurs in the range of 20–160°C with low mass change percent (4.6%) is related to elimination of physically adsorbed water surround the adsorbate and confirm the stability up to 160 °C; the second step shows 6% decrease in weight by heating up to 400°C, which could be assigned the decomposition of polyaniline in construction of nanocomposite.

X-ray diffraction (XRD) gives further information about the phase structure of the PANI/Fe₃O₄. The XRD patterns of the PANI/Fe₃O₄ nanocomposite and Fe₃O₄ are presented in Figure 6. As it can be seen from the pattern for PANI/Fe₃O₄ nanocomposite, the characteristic peaks locate at 2θ = 32.2⁰, 35.5⁰, 43.2⁰, 53.5⁰, 57.1⁰, 62.9⁰ can be indexed to the cubic type Fe₃O₄ (220), (311), (400), (422), (511) and (440) crystal planes³⁰. These results display the successful modification of Fe₃O₄ nanoparticles surface with polyaniline. It is clear that the diffraction pattern of the PANI/Fe₃O₄ do not show considerably change in peak position compared to the Fe₃O₄ nanoparticle, suggesting that there is no strong interaction between PANI and Fe₃O₄ and consequently the crystalline structure of the Fe₃O₄ nanoparticle was properly preserved in the presence of polyaniline³¹.

3.2. Optimization of parameters:

To achieve the highest extraction efficiency of naproxen by PANI/Fe₃O₄, the influence of some parameters were investigated in the one-at-a-time procedure. The influential parameters include the mass ratio of polyaniline to magnetic nanoparticles, the amount of sorbent, sample pH, ionic strength, extraction time, desorption time, type and volume of elution solvent.

3.2.1. Mass ratio of PANI to MNPs

Different mass ratios of PANI to MNPs were considered to achieve effective extraction of naproxen from aquatic media. The effect of polyaniline content was evaluated in the range of 0-24 mg, while the content of Fe₃O₄ remained at a constant level of 10 mg. According to the obtained results, the extraction efficiency increased by increasing PANI amount up to 20 mg, but no obvious change occurred at higher amounts of PANI. These results also indicate that bare Fe₃O₄ cannot extract naproxen, and PANI has a vital role in the extraction process. Thus, the mass ratio of 2.2 was selected as the optimum value throughout the experiments.

3.2.2. Amount of sorbent

Sorbent amount is one of the important parameters influencing the extraction efficiency in SPE methods. To elucidate the effect of this parameter on the extraction efficiency, a series of experiments were performed by using various amounts of PANI/Fe₃O₄ nanocomposite (22:10) in the range of 2 to 35 mg for the extraction of naproxen with concentration of 1 µg·mL^-1 adjusted at pH 5. As can be seen in Figure 8, increasing the amount of sorbent has led to increase of the extraction efficiency, which can be attributed to this fact that a higher amount of adsorbent supplies more active sites for adsorbate and increase the adsorption efficiency³². It was found that by using 20 mg of the PANI/Fe₃O₄, the quantitative extraction of naproxen from aqueous phases could be achieved. But no obvious change occurred, when the amount of adsorbent increased more than 20 mg. Considering these results, indicating high adsorption efficiency with a low amount of sorbent, it can be ascribed to high dispersity of sorbent in throughout the sample solution and the high contact area between naproxen molecules and PANI/Fe₃O₄, which increase the mass transfer of naproxen. Taking into account these results, 30 mg of PANI/Fe₃O₄ with the mass ratio of 2.2 was used for the following studies.

3.2.3. Effect of pH on the adsorption process

The sample pH values significantly can influence the extraction efficiency by affecting the sorbent surface charge density and subsequently affecting the interactions between the analyte and the adsorbent. In fact, the properties of the sorbent and the analyte elucidate whether adsorption is desirable in acidic or basic solutions ¹³. Whereas magnetic nanoparticle could be partially dissolved in analyte solution with pH below 2 ²², the effect of pH was considered over the range from 2 to 9. As can be seen from Figure 9 the pH has significant effect on naproxen adsorption and the extraction efficiency is higher in acidic solutions than those in neutral and alkaline conditions. The ionic or neutral form of naproxen in aqueous solution depends on the pH of the solution, and thus it can influence the extraction efficiency. Naproxen is planar compounds, which can form both strong π–π bonding and hydrophobic interactions with magnetic polyaniline. At pH above 4, the carboxylic acid group (pK_a= 4.2) in naproxen induces a negative charge to its structure and consequently, increases the affinity of the analyte with sample matrix. In such a case, the role of hydrophobic interaction would be decreased; while, the π–π bonding interaction is still strong enough to keep the analyte absorbed onto the magnetic polyaniline, resulting in a decrease of naproxen adsorption efficiency. Therefore, sample solution with pH of 5 was found to be the appropriate value in the following study.

3.2.4. Time dependency of the process

The contact time is an important parameter influencing the extraction efficiency and verifying the kinetics of the process. Contact time profiles for three initial concentrations of naproxen were investigated by increasing the contact time from 1 to 200 min. It is evident from Figure 10 that the adsorption increases with contact time up to 30 min and then reached the equilibrium condition. Beyond the 30 min, adsorption remains without significant change, indicating the equilibrium condition for extraction process. Therefore, 30 min of contact time is utilized for the rest of experiments.

3.2.5. Effect of ionic strength

The salt content of the aqueous phase affects the electrostatic and nonelectrostatic interactions between the sorbent and the analytes ³³. To investigate the effect of salt content on the adsorption of naproxen, the extractions were performed in the presence of KCl (0–4000 mM). The results presented in Figure 11 indicate that naproxen extraction efficiency was decreased at higher concentration of KCl. In this experiment, the increase of salt concentration result in adverse effect on the extraction efficiency and therefore further experiments were carried out without salt addition. For hydrophobic analyte, the presence of salt stimulates the transfer of the non-polar component to the water surface, minimizing the interaction of an analyte with adsorbent and consequently reduces the extraction efficiency of analytes. This fact was also pointed out previous ^{34, 35} by other authors as “oil effect”. On the other hand, the increase of ionic strength led to increasing the viscosity of aqueous phase, which operated like a barrier in the mass-transfer process and increases the required time for extraction.

3.2.6. Influence of eluting solvent

Selecting the suitable solvent that effectively elutes the adsorbed analyte with minimum volume in the shortest time is required for optimization of the desorption process. In this study, different kinds of eluents e.g., ethanol, methanol, 1,4-dioxane and dimethyl formamide were applied to select the best ones which can elute efficiently naproxen from the sorbent. It is worth to mention that previous result²⁵ indicate that the fluorescence intensity of naproxen was quenched in the acetonitrile, chloroform, dichloromethane and toluene. So these solvents did not use in the following experiments. In these experiments, 50 mL aqueous solutions containing 1 µg mL^-1 of naproxen adjusted at pH 5 were contacted with 30 mg of the sorbents for 30 minutes. The mixture was then separated magnetically and the naproxen remained in the aqueous phase was determined. The loaded adsorbent was then washed two times with distilled water and then dried at 40°C. The adsorbed naproxen was desorbed from the surface of adsorbent by 10 mL of ethanol, methanol, 1,4-dioxaneand dimethyl formamide solvents in 60 minutes to ensure that equilibrium was obtained (Figure 12). This investigation showed that ethanol is able to desorb the naproxen from adsorbent with the highest efficiency compared to other eluting solvents; this may be related to the high solubility of naproxen in ethanol ³⁶. So, ethanol was utilized as desorption solvent in the subsequent experiments. In order to achieve a complete desorption in the same condition (C₀: 1 µg·mL^-1, pH: 5 and V_aq: 50 mL), the solvent volume and desorption time were investigated in the range of 1–4 mL and 2–80 min, respectively. The results showed that desorption process would be complete when 3 mL of ethanol was used for desorption for 40 min. This condition is efficient for quantitatively transfer of adsorbed naproxen to elute solvent.

3.2.7. Reusability of adsorbent

For practical applications, the regeneration and reusability of the spent adsorbent with high adsorption efficiency and the renewal of its primary characteristic during reuse are of prime economic interest 37. The reusability of consumed adsorbent is related to its stability, which has a critical role in the practical application and reduces the environmental pollution. So efforts were done to regenerate the used adsorbent and utilize it repeatedly in several adsorption-desorption cycles. The regeneration of PANI/Fe3O4 nanocomposite was verified by monitoring the change in efficiency of the adsorption of naproxen via several adsorption/desorption cycles. In general, after adsorption process at optimum condition, PANI/Fe3O4 nanocomposite was magnetically retrieved and elution process was carried out by ethanol under the optimized condition and then reused for the next experiment with a fresh solution of naproxen. The results (Table 1 and Figure 13) indicate that after ten consecutive adsorption-desorption cycles the percentage of adsorption efficiency decrease around 13.4%, which shows the high reusability and stability of PANI/Fe3O4magnetic nanocomposite in extraction of naproxen

Kinetic data modeling

Kinetic data for adsorption process provide an elucidate insight about the prediction of the mechanism of absorption. To estimate the kinetic adsorption constant and find the rate-controlling steps like chemisorption or diffusion process, the experimental data at three concentration of naproxen were fitted using three kinetic models: pseudo-first order, pseudo-second order and Weber-Morris diffusion model. The pseudo-first-order rate equation is expressed as follows³⁸:

log⁡qe-qt=logqe-k12.3t

(7)

Where, qe and qt have the same meaning as above mentioned, k1 is the rate constant of the first-order model (min^-1). The values of k1, q_e and correlation coefficient were evaluated from linear plots of log (qe-qt) versus t for the pseudo first order model. It is important to emphasize that in many instances, experimental data do not fit over the whole range of contact time with the pseudo-first-order model properly and this model only at the primary stage of adsorption applicable to experimental data. The pseudo-second-order rate equation is expressed as follows³⁹:

tqt=1k2qe2+tqe

(8)

Where k₂ (g·mg^-1·min^-1) is the rate constant of pseudo-second-order model. The graph of t/q_t versus t gives a straight line that the qe and k2 are evaluated from the slopes and intercepts of this line. The pseudo-second-order kinetics model is based on the assumption that the adsorption process is controlled by the pseudo-chemical reaction.

The possibility of intraparticle diffusion was considered by using the well-known type of diffusion equations that given by Weber-Morris⁴⁰:

qt=kit0.5+I

(9)

Where k_i (mg·g^-1·min^-0.5) and I are (mg·g^-1) Weber-Morris diffusion constants. When diffusion process is the rate-controlling step for the adsorption, the plot of q_t versus t^0.5 will give intraparticle-diffusion rate constant as the slope and I as the intercept. Furthermore, when the plot of q_t versus t goes through the origin, it is indicated that the intraparticle diffusion is the sole rate-limiting step. The kinetic adsorption constant along with related correlation coefficients for pseudo-first-order kinetics and pseudo-second-order kinetics model at three initial concentrations of naproxen are presented in Table 4. The results show that the adsorption mechanism of naproxen does not obey from the pseudo-first-order model. It should be noted that however correlation coefficients (R²) value from fitting the experimental data with the pseudo-first-order model show relatively high value, but calculated q_e value does not close to experimental q_e value ⁴¹. It seems that experimental data fit satisfactorily with pseudo-second-order adsorption model since the correlation coefficient values (0.997 to 0.998) obtained from this model are close to unity and the calculated q_e value shows good agreement with the experimental ones. Consequently, chemisorption process is the rate-controlling step in adsorption of naproxen. It should be noted that the amount of adsorbed naproxen at equilibrium (q_e) increases by increase of initial naproxen concentrations. This trend can be explained in this way that increasing the initial concentration of naproxen prevail the mass transfer resistances of adsorbate between the solution and adsorbent and increase the amount of adsorbed at equilibrium (q_e) ^39,42.

Generally, in solid–liquid sorption process the adsorbate transfer carried out in several steps; including the mass transfer of adsorbate in bulk solution (external diffusion), diffusion of adsorbate from boundary layer film onto the external surface of adsorbent and finally intraparticle diffusion ³³ in the solid phase and within the porous structure of the adsorbent. The kinetic adsorption constants along with R² values for Weber-Morris model are summarized in Table 5. For three initial concentration of naproxen, the plot of q_t versus t^0.5 shows two linear lines with two slopes (k_i) and two intercept (I) ⁴³. The first straight line with a steep slope is assigned to the diffusion of adsorbate via fluid phase to external surface of the adsorbent. Afterward, diffusion of naproxen continues with the intraparticle diffusion into the solid phase of adsorbent through the pores, which constitutes the second straight line with the mild slope. The resulting straight line does not pass through zero, which indicates that intraparticle diffusion is not the rate-limiting step. The I value give insight about the boundary layer thickness, the high thickness means there is high resistance in external adsorbate transfer, which can be decreased by stirring and suggest that intraparticle? diffusion process is the rate-limiting step.

3.2.6. Thermodynamics of the adsorption process

Thermodynamic studies verify the influence of temperature on adsorption equilibrium and elucidate the feasibility, spontaneity, endothermicity or exothermicity of the adsorption process. Negative values of ΔG⁰ (free energy change) support the spontaneous of the process. The magnitude and sign of ΔH⁰ (enthalpy change) provide helpful information about nature of the adsorption; the ΔH⁰ value for physical adsorption normally doesn’t exceed from 4.2 kJ·mol⁻¹, while in chemical adsorption process due to the strong interaction between the adsorbate and adsorbent ΔH⁰ value reach more than 21 kJ·mol^{−1 28,44}. ΔS⁰ (entropy change) parameter is a criterion of disorder of the adsorption process. The following equation was applied for estimation of these parameters:

∆G0=-RTlnKd

(3)

Where K_d is thermodynamic distribution coefficient and determined by using the following equation;

Kd=qeCe

(4)

In this equation q_e denotes the amount of analyte adsorbed onto the adsorbent per liter of the solution at equilibrium condition and C_e denotes the equilibrium concentration (mg·L^-1) of analyte in solution. Considering to the relationship between the changes of free energy with change of enthalpy and change of entropy as follow:

∆G°=∆H°-T∆S0

(5)

It is possible to estimate G°, H° and S° from the slope and intercept of the plot lnK_d versus T^-1 from the following equation:

lnKd=-∆H0RT+∆S0R

(6)

In order to verify the influence of temperature on the process, the adsorption of naproxen by the PANI/Fe₃O₄ was examined in the range of 293-318K. The results reveal that the naproxen adsorption process slowly increased with an increase in temperature. The corresponding thermodynamic values G°, H° and TS° are -5.04, 9.22 and 14.22 kJ·mol^-1, respectively. The positive value of enthalpy indicates naproxen adsorption process is endothermic. The H° value less than 21 kJ·mol^-1 can be taken as evidence for electrostatic interaction between the adsorbate and adsorbent³². Furthermore, polyaniline with conjugated π structure interacts efficiently with aromatic compounds via hydrophobic interactions and π- π interaction. The positive entropy change is the final outcome of two interaction in the adsorption process; the decrease in entropy due to adsorption of the analyte molecules on the adsorbent which decrease the randomness and on the other hands, an increase of enthalpy due to desolvation of the analyte molecules. As regards, dehydration of analyte molecules causes more randomness than was decreased via adsorbate naproxen, thus leading to positive enthalpy change.

Validation of the method

The developed method using the prepared PANI/Fe₃O₄ nanocomposite was employed for quantitative analysis of naproxen in spiked distilled water samples. The spectrofluorimetric intensity versus naproxen concentration was investigated over the range from the 50 to 1000 ng·mL^-1. Calibration curve presents linear dynamic range over the range of 40 to 600 ng·mL^-1. A calibration equation was found to be Y=0.150C+7.64, (C is the concentration of naproxen in ng·mL^-1) with a correlation coefficient of 0.9972 (n=10). The limit of detection was 16.9 ng·mL^-1 (S/N=3) and the precision of the method, which estimated as the standard deviation of recovery of naproxen at the concentration level of 1 µg·mL^-1 was found to be 2.34% (n=8). Comparison of the developed method with other related methods that previously reported for measurement of naproxen (Table 2) presents comparable detection limit and dynamic linear range. To demonstrate the applicability of the proposed method for quantitative analysis of naproxen in real samples, PANI/Fe₃O₄ sorbent was employed for extraction of naproxen in tap water, human plasma and human urine. Considering to the change of extraction recovery for naproxen in various samples, it is possible to demonstrate the matrix effect and efficiency of presented method for the analysis of naproxen. It should be mentioned that no signal was observed from samples before spiking naproxen solution. The relative recovery percentages for three independent replicate experiments were estimated by dividing the extraction recovery of naproxen in the real samples by spiked sample in distilled water. Satisfactory recoveries (Table 3) along with acceptable reproducibility suggest the capability of this sorbent for extraction of naproxen in real samples with the complex matrix, confirming that the presence of interference doesn’t markedly reduce the extraction efficiency. It may be related to the favorable interaction between naproxen and sorbent in the adsorption process.

Conclusion

A simple, rapid, reproducible, convenient and efficient magnetic solid phase extraction method based on the PANI/Fe₃O₄ nanocomposite has been proposed for the analysis of naproxen. Higher surface area due to the porous structure of nanocomposite lead to the satisfying adsorption efficiency with the fewer amount of adsorbate. Another privilege of the present method is the convenient collection of the adsorbent with the external magnetic field, which eliminate the further filtration steps and make this process a low time-consuming method. From this study, it can be concluded that the developed MSPE base on PANI/Fe₃O₄ nanocomposite with high relative recovery, reusability and other good analytical results can be successfully employed for monitoring of naproxen in the real sample. These good characteristics along with intrinsic sensitivity of fluorescence spectroscopy provide a valid and effective method for determination of naproxen. The extraction efficiency of naproxen by PANI/Fe₃O₄ nanocomposite was investigated in view of equilibrium, kinetics and thermodynamics parameters. The developed method with high stability of sorbent, a low detection limit, wide linear range and good precision parameters along with intrinsic sensitivity of fluorescence spectroscopy provide a valid and effective method for monitoring of naproxen in the biological samples.