Removal of pyrene and benzo(a)pyrene micropollutant from water via adsorption by green synthesized iron oxide nanoparticles

Ferrous sulfate, sodium hydroxide, methanol, pyrene, benzo(a)pyrene and acetone, were purchased from Sigma-Aldrich, Germany. The stock solutions of pyrene and benzo(a)pyrene were prepared in acetone. Series of variable concentrations were prepared by the appropriate dilutions with distlled water.

An amount of 2.0 g of the dried pomegranate peel solid waste was placed in 200 ml beaker, followed by the addition of 100 ml of deionized water. The mixture was gently heated until the color of the aqueous solution changes to dark yellow. The extract was cooled to room temperature and filtered through a Whatman filter paper (no. 40). Phytochemical screening of pomegranate peel extract was qualitatively estimated according to the method described by Selvaraj et al [24].

An amount of 4.0 g ferrous sulfate was added to 1 l of distilled water. The mixture was stirred vigorously for 10 min, till complete dissolution of ferrous sulfate. A 20 ml aliquot of the previously prepared pomegranate peel solid waste extract was added to the ferrous sulfate solution. The mixture was stirred in a very low speed (20 rpm). A solution of 2.0 M NaOH was drop wise, added until the pH increased to 12. The mixture was further continuously stirred at low speed (20 rpm) to allow the precipitation of IONPs. The obtained dark brown precipitate was collected and washed several times with double distilled water. The centrifuge was utilized after each step of washing in order to collect the synthesized IONPs. The collected nanoparticles were left to dry overnight at 70 °C to obtain the IONPs powder. To confirm the role of the pomegranate peel extract on the green synthesis of IONPs, the above preparation procedure was repeated without the addition of this extract. The obtained precipitate was collected and dried at 70 °C overnight. The prepared precipitate was then examined.

The crystallinity of the synthesized IONPs was investigated by XRD (PANalytical–Empyrean) under the following conditions: scan axis: Gonio, start position 2θ  =  5.0129°, end position 2θ  =  79.9709°, step size 2θ  =  0.0260°, scan step time is 18.8700 sec, scan type: continuous, PSD length 2θ  =  3.35°, measurement temperature is 25 °C, anode material: Cu-Kα1 is 1.540 60 Å, Cu-Kα2 is 1.54443 Å, Cu-Kβ is 1.392 25 Å, ratio Kα2/Kα1  =  0.500 00, generator settings: 30 mA, 45 kV. SEM (quanta FEG 250) was employed to examine the surface morphology of the synthesized IONPs. In addition, EDX (AMETEK elemental analysis device) was used for the characterization of IONPs. Furthermore, HRTEM (JEOL JEM–2100 electron microscope) was used to gain the actual particle size of the synthesized IONPs.

For the detection of benzo(a)pyrene and pyrene concentrations, the high performance liquid chromatograph (HPLC) was used (Agilent 1200 HPLC). It is equipped with a 5 µm  ×  25.0 cm  ×  4.6 mm LC-C18 column, PDA detector and auto-sampler. The mobile phase consisted of 40% water and 60% acetonitrile. The solvent program was isocratic, the flow rate was 1.0 ml min⁻¹ and the injection volume was 5 µl.

2.5.1. Determination of IONPs optimum adsorption dose.

2.5.2. Study the effect of pH, temperature, contact time, TDS and initial concentration of PAHs on the efficiency of adsorption.

In order to perform the adsorption process more economically, it was important to study the reuse of the adsorbent. The regeneration and reuse of the green synthesized IONPs was investigated as follows: a mixed solution of 3 µg l⁻¹ of benzo(a)pyrene and 300 µg l⁻¹ pyrene was allowed to be in contact with 90 mg l⁻¹ green synthesized IONPs at pH 7 and room temperature for 150 min. The green synthesized IONPs were then separated by centrifugation. The collected green synthesized IONPs were rinsed 3 times with 3 ml methanol each to remove pyrene and benzo (a) pyrene. The concentrations of pyrene and benzo(a)pyrene in the separated supernatant solutions were determined by HPLC, until no pyrene and benzo(a)pyrene were detected. The rest of the methanol supernatant solution that contains pyrene and benzo(a)pyrene was submitted to rotary evaporator to get rid of the methanol. The obtaind solid pyrene and benzo(a)pyrene could be reused again.

The regenerated IONPs were reused again for adsorbing of pyrene and benzo(a)pyrene for several times according to the previously described procedure. The purpose is to investigate the effect of IONP reuse on the adsorption capacity. The regeneration and reuse process was repeated 5 times.

To study the efficiency of the iron oxide nanoparticles for the removal of the studied contaminants from drinking water, a semi-pilot plant was designed for a batch flow system (figure 1). Drinking water samples were brought to the pilot plant, where water was artificially contaminated with benzo(a)pyrene and pyrene at the concentration of 1 µg l⁻¹ of benzo(a)pyrene and 100 µg l⁻¹ pyrene. The physical and chemical characteristics of this water were carried out according to the standard methods [25].

Zoom In Zoom Out Reset image size

The semi-pilot plant (figure 1) consists of a mixing tank (1.0  ×  0.8  ×  0.8 m³) and is equipped with mechanical stirrer in middle of the tank. In this mixing tank the artificial contaminated tap water was placed and the iron oxide nanoparticles were added at the pre-determined dose, namely, 110 mg l⁻¹. The stirrer is to serve for the complete mixing of the contaminated tap water with the iron oxide nanoparticles. The mixture was stirred for 150 min, continuously, after which the stirred mixture was allowed to flow into the first sedimentation tank followed by the second sedimentation tank. Each sedimentation tank is partitioned by baffle to allow the settling down of the particles. The dimensions of each sedimentation tank are (0.8  ×  0.35  ×  0.35 m³). Each sedimentation tank is equipped with two magnetic bars in the bottom to enhance the settling of the iron oxide nanoparticles to the bottom of the tanks. The final treated water was then collected for physical and chemical analyses. The settled iron oxide nanoparticles were collected and submitted for regeneration and reuse as described above.

Figure 2 illustrates the XRD pattern of the green syntheize iron oxide nanoparticles. There are two small observed peaks at 2θ positions, namely: 38.8062° and 55.4002° refers to Fe₂O₃ (figure 2(a)). This observation is an indication of the amorphous structure of the green syntheized IONPs. It has been reported that amorphous metal oxides show great industrial potential in solar energy transformation, sorption, purification processes and catalysis. In such applications, amorphous iron (III) oxide plays a key role. This is mainly due to its superior catalytic activity, super-paramagnetic behavior, and the larger specific surface area of the nano-particles. The super-catalytic behavior of the amorphous IONPs is more effective than the nano-crystalline, polymorphs or particles of metallic iron of the same diameter [26]. On the other hand, figure 2(b) shows the XRD pattern of the chemically synthesized iron oxide particles. It is only one small peak at 2θ  =  12.9662°, as an indication of the amorphous formation of iron hydroxide particles.

Zoom In Zoom Out Reset image size

Meanwhile, figure 3 illustrates the FTIR spectra of green synthesized particles and chemically synthesized particles, respectively. In this figure, the presence of phytochemical compounds on the surface of the green synthesized particles, indicates the following differences:

• (1)  
  Appearance of a new low intensity peak at 3690.12 cm⁻¹ (due to free OH stretching vibration, mainly for phenols), 3233 cm⁻¹ (due to intermolecular hydrogen bond stretching, vibration), 2418.3 cm⁻¹ (due to C–H stretching, vibration), 1800.22 cm⁻¹ (due to C=O stretching, vibration), high intensity peak at 1105.98 cm⁻¹ (due to C–N stretching, vibration), and small peak at 1020.16 cm⁻¹ (due to C–O H stretching vibration), 785.85 cm⁻¹ (due to  =C–H=CH₂, out-of-plane, bending vibration).
• (2)  
  The shift of the peak at 3761.47 cm⁻¹ (due to vibrational overtones from charge transfer between O and Fe³⁺) to 3769 cm⁻¹, and high intensity peak at 1446.35 cm⁻¹ (due to Fe–OH bending vibration) to low intensity peak at 1440.56 cm⁻¹.
• (3)  
  Disappearance of the high intensity peak at 3427.85 cm⁻¹ (due to OH (involved in hydrogen bond) stretching vibration), low intensity peak 3250.43 cm⁻¹ (due to stretching vibration of OH involved in hydrogen bond), 2511.83 cm⁻¹ (due to O–H stretching vibration), 2406.73 cm⁻¹ (due to O–H stretching vibration), 1791.55 cm⁻¹ (due to OH bending vibrations), the low intensity peak at 1300 cm⁻¹ (due to O–H, in-plane, bending) vibration), the high intensity peak at 1114.65 cm⁻¹ (due to OH–groups, in-plane deformation, bending vibration), 795.5 cm⁻¹ (due to OH bending vibration) and the peak at 458 cm⁻¹ (due to lattice vibrations involving OH⁻).
• (4)  
  High intensity peak at 486.9 cm⁻¹ refers to Fe–O stretching vibration.

Zoom In Zoom Out Reset image size

Furthermore, the SEM images (figures 4(a) and (b)) show the surface morphology of green synthesized particles and chemically synthesized particles, respectively. Figure 4 indicates a difference in the surface morphology. The surface of the green synthesized particles (figure 4(a)) is more porous than the chemically synthesized iron hydroxide particles (figure 4(b)).

Zoom In Zoom Out Reset image size

In addition, EDX spectrum (figure 5(a)) gives the elemental analysis of the green synthesized iron oxide in the metallic form. The atomic weight percentages of iron and oxygen were 30.36% and 54.86%, respectively, with a small percentage of carbon (14.79%). The obtained results confirm the presence of carbon on the surface of the green synthesized iron oxide, that comes from the phytochemicals of the pomegranate extract. Figure 5(b) also indicates that the EDXs of the green synthesized iron oxide is in the oxide form. Here, the oxidized form is Fe₂O₃ at weight percentage of 44.29%. Figure 6(a), at the mean time, represents the EDXs of the chemically synthesized particles under the same conditions. Furthermore, the atomic weight percentages were 36.56% iron, 44.33% oxygen with a small percentage of sulfur (19.11%). Meanwhile, the oxide form in figure 6(b) shows that the iron oxide was not formed, besides, sulfur was in the oxide form as SO₃. These facts reveal that the obtained particles were not iron oxide, but it is iron hydroxide particles as indicated from XRD.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Further investigation was conducted for the determination of the actual particle size of the green synthesized IONPs using HRTEM (figure 7(a)). The average particle size was about 2.7 nm. This image (figure 7(a)) shows that the green synthesized IONPs has a narrow size distribution. This reveals that such particles have disordered arrangement, and the green synthesized IONPs are amorphous. These results are compatible with the XRD results. Small nanosize particles strongly favored formation of agglomerates due to the high density number of these paricles, while larger ones stay mainly unagglomerated [27]. According to Weatherill et al [28] the iron oxide nanoparticles with small nano size have the tendency to form agglomerates. This agglomeration is caused by the magnetostatic (magnetic dipole–dipole) interactions between the iron oxide nanoparticles [29]. Using a higher magnification (figure 7(b)), it is possible to detect the lattice planes of individual particles, where the interplanar spacing is 0.35 nm.

Zoom In Zoom Out Reset image size

At equilibrium q_e (mg g⁻¹), the maximum amount of PAHs adsorped on the surface of the green synthesized IONPs was calculated according to the following equation

$$\begin{align} \newcommand{\e}{{\rm e}} \displaystyle {{q}_{{\rm e}}}=\frac{\left( {{C}_{0}}-{{C}_{{\rm e}}} \right)V}{m},\nonumber \end{align} \tag{ 1 }$$

where C₀ (mg l⁻¹) is the initial concentration of pyrene or benzo(a)pyrene, and C_e (mg l⁻¹) is the concentrations of pyrene or benzo(a)pyrene at equilibrium. V(l) is the volume of the solution containing pyrene and benzo(a)pyrene (in combination), and m(g) is the mass of the green synthesized IONPs.

The percentage of pyrene or benzo(a)pyrene removal (R) was calculated as follows

$$\begin{align} \newcommand{\e}{{\rm e}} \displaystyle R( \%)=\frac{{{C}_{0 ~ }}-{{C}_{{\rm e}}}}{{{C}_{0}}} ~ \times 100.\nonumber \end{align} \tag{ 2 }$$

3.3.1. Effect of iron oxide dose on the adsorption process.

3.3.2. Effect of pH on the adsorption process.

The pH of the tested solution plays an important role in water chemistry. It affects the properties of the solutes as well as the surface charge of the adsorbents. Figure 8 shows the effect of different pH (from 4 to 8) on the removal efficiency of pyrene and benzo(a)pyrene. At these pH levels, no change was observed in the solution, but above pH 8, the color of the solution changed into light brown indicating the dissociation of IONPs. The results (figure 8) indicate that the percentage of removal decreased as the pH increased. The optimum pH for the adsorption is 7 (i.e. the neutral pH is the optimum pH for the adsorption). This could be attributed to the variable solubility of pyrene and benzo(a)pyrene at different pH values [30].

Zoom In Zoom Out Reset image size

3.3.3. Effect of temperature on the adsorption process.

Figure 9 shows the effect of different temperatures (from 20 to 50 °C) on the adsorption of pyrene and benzo(a)pyrene. The results indicate that the adsorption decreased by increasing the temperature (figure 9). The optimum temperature was 20 °C (i.e. the adsorption can efficiently occur at room temperature).

Zoom In Zoom Out Reset image size

3.3.4. Effect of contact time on the adsorption process.

3.3.5. Effect of initial concentration of pyrene and benzo(a)pyrene on the adsorption process.

3.3.6. Effect of total dissolved solids on the adsorption process.

Total dissolved solids (TDS) plays an important role in water quality [32]. Thus, the effect of the total dissolved solids on the adsorption capacity was investigated under the optimized conditions. Figure 10 illustrates that increasing the level of NaCl increases the adsorption capacities of pyrene and benzo(a)pyrene. It was previously reported that the presence of inorganic ions (including Ca²⁺, K⁺, Na⁺, ${\rm SO}_{4}^{2-}$ and Cl⁻) binds water molecules tightly into hydration shells; thus; reduces the solubility of PAHs in water [33]. Therefore, the cavity volume that accommodates organic solute, forces the PAHs molecules onto the surface of the adsorbent [33]. This is mainly due to the fact that the presence of the inorganic salts increases the ionic strength of the aqueous solution, and affects the solubility of the organic molecules. This is known as salting-out effect (i.e. the engagement of water molecules around the ionic salt that reduces the available water and affect the dissolution of the PAHs [33].

Zoom In Zoom Out Reset image size

The values of Gibbs free energy ΔG⁰ (kJ mol⁻¹), enthalpy ΔH⁰ (kJ mol⁻¹) and entropy change ΔS⁰ (K J mol⁻¹), of the adsorption process of pyrene and benzo(a)pyrene on the surface of IONPs, were evaluated according to the following equations

$$\begin{align} \newcommand{\e}{{\rm e}} \displaystyle {{K}_{{\rm C }~ }}= ~ \frac{{{C}_{{\rm ads}}}}{{{C}_{{\rm e}}}},\nonumber \end{align} \tag{ 3 }$$

where C_ads is the concentration (mg l⁻¹) of pyrene and benzo(a)pyrene adsorbed on the surface of the IONPs at equilibrium, and C_e is the concentration (mg l⁻¹) of pyrene and benzo(a)pyrene remained in the solution at equilibrium.

$$\begin{align} \newcommand{\e}{{\rm e}} \displaystyle \Delta G_{{\rm ads}}^{{\rm 0}}= ~ -RT\ln {{K}_{{\rm C }~ }},\nonumber \end{align} \tag{ 4 }$$

$$\begin{align} \newcommand{\e}{{\rm e}} \displaystyle {\rm ln}~{{K}_{{\rm C }~ }}=\frac{\Delta S_{{\rm ads}}^{{\rm 0}}}{R}-\frac{\Delta H_{{\rm ads}}^{{\rm 0}}}{RT},\nonumber \end{align} \tag{ 5 }$$

where $\Delta H_{{\rm ads}}^{{\rm 0}}$ (K J mol⁻¹) is the standard enthalpy, and $\Delta S_{{\rm ads}}^{{\rm 0}}$ (J mol⁻¹ K⁻¹) is the entropy of the adsorption process. These values were evaluated according to equation (5). By plotting ${\rm ln}{{K}_{{\rm C}}}$ versus $1/T$ (figure 11), a straight line was obtained where $(-\Delta H_{{\rm ads}}^{{\rm 0}}/R)$ is the slope, R  =  8.314 J mol ⁻¹ K⁻¹ is the gas constant and T is the temperature (K).

Zoom In Zoom Out Reset image size

The thermodynamic parameters are given in table 4. The negative value of $\Delta G_{{\rm ads}}^{{\rm 0}}$ indicates the spontaneous nature of the adsorption process. The negative value of $\Delta H_{{\rm ads}}^{{\rm 0}}$ indicates the exothermic nature of the adsorption process. The negative value of $\Delta S_{{\rm ads}}^{{\rm 0}}$ indicates the change in the randomness at the IONPs interface during the adsorption process.

The kinetic study on the adsorption of pyrene and benzo(a)pyrene on the surface of green synthesized IONPs was investigated according to the Langmuir–Hinshelwood kinetic model [34]. Langmuir pseudo first order (equation (6)), and pseudo second order kinetic models (equation (7)) were applied to obtain the adsorption rate constants.

$$\begin{align} \newcommand{\e}{{\rm e}} \displaystyle \frac{1}{{{q}_{{\rm t}}}}= ~ \frac{{{K}_{1}}}{{{q}_{{\rm e}}} ~ t}+\frac{1}{{{q}_{{\rm e}}}},\nonumber \end{align} \tag{ 6 }$$

where q_e (mg g⁻¹) and q_t (mg g⁻¹) are the amount of pyrene and benzo(a)pyrene adsorbed on the surface of IONPs at equilibrium and t(min), respectively. K₁ (min⁻¹) is the rate constant in the pseudo-first-order adsorption process. By plotting 1/q_e against 1/t the value of K₁ and q_e can be determined

$$\begin{align} \newcommand{\e}{{\rm e}} \displaystyle \frac{1}{{{q}_{{\rm e}}}}=\frac{1}{{{K}_{2}}q_{{\rm e}}^{2}}+\left( \frac{1}{{{q}_{{\rm e}}}} \right)t,\nonumber \end{align} \tag{ 7 }$$

where K₂ (g mg⁻¹ min) is the pseudo-second-order rate constant. By plotting 1/q_e against t, the K₂ and q_e can be determined. The initial h₀ (mg g⁻¹ min) is defined as follows:

$$\begin{align} \newcommand{\e}{{\rm e}} \displaystyle {{h}_{0}}={{K}_{2 ~ }}q_{{\rm e}}^{2}.\nonumber \end{align} \tag{ 8 }$$

Figures 12(a) and (b) exhibit the pseudo-first-order plot of pyrene and benzo(a)pyrene, respectively. Meanwhile, figures 12(c) and (d) give the pseudo-second-order plot of pyrene and benzo(a)pyrene, respectively. The calculated values, of the rate constants (K₁, K₂), h_0, q_e, and the correlation coefficient of both models, are given in table 5. The pseudo-second-order model plot is linear, based on the assumption that the rate-limiting factors is chemisorption. According to Robati [35] the removal from a solution is due to physico-chemical interactions between the two phases.

Zoom In Zoom Out Reset image size

This study was carried out at different concentrations of pyrene and benzo(a)pyrene, where the equilibrium adsorption isotherm on the surface of the green synthesized IONPs was examined. This investigation was conducted under the optimized conditions of temperature, pH and the dose of the green synthesized IONPs. Figure 13 represents the isotherm plots of pyrene and benzo(a)pyren. The two isotherm models, namely, Langmuir (equation (9)) and Freundlich (equation (10)) were utilized for describing the experimental isotherm data:

$$\begin{align} \newcommand{\e}{{\rm e}} \displaystyle \frac{{{C}_{{\rm e}}}}{{{q}_{{\rm e}}}}= ~ \frac{1}{{{K}_{{\rm L}}}} ~ +\frac{{{a}_{{\rm L}}}}{{{K}_{{\rm L}}}}{{C}_{{\rm e}}},\nonumber \end{align} \tag{ 9 }$$

$$\begin{align} \newcommand{\e}{{\rm e}} \displaystyle \log {{q}_{{\rm e}}}=\log {{K}_{{\rm F}}}+~\frac{1}{n}\log {{C}_{{\rm e}}},\nonumber \end{align} \tag{ 10 }$$

where q_e (mg g⁻¹) is the maximum adsorption capacity at equilibrium, C_e (mg l⁻¹) is the residue concentration of pyrene or benzo(a)pyrene detected in the solution at equilibrium, and K_L and a_L (l g⁻¹) are Langmuir constants. In equation (10) K_F and n are the relative adsorption capacity constant and the adsorption intensity constant, respectively.

Zoom In Zoom Out Reset image size

Figures 14(a) and (b) exhibit the Langmuir isotherm model plots of pyrene and benzo(a)pyrene, respectively. Meanwhile, figures 14(c) and (d) show the Freundlich isotherm model plots of pyrene and benzo(a)pyrene, respectively. The adsorption isotherm parameters of Langmuir and Freundlich isotherm models, for the adsorption of pyrene and benzo(a)pyrene, are given in table 6. In the present study Langmuir isotherm model plot of pyrene and benzo(a)pyrene are linear. This indicates that the Langmuir isotherm model fits the adsorption process better than Freundlich isotherm model.

Zoom In Zoom Out Reset image size

It is worth mentioning that the classical Langmuir isotherm model is an equation with the best theoretical basis. It assumes that the adsorption is limited to one monolayer onto a surface containing a finite number of typical identical and homogeneous sites. It is, thus, assumed that molecules striking the surface have a given probability of adsorption. Consequently, molecules already adsorbed similarly should have a given probability of desorption. At equilibrium, therefore, equal numbers of molecules desorb and adsorb at any time. Such probabilities are related to the strength of the interaction between the adsorbent surface and the adsorbate. This classical model assumes uniform energies of adsorption onto the surface and assumes that there is no transmigration of the adsorbate on the surface plane [36, 37].

One purpose of this study is to evaluate the possibility of regenerating and reuse of the green synthesized IONPs for the adsorption of pyrene and benzo(a)pyrene. For this purpose, the used IONPs were thoroughly washed with methanol. Methanol was selected due to the following advantages:

• Methanol solubility of pyrene and benzo(a)pyrene are 1 mg ml⁻¹ and 0.1 mg ml⁻¹, respectively.
• It is non toxic compared to other solvents (i.e. benzene, toluene, chloroform, etc).
• It is easily removed by rotary evaporator.

The capacities of the adsorption process of pyrene and benzo(a)pyrene by the regenerated IONPs, during the five times of regeneration and reuse, are shown in figure 15. It was noticed that there is no remarkable decrease in the adsorption capacity of the regenerated IONPs toward pyrene and benzo(a)pyrene up to five cycles of regeneration and reuse (i.e. the iron oxide nanoparticles are a potential candidate for the adsorption of pyrene and benzo(a)pyrene even for five cycles). This indicates the high applicability of the reuse of this green synthesized IONPs.

Zoom In Zoom Out Reset image size

Table 7 shows that the green synthesized IONPs are efficient material for the adsorption of pyrene and benzo(a)pyrene from aqueous media and offers significant advantages over many of those previously described using traditional adsorbents.

Table 8 gives the physical and chemical characteristics of the artificial contaminated tap water with the studied PAHs. Table 9 illustrated the results of the removal of pyrene and benzo(a)pyrene by using different doses of IONPs namely 90, 110 and 130 mg l⁻¹, where the contact time was 150 min. The results showed that removal rate of pyrene and benzo(a)pyrene varied according to the implemented dose of the IONPs. These results indicated that the use of 110 mg l⁻¹ of IONPs decreased the concentration of pyrene and benzo(a)pyrene from 100 and 1 µg l⁻¹ to 1.7 and 0.02 µg l⁻¹, respectively, and the corresponding rate of removal reached 98.3 and 98.0%, respectively. Also the use of 130 mg l⁻¹ of IONPs decreased the concentration of pyrene and benzo(a)pyrene from 100 and 1 µg l⁻¹ to 1.5 and 0.01 µg l⁻¹, respectively, and the corresponding rate of removal reached 98.5 and 99.0%, respectively. Therefore, 130 mg l⁻¹ of IONPs can be used as the optimum dose for the removal of pyrene and benzo(a)pyrene from drinking water. Thus it can be concluded that IONPs can be used to eliminate the PAHs in drinking contaminated water and can be employed in the dringking water treatment plants (DWTP) as a polishing tool.