3.1. Characterizations of the Prepared Activated Carbon
The textural characteristics of GAC and Fe-doped GACs are illustrated in
Figure 2a,b, respectively. According to IUPAC categorization, the N
2 adsorption/desorption isotherm of GAC and Fe–GAC samples can be categorized into Type IV isotherm with H4 hysteresis loop. Type IV isotherm indicates the mesoporous structure of GAC, initially starting with monolayer and multilayer adsorption followed by pore condensation (see
Figure 2a). The specific surface areas of raw GAC, Fe–GAC 1, Fe–GAC 5, and Fe–GAC 20 were found to be 829.6, 831.9, 848.2, and 750.0 m
2 g
−1, respectively. The increased specific surface area of Fe–GAC 1 and Fe–GAC 5 is attributed to the contribution of Fe
2O
3 nanoparticles onto the GAC surface. It was also noticed that at higher loadings of Fe
2O
3 (Fe–GAC 20), the specific surface area decreased due to possible pore blockage via the exceeded amount of Fe
2O
3 nanoparticles [
27]. The BJH pore size distribution of raw and modified GACs is presented in
Figure 2b. The average pore sizes of raw GAC, Fe–GAC 1, Fe–GAC 5, and Fe–GAC 20 were found to be 2.31 nm, 2.40 nm, 2.37 nm, and 2.36 nm, respectively, confirming the presence of abundant mesoporous.
The XRD patterns (
Figure 2c) of raw GAC and Fe–GAC samples demonstrated crystalline peaks for GAC and Fe
2O
3 nanoparticles. The peaks located at 21.44° (002) and 26.74° (002) in GAC correspond to the graphitic carbon structures, typical for all carbon materials. For Fe–GAC samples, new characteristic diffraction peaks originated at 2θ of 33.1° (104), 35.54° (311), and 54.00° (422), respectively. These peak positions validated the formation of Fe
2O
3 nanoparticles on the surface of GAC.
Raman spectroscopy was used to investigate the ordered and disordered crystal structures of raw GAC and Fe–GAC samples, and the outcomes are clarified in
Figure 2d. All the samples possessed two characteristic peaks of carbon materials denoted as D-band at 1337.20 cm
−1 and G-band at 1598.71 cm
−1, respectively. The G-band is linked with the stretching vibrations of the sp
2-bonded pairs, suggesting full graphitization of the GAC [
28]. On the other hand, the D-band is related to the sp
3 defect sites, which indicates the existence of disorder in the structure of the carbon material [
29]. According to
Figure 2d, the intensities of D- and G-bands were seen to decrease with an increase in the Fe
2O
3 loadings, suggesting the successful doping of Fe
2O
3 nanoparticles on the GAC surface.
The FTIR spectra of GAC and Fe–GAC materials were inspected in the range of 4000–400 cm
−1 and the outcomes are displayed in
Figure 3. The peak at 3475 cm
−1 signifies the presence of O-H stretching vibration corresponding to a hydroxyl functional group. The intensity of the -OH peak was seen to significantly improve with the Fe
2O
3 loadings, suggesting an increased number of O-H functional groups after the modification of GAC. The characteristic absorption peak at 2849 cm
−1 can be assigned to an O-H stretching vibration. Fe–GAC materials exhibited a small peak at 1647 cm
−1 that corresponds to the presence of either C=C or C=O stretching vibrations. A significantly broad peak at 1424 cm
−1 was observed in Fe–GAC 20 and can be given to O-H bending vibration from either carboxylic acid or alcohol. Furthermore, the broad band at 1080 cm
−1 can be given to C-H and C=C bending vibrations. The band at 665 cm
−1 proves the successful doping of Fe
2O
3 nanoparticles on the GAC surface due to the existence of the FeOOH functional group. The presence of similar absorption peaks in raw GAC and Fe–GAC materials suggests that the GAC preserved its structure after modification by creating more functional groups on the surface. Furthermore, similar trends were observed in the literature suggesting the destruction of organic structures and the addition of more acidic groups on the GAC surface after modification [
30]. Therefore, the presence of more acidic groups enhances the number of active binding sites, which in turn might offer more adsorption sites for the binding of Pb(II) and Cr(T) ions.
The morphological properties and chemical composition of raw GAC and Fe–GAC samples were studied using FE-SEM and EDS, respectively, and the outcomes are demonstrated in
Figure 4a–d. Raw GAC (
Figure 4a) showed a porous morphology with a wide range of mesopores and macropores. Fe–GAC materials in
Figure 4b–d clearly showed a well-defined distribution of Fe
2O
3 nanoparticles on the surface of the GAC with an average particle diameter ranging between 100–400 nm. EDS results displayed in
Table 1 illustrate that raw GAC was primarily composed of trace amounts of silicon (Si), aluminum (Al), and iron (Fe) as impurities
. It was noticed that after the loading of Fe
2O
3, the average mass of C decreased with increasing Fe
2O
3 loading, whereas corresponding masses of Fe increased to 1.84, 6.04, and 21.90 for Fe–GAC 1, Fe–GAC 5, and Fe–GAC 20, respectively.
3.2. Effect of Fe2O3 Loading on the Elimination of Pb(II) and Cr(T) Ions
Screening tests were performed to illustrate the ideal Fe
2O
3 loading for the adsorption of Pb(II) and Cr(T) ions from water. The adsorption results presented in
Figure 5 demonstrated an increase in the adsorption uptake of Pb(II) and Cr(T) by raising the Fe
2O
3 loading from 1 to 5 wt.%. A further increase in the Fe
2O
3 loading to 20 wt.% led to a drastic decrease in the adsorption uptake from 9.1 mg g
−1 to 3.1 mg g
−1 and 9.9 mg g
−1 to 7.8 mg g
−1 towards Pb(II) and Cr(T) ions, respectively, which is lower than that of raw GAC (i.e., 5.9 mg g
−1 for Pb(II) and 8.8 mg g
−1 for Cr(T). The reduction in the adsorption uptake of Fe–GAC 20 can be explained by the possible agglomeration of the Fe
2O
3 nanoparticles on the surface of the GAC, as seen in the SEM micrograph in
Figure 4d. Similar outcomes were gained by Monika Jain et al., who reported that more surface-free binding sites are covered with increased metal ion concentration, causing a decline in the adsorption uptake [
31]. The adsorption outcomes illustrated that Fe–GAC 5 (5 wt.%) showed good efficiency for both Pb(II) and Cr(T) ions as compared to raw GAC, Fe–GAC 1, and Fe–GAC 20. The adsorption uptake of Fe–GAC was highly dependent on the specific surface area and the existence of mesoporous structure. The higher adsorption uptake of Fe–GAC 5 accounted for the highest textural characteristics, which provided more adsorption sites for binding Pb(II) and Cr(T) ions. It is worth mentioning that doping Fe
2O
3 nanoparticles did improve the adsorption uptake of GAC, but to a certain limit, above which it exhibited a negative effect. Therefore, Fe–GAC 5 was picked up as the optimum modified GAC for subsequent studies.
3.3. Effect of Fe–GAC 5 Dose
The impact of Fe–GAC 5 amount on the elimination efficiency and adsorption uptake of Pb(II) and Cr(T) ions was investigated by fluctuating the Fe–GAC 5 dosage from 1 to 5 g L
−1 and the outcomes are demonstrated in
Figure 6a,b. The results reveal an increase in the elimination efficiency with raising Fe–GAC 5 dosage. For example, the elimination efficiency was found to increase from 53.3% to 90.8% for Pb(II) and from 74.8% to 98.2% for Cr(T) with increasing Fe–GAC dosage from 1 to 5 g L
−1. However, further increasing the Fe–GAC 5 dosage from 3 to 5 g L
−1 was found to have an insignificant effect on the elimination efficiency. The increase in the elimination efficiency with Fe–GAC 5 dosage is attributed to the accessibility of more Fe–GAC 5 surface area and the availability of active adsorption sites alongside fixed concentrations of Pb(II) and Cr(T) ions in the water. On the other hand, the adsorption uptake was found to decrease with the Fe–GAC 5 dosage, most probably due to the opposite relationship between the adsorption uptake and the Fe–GAC 5 mass. Typically, at higher doses, a larger Fe–GAC 5 surface area becomes exposed for the adsorption of Pb(II) and Cr(T) ions, which results in more competition for the ions to fill the active sites. This behavior causes many active sites on the Fe–GAC 5 surface to remain unsaturated, which therefore leads to a decreased adsorption uptake [
32]. Similar trends were stated in the literature for the remediation of Cr(T) and Pb(II) ions using paper mill sludge and date pit ACs, respectively [
33,
34]. In this study, an Fe–GAC 5 dosage of 1.75 g L
−1 is recommended for the rejection of both Pb(II) and Cr(T) ions from water and has been used for the remaining experiments.
3.4. Effect of pHi
The pH
i of the solution plays a vibrant role in the rejection of pollutants due to the fact that the pH does not only change with the nature of the pollutant species but also affects the adsorbents’ surface charge in the liquid phase. At first, the pH
ZPC of raw GAC and Fe–GAC 5 were studied using the pH drift method and the results are displayed in
Figure S1. The pH
ZPC is the pH point where the charge on the adsorbent surface is neutral (zero net charge). It was noticed that the surface of raw GAC was almost neutral, holding a pH
ZPC value of 7.2 while the pH
ZPC of Fe–GAC 5 was found to be 5.6. The influence of pH
i on the adsorption uptake of Pb(II) and Cr(T) ions from water using Fe–GAC 5 is displayed in
Figure 7. The outcomes clearly show an increase in Pb(II) adsorption uptake by raising the pH
i from 2.1 to 6.1. After that, the adsorption uptake was found to decrease at pH 8.2. At pH 2.1, the Fe–GAC 5 and Pb(II) ions are both holding positive charges, resulting in low adsorption uptake due to the electrostatic repulsions between the Fe–GAC 5 and the adsorbate. Raising the pH
i from 2.1 to 5.6 undoubtedly decreased the magnitude of electrostatic repulsions, which in turn resulted in an improved adsorption uptake from 3.9 to 5.1 mg g
−1. The maximum adsorption uptake of 5.6 mg g
−1 was recorded at pH 6.1. At this pH, the Fe–GAC 5 is negatively charged since the solution pH exceeds the pH
ZPC (5.6) and Pb(II) ions have a positive charge, hence the electrostatic repulsions no longer exist, and the adsorption process is dominated by the electrostatic interactions. However, at pH 8.2, a high concentration of hydroxyl ions (OH
−) competes with Pb(II) ions for the adsorption sites causing Pb(II) ions to precipitate, decreasing the adsorption uptake to 4.8 mg g
−1 [
35,
36].
For Cr(T) removal (see
Figure 7), a decrease in the adsorption uptake with increasing pH values was observed, indicating more adsorption uptake at low pH values. It is worth mentioning that Cr(T) ions generated from potassium dichromate salt exist in different forms (CrO
42−, Cr
2O
72− and HCrO
4−) by changing the pH
i from 2.0 to 12.0, implying negatively charged Cr(T) ions within the pH range 2 to 12. The higher adsorption uptake at pH 2.1 was ascribed to the electrostatic interactions among the negatively charged Cr(T) ions and the positively charged Fe–GAC 5. The higher number of positive charges on the Fe–GAC 5 at pH 2.1 is generally accredited to the protonation of the functional groups on the Fe–GAC 5 surface owing to the high concentration of H
+ ions. Moreover, the reduction of Cr(III) to Cr(VI) is more likely to take place at low pH such as pH 2.1, resulting in an improved adsorption uptake. Similar trends were observed by Yang et al. and Ihsanullah et al. [
37,
38]. It was also reported that Equation (8) is feasible to explain the higher adsorption uptake under acidic conditions.
Raising the pHi from 2.1 to 5.6 was found to decrease the adsorption uptake from 8.24 to 5.8 mg g−1. This is credited to the decline of the extent of electrostatic interactions between Cr(T) ions and Fe–GAC 5 surface. Further increasing the solution pH, higher than the pHZPC (>5.6) resulted in a negatively charged Fe–GAC 5. Moreover, the Cr(T) ions generated from potassium dichromate salt are available in water as CrO42− and Cr2O72−. As such, the adsorption uptake was found to decrease to 3.8 mg g−1 at pH 8.2 owing to the electrostatic repulsions. Further increase in the solution pH from 9.8 to 11.8 was found to decrease the adsorption uptake from 3.47 to 2.2 mg g−1. Typically, increasing the solution alkalinity would increase the magnitude of the negative charges on the Fe–GAC 5 surface, resulting in more electrostatic repulsions with the negatively charged Cr(T) ions and hence decreasing the adsorption capacity. Moreover, a high concentration of OH− ions would compete with Cr(T) ions on the accessible adsorption sites subsequent to a further decrease in the adsorption uptake.
3.5. Effect of Contact Time and Adsorption Kinetics
The impact of
T on the adsorption uptake of Pb(II) and Cr(T) ions onto the Fe–GAC 5 is displayed in
Figure 8. As depicted, rapid adsorption had taken place in the first 1 h, then a slower rate of adsorption was observed until achieving equilibrium after 24 h. At the initial adsorption stages, the existence of high adsorbate concentration in the solution led to the diffusion of the ionic species from the bulk to the surface of the Fe–GAC 5, explaining the rapid increase in the adsorption uptake in the first 1 h. With increasing
T, the surface adsorption binding site on the Fe–GAC 5 gets occupied with time, causing less improvement in the adsorption uptake as no more sites are available for binding. However, the adsorption uptake for Pb(II) and Cr(T) ions increased from 12.2 to 15.6 mg g
−1 and from 9.4 to 12.7 mg g
−1 by rising the
T from 60 min to 1260 min, respectively. This slow improvement in the adsorption uptake is attributed to the diffusion of Pb(II) and Cr(T) ions within the internal micropores and mesoporous of the Fe–GAC 5. Once equilibrium had been established, the adsorption sites were less likely to be occupied and the adsorption uptake gradually balanced after 24 h of adsorption.
Data generated from the kinetic adsorption studies for Pb(II) and Cr(T) ions were fitted to several kinetic models.
Table 2 shows all calculated kinetic model factors and correlation factor (R
2) for the tested models. The pseudo-first order model (see
Figure S2a) for Pb(II) and Cr(T) removal exhibited low values of R
2 (<0.97). Moreover, the q
e value obtained from the model did not match or fit the experimental q
e value, implying that the pseudo-first order model does not define the elimination of Pb(II) and Cr(T) ions by Fe–GAC 5. The pseudo-second order model (see
Figure S2b for the trend line fits) demonstrated an R
2 value > 0.99 for Pb(II) and Cr(T) ions. Additionally, the calculated q
e value from the model was seen to approach the experimental q
e for Pb(II) and Cr(T) ions onto the Fe–GAC 5. The pseudo-second order rate constant K
2 was found to be higher for Pb(II) adsorption than Cr(T), suggesting that Pb(II) was adsorbed at a faster rate by Fe–GAC 5 than Cr(T) ions. These outcomes validated that both Pb(II) and Cr(T) kinetics obeyed the pseudo-second order model.
The intra-particle diffusion model (see
Figure S2c) was applied to inspect the mechanism of the adsorption process. Typically, the intra-particle diffusion controls the adsorption process if the trend line between q
t and t
1/2 forms a straight line and passes through the origin. The R
2 value generated from the intra-particle diffusion model fit was found to be 0.77 and 0.85 for Pb(II) and Cr(T), respectively, suggesting that the adsorption of both ionic species onto the Fe–GAC 5 was controlled by different steps (i.e., intra-particle diffusion and surface film diffusion). The adsorption of Pb(II) and Cr(T) ions onto the Fe–GAC 5 initially took part at the external surface of the Fe–GAC 5, which then diffused into the inner porous structure until equilibrium was reached.
The Elovich kinetic model (see
Figure S2d) was found to produce an R
2 value = 1, indicating the compliance between the kinetic data and the Elovich model. Moreover, the Elovich model stands for the heterogeneous surfaces and chemisorption process. Accordingly, the kinetic results generated from this work suggest a chemisorption process for the elimination of Pb(II) and Cr(T) ions onto the heterogeneous adsorption sites on the Fe–GAC 5 surface. However, the adsorption mechanism will be further investigated in the following sections.
3.6. Adsorption Isotherm
The isotherm curves optimized by the Excel solver are displayed in
Figure 9a,b.
Table 3 displays the model parameters and statistical evaluations (SSE and R
2) of each fitted model. According to
Table 3, Sips (R
2 = 0.96; SSE = 7.10) and Langmuir (R
2 = 0.95; SSE = 8.67) models best fitted the equilibrium data for Cr(T) removal, proposing a monolayer coverage of the Cr(T) ions onto the surface of the Fe–GAC 5 with some likely heterogeneity in the adsorption sites. For Pb(II) removal, all isotherm models were in agreement with the equilibrium data representing R
2 values greater than 0.99 and SSE values less than or equal to 1.5, suggesting different removal mechanisms in the adsorption process. Accordingly, the agreement of different models for the elimination of Pb(II) and Cr(T) ions suggests a combination of chemical and physical adsorption processes on the Fe–GAC 5 surface. The Freundlich coefficient n
F values for both Pb(II) and Cr(T) ion adsorption were found to be greater than 1, signifying favorable removal of Pb(II) and Cr(T) ions by Fe–GAC 5. Typically, the value 1/n
s in the sips model gives information on the homogeneity (1/n
s ≈ 1) and heterogeneity (1/n
s > 1) of the active adsorbent sites. The values of 1/n
s obtained from this study indicated the heterogeneous and homogeneous adsorption sites of the Fe–GAC 5 towards Pb(II) and Cr(T) ions. The maximum adsorption capacities obtained from the Langmuir equation were found to be 11.9 and 22.1 mg g
−1 for Pb(II) and Cr(T) by Fe–GAC 5, respectively.
Table 4 displays a comparison between our Fe–GAC 5 and other carbon-based materials in terms of preparation technique, experimental conditions, and corresponding adsorption capacities for the rejection of Pb(II) and Cr(T) ions from water. It can be seen that the Fe–GAC 5 illustrated one of the best-stated adsorption uptakes for Pb(II) ions compared to other carbonaceous materials. Moreover, the adsorption uptake of Cr(T) onto the Fe–GAC 5 is also comparable to the adsorption uptakes onto other adsorbents.
3.10. Removal of Pb(II) and Cr(T) from Real Brackish Water
Adsorptive rejection of Pb(II) and Cr(T) ions was performed on real brackish water, collected from underground water well in Ras Al Khaimah, United Arab Emirates. The in-house characterization of the obtained water was, conductivity 6.6 μS cm
−1, total organic carbon 1.09 mg L
−1, chemical oxygen demand 13.8 mg L
−1, and total dissolved solids 5.3 g L
−1. In two different beakers, 1 L of this water was spiked with 1 mg L
−1 Pb(II) and another liter was injected with 300 µg L
−1 Cr(T). These concentrations of Cr(T) ions were selected based on the in-house analysis of different wells in Ras Al Khaimah areas while Pb(II) ions were not detected; hence, it was spiked with 1 mg L
−1. The influence of Fe–GAC 5 dosage on the elimination of Pb(II) and Cr(T) ions from the actual well water is demonstrated in
Figure 12. The outcomes illustrated that the elimination efficiency drastically improved with increasing the Fe–GAC 5 dose. The Fe–GAC 5 doses of 9 and 3 g L
−1 were sufficient to decrease the concentrations of Pb(II) and Cr(T) ions to 0.0 and 0.027 mg L
−1, respectively. Fortunately, these results were in agreement with WHO standards of Pb(II) and Cr(T) in drinking water [
50]. In comparison to the elimination of Pb(II) and Cr(T) from synthetic water, the Fe–GAC 5 dose of 1.75 g L
−1 was able to achieve a removal between 60–80% for [
] of 40 mg L
−1. The increased Fe–GAC 5 dose required to remove pollutants in real water was due to the presence of other organic pollutants competing for the Fe–GAC 5s’ active sites, leading to higher chances of pore blockage.
3.11. Spent Adsorbent Characterization and Removal Mechanism
Figure 13a,b shows the FE-SEM images of Fe–GAC 5 after the adsorption of Pb(II) and Cr(T) ions, respectively. The EDS analysis of Fe–GAC 5 after adsorption indicated the presence of Pb(II) and Cr(T) ions on the Fe–GAC surface, suggesting the successful removal of the ionic species by Fe–GAC 5 (
Table 6). After the adsorption of Cr(T), the increased oxygen to carbon (O/C) ratio on the Fe–GAC 5 indicated the oxidation of the carbon material with a likely reduction of Cr(IV) ions to Cr(III).
Figure 14 presents the FTIR spectra of Fe–GAC 5 over the adsorption of Pb(II) and Cr(T) ions. A minor move in the position of the O-H group from 3475 cm
−1 to 3433 cm
−1 and 3450 cm
−1 and mere disappearance of the peak stretch was observed upon the adsorption of Pb(II) and Cr(T) ions by Fe–GAC 5, respectively. This is due to the complexation of the -OH group with Pb(II) and Cr(T) ions [
51]. Similarly, the peak at 1647 cm
−1 was also found to decrease to 1577 cm
−1 and 1588 cm
−1 after the adsorption of Pb(II) and Cr(T) ions, respectively. This can be elucidated based on the complexation of the C=O surface groups on the Fe–GAC 5 surface with Pb(II) and Cr(T) ions. In addition, peaks at 2849 and 1080 cm
−1 showed no noteworthy shift after Pb(II) ad Cr(T) adsorption, which suggests that C-H and C=C groups on the Fe–GAC 5 surface did not participate in any complexation with the ionic species. It is worth mentioning that the red shifts of the bands with decreasing intensities are attributed to the successful adhesion of Pb(II) and Cr(T) ions on the Fe–GAC 5 surface [
52]. Moreover, the appearance of new peaks at 470 and 793 cm
−1 can be attributed to the stretching vibration of the M-O (Pb-O and Cr-O) and M-O-M (Fe-O-Pb and Fe-O-Cr) bond, implying that the adsorption and interaction of the metal ions to the Fe–GAC 5 surface was achieved. Similar observations were reported in previous studies [
30,
31,
50,
51].
According to the adsorption results, different adsorption mechanisms were found to control the deletion process of Pb(II) and Cr(T) ions by the Fe–GAC composite. The isotherm modeling indicated that both chemical and physical adsorption processes occurred. Moreover, the kinetic data demonstrated that both surface film diffusion and intra-particle diffusion processes participated in the adsorption process. As revealed from the pH study, the elimination of both Pb(II) and Cr(T) ions was driven by electrostatic interactions. It was also suggested that possible reduction of Cr(IV) to Cr(III) took place at very low pH (under acidic conditions). Furthermore, desorption experiments suggested that Cr(T) removal involved chemisorption due to the lower desorption efficiency of HCl solutions caused by the strong chemical bond formation between the Cr(T) ions and Fe–GAC material. Additionally, the FTIR analysis after Pb(II) and Cr(T) adsorption suggested a surface complexation of Pb(II) and Cr(T) ions onto the surface of the Fe–GAC 5.