3.1. Characterization study
Figure 1 shows the FT-IR spectra of Fe3O4 (a), Fe3O4@SiO2 (b), Fe3O4@SiO2-Cl (c), Fe3O4@SiO2(OH)2 (d), Fe3O4@SiO2-TCT (e) and Fe3O4@SiO2-TCT-PVA (f) in the range from 500 to 3700 cm− 1. The presence of vibration band at 570 cm− 1, being due to stretching Fe-O demonstrates the presence of Fe3O4(Fig. 1a) (Dindarloo Inaloo et al. 2020c). In the FT-IR spectrum of MNPs, the absorption bands around 3400 cm− 1 is related to the stretching vibrations of hydroxyl group which absorbed by MNPs. Moreover, there is a peak at 1620 cm− 1 is assigned to the hydroxyl bending group (Fig. 1a) (Sardarian et al. 2018). In the spectrum of Fe3O4@SiO2, there is a characteristic absorption bands at 1100 and 850 cm− 1which are attributed to the asymmetric and symmetric stretching of the Si-O-Si group, respectively (Fig. 1b). According to these data, the SiO2 group has been successfully coated on the surface of Fe3O4 MNPs. The IR spectrum of 3-chloromethoxypropylsilane functionalized Fe3O4@SiO2 indicated absorption peaks at 2954, 1444 and 702 cm− 1 related to the vibration of C-H (stretching), CH2 (bending) and C-Cl (stretching) bonds, respectively (Fig. 1c) (Dindarloo Inaloo et al. 2020b). FTIR spectra of Fe3O4@SiO2(OH)2 indicate a new absorption band at 1354 cm− 1 (elimination of the C-Cl peak) for the C-N band (Fig. 1d). The resulted compound reacted with cyanuric chloride to give TCT functionalized Fe3O4@SiO2(OH)2 MNPs. In the FTIR spectrum of Fe3O4@SiO2-TCT, absorption bands at 1512, 1553, and 1724 cm− 1 are attributed to C = N bond and the peaks observe at 2781–2978 and 1481 cm− 1 correspond to the stretching bond of C–H and bending bond of CH2 (Fig. 1e). The peak at 1100 cm− 1 can be assigned to C-Cl stretching of cyanuric chloride which overlaps with the vibrating band of Si-O-Si (stretching) (Fig. 1e). Figure 1f is related to the FT-IR spectrum of Fe3O4@SiO2-TCT-PVA NPs. As it is indicated in the Fig. 1f., the peaks at 2931–2970, 1471, 1085 and 573 cm− 1 are related to C-H (stretching vibration), CH2 (bending), Si-O-Si (asymmetric stretching) and Fe-O (stretching vibration), respectively. On the other hand, the weak peaks at 1724, 1543 and 1512 cm− 1 are due to the vibration mode of C = N in the cyanuric chloride in the adsorbent structure (Fig. 1f). These bands prove the modification of surface of Fe3O4@SiO2 NPs with organic layers.
The investigation of samples structure is performed by using powder X-ray diffraction (XRD) method. As it is observed from the XRD data, the standard Fe3O4 crystal with spinel structure icludes six diffraction peaks at 2θ = 30.1o, 35.4o, 43.1o, 53.4o, 57o and 62.6o which are assigned to the (220), (311), (400), (422), (511), and (440) planes (Fig. 2a). These peaks are indexed to the inverse cubic spinel structure of Fe3O4 (JCPDS card no. 19-0629) (Sardarian et al. 2021), were also observed for Fe3O4@SiO2 and Fe3O4@SiO2-TCT-PVA nanoparticles (Fig. 2b, c). After being coated with SiO2, the intensity of Fe3O4 phase decreased and the phase composition of MNPs remained intact after surface modification on nanoparticles (Fig. 2b, c). Moreover, the broad peak from 2θ = 10-20o is proved an amorphous silica phase in the shell of the silica-coated Fe3O4 nanoparticles. In the structure of Fe3O4@SiO2-TCT-PVA NPs, the broad peak is shifted to lower angles as a result of the synergetic effect of amorphous silica and polyvinyl alcohol polymer (Fig. 2c).
The investigation of the size, shape and morphology of the nanocomposites are carried out using transmission electron microscopy (TEM) and Field emission scanning electron microscopy (FE-SEM) methods (Fig. 3). As it is indicated in Fig. 3a, the Fe3O4 seeds are nearly spherical in shape and average grain size of < 20 nm. Figure 3b,c indicates the FE-SEM images of the Fe3O4@SiO2 and Fe3O4@SiO2-TCT-PVA NPs. Clearly, the Fe3O4@SiO2 are 15–25 nm and spherical, while the later-obtained Fe3O4@SiO2-TCT-PVA NPs are a larger 35–55 nm. Furthermore, the TEM image discloses the nearly spherical shape of the Fe3O4 MNPs in average size of 12 nm. The size and shape of Fe3O4@SiO2-TCT-PVA NPs are in agreement with the values calculated from XRD analysis (Fig. 3d).
Furthermore, Fig. 3e and f indicates the TEM images of Fe3O4@SiO2 and Fe3O4@SiO2-TCT-PVA. The results prove the spherically shaped of Nps which is proved the formation of uniformly distributed nanosized particles with an average diameter of 20 and 45 nm, respectively. In order to extend the characterization of the adsorbent, DLS analysis of MNPs was performed. Figure 3g-i shows the size distributions of Fe3O4, Fe3O4@SiO2 and Fe3O4@SiO2-TCT-PVA which is in the range of 8–16 nm, 16–24 nm and 38–52 nm, respectively.
The thermal stability of the Fe3O4@SiO2-TCT-PVA nanoparticles are explored via TGA assay (Fig. 4A). As illustrated in Fig. 4A, two weight loss steps is observable in the TGA plot. The first step occurs between the temperatures of 100 and 200 oC and it is related to the evaporation of residual organic solvent, adsorbed water molecules and loss of surface hydroxyl moieties of the nano adsorbent. The second weigh loss appears after 200°C, which is correspond to the elimination of organic part which is supported on the Fe3O4@SiO2 NPs. The TGA curves of the Fe3O4@SiO2, Fe3O4@SiO2-Cl, Fe3O4@SiO2-TCT and Fe3O4@SiO2-TCT-PVA NPs are illustrated in Fig. 4A with the weight losses of 12.1%, 24.3%, and 52.7%, respectively in the range of 150–700 oC.
Energy dispersive X-ray spectroscopy (EDX) prove powerful tool for accomplishing the elemental analysis of various samples. The EDX spectrum of Fe3O4@SiO2-TCT-PVA indicates all predicted elements in the adsorbent structure, such as carbon, nitrogen, oxygen, silicon and iron as indicated in Fig. 4B. The magnetic properties of nanoparticles are evaluated at room temperature using VSM in an external magnetic field within the range of -6 kOe to + 6 kOe. As it is observed from Fig. 4C, the Fe3O4 and Fe3O4@SiO2-TCT-PVA possess magnetization saturation values of 64.8 and 25.9 emu g− 1, respectively (Fig. 4Ca,b). Neither remanence nor coercivity is observed which is proved that these particles are superparamagnetic.The immobilizing of magnetite nanoparticles resulted in the decrease in the magnetic strength of the composite on account of the weight contribution from the nonmagnetic portion (Fig. 4Cb). As it is observed from these datasets, the magnetization saturation value of Fe3O4@SiO2-TCT-PVA in compared with Fe3O4 is lower due to the grafting of organic compounds. However, the Fe3O4@SiO2-TCT-PVA structure still indicates strong magnetization which is provided an easy and efficient route to separate nanoparticles from a suspension system under an external magnetic field. The magnetic separation is quickly fulfilled by supplying an external magnetic field. Moreover, there is a possibility to re-disperse nanoparticles quickly with a slight shake once the magnetic field is removed. (Fig. 4D)
3.2. Selection of materials
The adsorbent type plays a crucial role on the removal efficiency of metal ions. Accordingly, Fe3O4@SiO2 core-shell nanoparticles functionalized with polyvinyl alcohol was utilized for the removal of Cu(П) and Cd(П) to benefit from the excellent magnetic property of magnetite NPs as well as high adsorption capability of polyvinyl alcohol. To protect the Fe3O4 NPs in strong acidic medium, which is need for desorption of target metal ions, the surface of the Fe3O4 NPs was functionalized with the TEOS as silanizing agent. It is worth noting that silica layer (SiO2) is stable in acidic strong media. After that, the surface of Fe3O4@SiO2 core-shell nanoparticles was functionalized with polyvinyl alcohol, which has oxygen donor atoms and can enhance the adsorption capacity of the sorbent towards the target ions, as well as improve the dispersibility of the adsorbent in the aquatic extraction medium.
3.3. Optimization
3.3.1. Effect of nano-adsorbent amount
The adsorbent amount exhibits a profound effect on the metal adsorption (Ebrahimzadeh et al. 2012). In this way, various amounts of the nano-adsorbent (2.5–20 mg) were added to 50 mL of the target metal solutions. As Fig. 5a illustrates, Fe3O4@SiO2-TCT-PVA exhibits a positive effect in the adsorption% of Cu(II) and Cd(II) ions up to 15 mg and at the higher dosages (> 15 mg) more improvement was not observed in the adsorption efficiency. Accordingly, 15 mg of Fe3O4@SiO2-TCT-PVA MNPs in 50 mL sample solution was selected as the optimum nano-adsorbent dose.
3.3.2. Effect of pH
The pH as an important variable affects the adsorption efficiency of metal ions strongly. Accordingly, the pH of sample solution was varied within the range of 4 to 8, and the best adsorption performance was observed at pH 7 (Fig. 5b). At the low pH values, the active sites of Fe3O4@SiO2-TCT-PVA MNPs are protonated mostly as a result, protonated sites are not able to coordinate to Cu(II) and Cd(II) ions owing to the electrostatic repulsion between the positive moieties of the Fe3O4@SiO2-TCT-PVA nano-sorbent and the target ions (Jalilian et al. 2017).
3.3.3. Effect of the initial concentration of metal ions
The effect of initial concentration of Cu(II) and Cd(II) in the range of 0.1–0.6 mmol/L is evaluated on the adsorption percentage. For this purpose, the experiments were carried out by adding 15 mg of the nano-adsorbent to 50 mL solutions at pH 7.0 with a contact time of 25 min. According to the Fig. 5c, the highest removal efficiency was obtained at the concentration of 0.5 mmol/L, and 0.4 mmol/L, for Cu(II) and Cd(II), respectively. The improvement in the efficiency of adsorption are explained by the increase of mass gradient in between the metal solutions and Fe3O4@SiO2-TCT-PVA nano-sorbent as the initial metal concentration increases (Soni et al. 2020).
3.3.4. Effect of adsorption time
Effect of adsorption time was investigated within the range of 5 to 30 min. for this purpose, 15 mg of the nano-adsorbent were added to 50 mL sample solution at pH 7.0 with an initial concentration of 0.5 mmol/L, and 0.4 mmol/L, for Cu(II) and Cd(II), respectively. As it is indicated inn Fig. 5d, the optimum adsorption percentage was observed at the time of 20 and 25 min, for Cd(II), and Cu(II) ions, respectively. As Fig. 5d illustrates, the adsorption efficiency increases with increasing the contact time. The obtained curve exhibits a rapid adsorption rate at the first step due to the unoccupied active sites of the adsorbent and then slows down owing to the occupation of the active site of the nano-adsorbent by the metal ions (Soni et al. 2020).
3.4. Adsorption isotherms
Adsorption isotherms of target ions in the range of 25–45 oC are exhibited in Fig. 6a-b. As shown in the figure, the equilibrium adsorption capacity (qe) increases up to 35 oC. The adsorption data were fitted to the linear form of Langmuir and Freundlich models. Table 1 summarizes the Langmuir and Freundlich adsorption data. The obtained correlation coefficients approve that the adsorption process obey from the Langmuir isotherm rather than the Freundlich isotherm.
The calculated qm at 25 oC is 0.84, and 1.48 mmol/g for Cd(II), and Cu(II), respectively. The maximum adsorbed amount of the target ions increased from 1.48 mmol/g to 1.57 mmol/g, and 0.84 mmol/g to 0.99 mmol/g by increasing the temperature from 25 oC to 35 oC for Cu(II), and Cd(II), respectively. The high temperature result in extend the pore volume and surface area of the nano-adsorbent. This phenomena provides more possibility for metal ions to pass through the external boundary layer which leads to penetrate it more easily (Esmaeilpour et al. 2022). Moreover, the increase of extraction temperature improves the mass transfer of the target ions to the active sites of Fe3O4@SiO2-TCT-PVA MNPs. Furthermore, the increase of extraction temperature accelerates the complex formation between the heteroatoms and metal ions. The decrease of the adsorption after 35 oC can be attributed to the exothermic nature of adsorption process. Therefore, the adsorption order of the target ions is as follows: Cu(II) > Cd(II).
Table 1
Freundlich and Langmuir isotherm model parameters.
Temperature (oC)
|
Metal ion
|
Langmuir model
|
Freundlich model
|
qm (mmol/g)
|
KL (L/mmol)
|
r
|
n
|
KF
(mmol/g)
|
r
|
25
|
Cu(II)
|
1.48
|
2246
|
0.9999
|
13.3
|
1.72
|
0.6474
|
Cd(II)
|
0.84
|
182
|
0.9983
|
6.0
|
1.14
|
0.6253
|
35
|
Cu(II)
|
1.57
|
7937
|
0.9999
|
14.5
|
1.23
|
0.5179
|
Cd(II)
|
0.99
|
633
|
0.9998
|
7.6
|
1.28
|
0.7847
|
45
|
Cu(II)
|
1.37
|
1457
|
0.9998
|
12.5
|
1.60
|
0.5405
|
Cd(II)
|
0.89
|
796
|
0.9997
|
7.4
|
1.20
|
0.7179
|
3.4. Adsorption kinetic
The kinetic study of the adsorption for target metal ions in the Fe3O4@SiO2-TCT-PVA active sites was explored by fitting the experimental data to the Lagergren pseudo first-order and pseudo second-order model (Table 2). The assay was performed at various time intervals (0–25 min) and 35 oC using an initial concentration of 0.2 mmol L− 1 for both the metal ions. Table 2 depicts the rate constant values and correlation coefficients (r) from the assays and also the counterpart values of calculated from the mentioned models. The values computed from the pseudo second-order model are closer to the experimental data in compared with the values of pseudo first-order model. Moreover, the pseudo-second order kinetic model indicates a very high value of “r” for all kinetic data. These observations confirm that the adsorption kinetic obey from the pseudo second-order process. The pseudo second-order model occur in three steps including: 1) surface adsorption, 2) external liquid film diffusion, and 3) intra-particle diffusion.
Table 2
Kinetic parameters for the adsorption of target metal ions into Fe3O4@SiO2-TCT-PVA MNPs.
Ion
|
Experiment
|
Pseudo first-order model
|
Pseudo second-order model
|
qe (mmol/g)
|
qe
(mmol/g)
|
k1
(min− 1)
|
r
|
qe (mmol/g)
|
k2 (g/mmol.min)
|
r
|
Cu(II)
|
1.44
|
0.2399
|
0.2006
|
0.9802
|
1.4497
|
4.0154
|
0.9992
|
Cd(II)
|
0.93
|
0.4291
|
0.0998
|
0.9743
|
0.9332
|
0.7406
|
0.9976
|
3.5. Durability and stability assays
The regeneration of a nano-adsorbent is one of the vital keys for its commercial applications. The main route for desorption of heavy metal ions from active sites of adsorbents is its elution with HCl solutions in the range of 0.01-1.0 mol/L. To evaluate the stability of nano-adsorbent in the acidic solution before the regeneration process, Fe3O4@SiO2-TCT-PVA MNPs was treated in acidic conditions and the leaching content of iron ions was evaluated (Table 3).
Table 3
Monitoring of the leached iron ions after treatment the various adsorbents saturated with HCl solutions for 24 h.
HCl solution concentration (M)
|
Leached Fe content (%)
|
Fe3O4
|
Fe3O4@SiO2
|
Fe3O4@SiO2-TCT-PVA
|
0.01
|
1.79
|
0.28
|
0.09
|
0.05
|
4.67
|
0.52
|
0.17
|
0.1
|
13.78
|
1.15
|
0.35
|
0.5
|
84.35
|
3.13
|
0.70
|
1.0
|
94.54
|
4.89
|
1.01
|
As depicted in Table 3, the leaching percentage for Fe3O4, and Fe3O4@SiO2, is 94.54% and 4.89% Fe, respectively, after suspending the material in 1.0 mol/L HCl solution for 24 h. This value is 1.01% for the Fe3O4@SiO2-TCT-PVA nanomaterials that affirms its high stability under the harsh acidic conditions (Table 3). Accordingly, 1.0 mol/L HCl solution was employed for desorption of Cu(П), and Cd(П), ions, and the results are illustrated in Fig. 7a. As Fig. 7a depicts, a small loss in the adsorption efficiency is observable after five consecutive adsorption-desorption cycles, which confirms the good performance and recyclability of the Fe3O4@SiO2-TCT-PVA nanomaterials. Figure 7b illustrates the FESEM micrograph of Fe3O4@SiO2-TCT-PVA nanomaterials after 5 adsorption-desorption assays and clarifies no remarkable change in the morphology of functionalized MNPs, though, a bit aggregation is unavoidable. In addition, DLS measurements were carried out to measure the particles size after performing the 5 adsorption-desorption assays. Figure 7c exhibits that this size distribution is centered at a value of 70 nm. These observations exhibit the excellent chemical stability and recyclability of the fabricated nanomaterial.