Adsorption of Pb(II) from Water onto ZnO, TiO2, and Al2O3: Process Study, Adsorption Behaviour, and Thermodynamics

This study is aimed at comparing the use of zinc oxide (ZnO), titanium dioxide (TiO2), and aluminium oxide (Al2O3) for removing lead ions from water through adsorption. The point of zero charge was obtained for ZnO, TiO2, and Al2O3 and was found to be 7.3, 7.1, and 9.0, respectively. The effect of pH, adsorbent dose, contact time, initial concentrations, and temperature was investigated in batch experiments. The optimal conditions obtained were 7, 2 g/L, 120 mins, 100 ppm, and 41°C, respectively, where the optimal removal efficiencies were 98.43%, 96.45%, and 85.50% for ZnO, TiO2, and Al2O3, respectively. In addition, analyses of adsorption kinetics, mechanisms, isotherms, and thermodynamics were performed. The adsorption kinetics of Pb(II) were compared to popular models, and it was found that the pseudo-second-order (PSO) model best fitted the Pb(II) uptake for all adsorbents at correlation coefficient ( R 2 ≥ 0.96 ). The adsorption isotherms of Pb(II) were also compared to popular models, and it was found that the Pb(II) uptake by TiO2 and ZnO was well-described by the Langmuir model ( R 2 ≥ 0.96 ) with maximum adsorption capacities of 55.04 and 58.88 mg/g, respectively. On the other hand, the behaviour of Al2O3 is described more accurately by the Dubinin-Radushkevich (D-R) model ( R 2 = 0.96 ), and the maximum adsorption capacity was 53.64 mg/g. The isotherm analysis proved that the limiting step of the adsorption process is the film diffusion mechanism. In addition, studying the heat of adsorption of Pb(II) implied that the adsorption is endothermic due to the positive values of enthalpy ( Δ H ° ≥ 30 ) for all adsorbents. The absorbents were characterized using scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDX) to study the morphology of surfaces and the chemical characterization of the adsorbents to ensure that adsorption is achieved. ZnO showed better performance for the uptake of lead followed by TiO2 then Al2O3.


Introduction
Contamination of aquatic environments by heavy metals is one of the most major environmental challenges because of their flexibility, aggregation, persistence, and nonbiodegradable nature. Some of the toxic heavy metals, in particular lead, can have severe and poisonous effects on human beings and marine organisms even at trace levels [1,2]. This problem is exacerbated in developing countries, where polluting industries are rapidly developed for various applications, such as mining operations, tanneries, batteries, fertilizer, pesticides, paper industries, and coke factories [3].
Various adsorbents are used to remove heavy metals, such as arsenic, cadmium, nickel, mercury, chromium, zinc, and lead, from aqueous solutions. These adsorbents include carbon-based adsorbents [4], bioadsorbents [5], low-cost adsorbents [6], and polymeric nanosorbents [7]. This study sheds light on lead, which is one of the most toxic heavy metals that attracts considerable attention from environmentalists. Lead is discharged into aquatic ecosystems from various industrial activities, such as storage batteries, ceramic glass industries, mining, plating, coating, and automotive industries. Another potential source for lead contamination is the wide utilization of agricultural chemicals such as fertilizers and fungicidal sprays [8]. In addition to the pollution of aquatic environments, the adverse effects of lead on human health have been well documented and characterized. Lead poisoning causes serious harm to the kidneys, damages the central nervous system and liver, and can result in cancer and brain damage, in addition to abnormalities in living creatures' organs [9,10].
In response to this, several techniques have been developed to remove harmful heavy metals from aqueous solutions and maintain environmental safety, such as chemical precipitation, ion exchange, membrane separation, adsorption, and electrochemical removal [11]. Among these techniques, adsorption has recently gained growing importance for being an efficient, effective, and economical technique. It provides high quality treated outputs, low operating cost, high design and operating flexibility, and high removal efficiency in the absence of interference of any competing additives [12].
A wide range of adsorbents was used to extract lead, such as activated carbon [13], modified alginate aerogel with melamine/chitosan [11], reduced graphene oxide-Fe 3 O 4 [8], and pea peel waste [14]. The use of zinc oxide (ZnO), titanium dioxide (TiO 2 ), and aluminium oxide (Al 2 O 3 ) as adsorbents for removal of different constituents from wastewater is promising in light of its advantages, such as availability, simplicity, nontoxic nature, resilience to corrosive elements, and a strong affinity for metal ions used in wastewater treatment [15]. Efficient removal of heavy metals was reported in the literature using ZnO nanocomposites and nanoparticles [16,17], TiO 2 composites and nanoparticles [2,18], and Al 2 O 3 nanocomposites and nanoparticles [19][20][21][22]. However, few studies compared the capacities of these three adsorbents in the removal of heavy metals.
Despite the studies mentioned above, limited studies compared the adsorption capacity and removal mechanism for lead adsorption onto zinc oxide (ZnO), titanium dioxide (TiO 2 ), and aluminium oxide (Al 2 O 3 ) under the same conditions. To our knowledge, no study compared the thermodynamics of these metal oxides for this particular case. Thus, this research is aimed at studying and comparing the use of these metal oxides for lead removal from water by adsorption. The effect of pH, adsorbent doses, contact time, initial concentrations, and temperature on lead removal was investigated. The experimental results were analyzed using pseudo-first order (PFO), pseudo-second order (PSO), Elovich equation, the intraparticle diffusion model, and the Boyd kinetic models to determine the most compatible adsorption kinetic and mechanism model. Subsequently, the equilibrium data were investigated using various isotherm models, such as Langmuir, Freundlich, Temkin, and Dubinin-Radushkevich (D-R) models, to determine the appropriate isotherms for the characterization of the behaviour of lead adsorption on metal oxides.

Materials and Methods
Lead nitrate (Pb (NO 3 ) 2 ), manufactured by Loba Chemie Pvt. Ltd., India, 99% extra pure, was used to prepare the contaminated solution. A 1000 ppm stock solution is prepared by dissolving 1.58 g of lead nitrate in 1000 mL of deionized water containing 10% (v/v) of hydrochloric acid (HCl) for 24 h to prevent the precipitation of Pb(II) due to hydrolysis. Other standard solutions were prepared by diluting this solution as required to acquire working solutions between 50 and 150 mg/L of Pb(II). The initial pH of any experimental solutions was modified by adding HCl or sodium hydroxide (NaOH) solutions.
Zinc oxide 99% pure (AR zinc white), titanium dioxide, 98% extra pure, and aluminium oxide active (basic), manufactured by Loba Chemie Pvt. Ltd., India, were used as adsorbents. The surface of materials before and after adsorption was studied using scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDX). By following procedures described in earlier studies [23,24], the points of zero charge for each of the three adsorbents using the pH drift method were found to be 7.3, 7.1, and 9.0 for ZnO, TiO 2 , and Al 2 O 3 , respectively, see Figure 1.
The adsorption experiments are divided into several batches, where each batch evaluates a specific parameter that affects the adsorption process and determines its optimal value among the tested range. The tested parameters were pH of the solution, adsorbent dose, contact time, initial concentration of Pb(II), and temperature. The effect of pH was investigated by adjusting the solution pH between pH 3 and 9 using 0.1 M of nitric acid (HNO 3 ) and NaOH. The adsorbent dosage tested range varied between 0.01 and 0.05 g. The effect of contact time was investigated by varying the contact time from 5 to 120 min. The impact of the initial concentration of lead was studied between 50 and 150 ppm, and the temperature effect was studied between 18 and 41°C. For each batch, the solution flasks were shaken to ensure that the solution reached equilibrium (maximum adsorption capacity). At the end of each batch, the solution was centrifuged, filtered using a filter paper (Whitman number 45), and analyzed using inductively coupled plasma (ICP) spectrometry (model Ultima 2 JY Plasma) to measure the final concentration of lead in water samples complies to Environmental Protection Agency (EPA-US) measurement technique [25]. The adsorption percentage and the capacity of adsorption can be calculated according to Equation (1) and Equation (2), respectively, [26]: where q e is the mass of metal ions adsorbed per unit weight of the adsorbent at equilibrium (mg adsorbate/g adsorbent), C 0 is the initial concentration of the metal ions (mg/L), C e is the equilibrium concentration of the metal ion (mg/L), V is the volume of the metal ion solution (mL), and m is the mass of the adsorbent (mg).  Figure 2(a). The adsorption percentage increased as the initial pH of the solution increased from pH 3.0 to 7.0 for all adsorbents used. The surface of the adsorbents became more negatively charged, and subsequently, the electrostatic attraction (ion exchange) between the metal ions and surface of the adsorbents probably increased. The increase in adsorption percentage may be attributed to the interaction of Pb 2+ , Pb(OH) + , and Pb(OH) 2 with the functional groups present on the surface of the adsorbents due to ion exchange mechanism or by hydrogen bonding [27] as shown in Figure 3 and the following reactions: Ion exchange mechanism:

Results and Discussion
Hydrogen bonding: At higher pH values (7.0-9.0), a decrease in the adsorption percentage was observed in ZnO and TiO 2 due to the formation of soluble hydroxyl complexes. However, for Al 2 O 3 , an increase in adsorption percentage was observed since the value of pH at the point of zero charge (pH PZC ) for Al 2 O 3 is 9.0, see Figure 1. As the pH of the solution increases, the adsorbent surface becomes less positively charged which allows for more adsorption of Pb(II) cations. For the following batches, an optimal pH value of 7.0 is used. Below this optimal value, competition between Pb(II) ions and H 3 O + reduces the adsorption efficiency, while higher pH values result in hydrolysis of lead species which encourages precipitation and prevents quantitative adsorption.

Effect of Adsorbent
Dosage. This batch of experiments was conducted with different concentrations of adsorbents, namely, 0.01, 0.02, 0.03, 0.04, and 0.05 g, at the optimal initial pH value, with 25 mL of the contaminated solution. The initial Pb(II) concentration was 100 ppm at room temperature of 29 ± 1°C, and the blend was shaken for 120 min at a constant stirring velocity of 150 rpm. The final Pb(II) concentration values were measured, and the adsorption percentage was calculated using Equation (1). The maximum removal efficiencies achieved were 95.27%, 90.46%, and 70.8% for ZnO, TiO 2 , and Al 2 O 3 , respectively, at a corresponding adsorbent dosage of 0.05 g as shown in Figure 3(b). The adsorption percentage increases as the doses of the adsorbents in the solution increase from 0.01 to 0.05 g for all adsorbents used. These results can be interpreted as increasing the adsorbent dose from 0.01 to 0.05 g provides more active binding sites for the adsorption process. Thus, more Pb(II) ions are adsorbed onto the surface of the adsorbent since the competition between molecules for active binding sites is reduced. The optimal dose obtained was 0.05 g, applied for the subsequent batches.

Effect of Contact
Time. This batch of experiments was conducted at different contact times, namely, 15, 30, 60, 90, and 120 min, with optimal pH value and adsorbent dose of 7.0 and 0.05 g, respectively, at room temperature of 29 ± 1°C and 25 mL of contaminated solution. The initial Pb(II) concentration was 100 ppm at a constant stirring velocity of 150 rpm. The final Pb(II) concentration values were measured, and the adsorption percentage was calculated using Equation (1). The maximum removal efficiencies obtained were 96.17%, 91.75%, and 79.5% for ZnO, TiO 2 , and Al 2 O 3 , respectively, at a contact time of 120 min as illustrated in Figure 2(c). The adsorption percentage increases as the process contact time increases from 15 to 120 min for all the adsorbents used. However, it appears that the rate of lead uptake is the greatest within the initial 15 min because of the availability of active binding sites on the surface of the adsorbents. Consequently, it can be seen that the equilibrium of lead uptake by all the adsorbents is reached within 90 min, following which there is an insignificant change in lead uptake until 120 min.

Effect of Initial Concentration of Pb(II)
. This batch was conducted at different initial concentrations of Pb(II), namely, 50, 75, 100, 125, and 150 ppm, at the optimal pH value and the optimal adsorbent dose obtained from previous batches with 25 mL of contaminated solution at room temperature of 29 ± 1°C. The blend was shaken for 120 min at a constant stirring velocity of 150 rpm. The final Pb(II) concentration values obtained were measured, and the adsorption percentage was calculated using Equation (1). The maximum removal efficiencies were 96.17%, 91.75%, and 79.50% for ZnO, TiO 2 , and Al 2 O 3 , respectively, at an initial concentration of 100 ppm for all adsorbents as shown in Figure 2(d). The results suggest that the percentage of

Adsorption Science & Technology
Pb(II) removal by adsorption increases as the initial Pb(II) concentration increases to a certain extent. The adsorption percentage increased from 93.55 to 96.17% for ZnO, from 85.88 to 91.75% for TiO 2 , and from 61.50 to 79.50% for Al 2 O 3 as the initial concentration of Pb(II) is increased from 1 to 100 ppm. Thereafter, the percentage of Pb(II) removal decreases as the initial Pb(II) concentration increases to a value of 150 ppm. This can be interpreted as at a fixed amount of adsorbent; the high initial Pb(II) concentration leads to an excess amount of available particles compared with the active binding site on the adsorbent surface, so that the percentage of Pb(II) removal by adsorption decreases.
3.1.5. Effect of Temperature. This batch was conducted at different temperature values, namely, 18, 29, and 41°C, at an optimal pH of 7.0, an optimal dose of adsorbents equal to 0.05 g, and a 25 mL contaminated solution of 100 ppm Pb(II) as an optimal initial concentration. The blend was shaken for 120 min at a constant stirring velocity of 150 rpm. The final Pb(II) concentration values were measured, and the adsorption percentage was calculated using Equation (1). The maximum removal efficiencies were 98.43%, 96.45%, and 85.50% for ZnO, TiO 2 , and Al 2 O 3 , respectively, at a temperature of 41°C for all adsorbents as shown in Figure 2(e).
Log q e − q t ð Þ= Logq e − k 1 2:303 t, ð5Þ where q e is the mass of absorbed pollutant per unit weight of adsorbent at equilibrium (mg/g), q t is the adsorbed amount of adsorbent at time t (mg/g), k 1 is the PFO equation constant (min -1 ), t is time (min), k 2 is the rate constant of PSO equation (g/mg·min), a is a parameter that describes the initial rate of adsorption (mg/g·min), and 1/b is a parameter relative to the number of adsorption sites on the surface of the adsorbents (mg/g). The PSO model's correlation coefficient (R 2 ) compared to the adsorption results was greater than that of the PFO model and Elovich equation, which suggests that the PSO model best describes the kinetics of the adsorption process. This also indicated that chemisorption was the reaction's rate-limiting part, as shown in Figures 4(a)-4(c). The kinetic parameters of these models are summarized in Table 1. 3.3. Adsorption Mechanisms. The adsorption process consists of three controlling steps: (1) the motion of the adsorbate from the bulk liquid to the surrounding film of the adsorbent, a process known as film diffusion, (2) the transport of the adsorbate from the film to the adsorbent surface, a process known as surface adsorption, and (3) the transmission of the adsorbate to the internal active sites, a process known as intraparticle diffusion [31]. The active sites sequentially attach to these metal ions. Step (2) Step (1

Adsorption Science & Technology
The limiting step of adsorption refers to the slowest step in the process that controls the total adsorption rate. To identify the limiting step of the adsorption mechanism, several kinetic models are commonly used to investigate the rate-limiting step; this included the intraparticle diffusion model as well as the Boyd model. The intraparticle diffusion model was applied to determine the adsorption mechanism. The model is described by Equation (8) [32].
where q t is the amount of Pb(II) adsorbed at time t (mg/g), k id is the intraparticle diffusion rate constant (mg/g·min 0.5 ), and C is a constant that is related to the thickness of the boundary layer. Figure 5(a) shows that the plot of Pb(II) uptake against t 1/2 is not a simple line passing through the origin, indicating

Adsorption Science & Technology
Furthermore, the Boyd kinetic model is used to determine the slowest step that limits the adsorption process. The model is described by Equation (9) [36], and the results are presented in Figure 5(b).
For any adsorbent, a linear plot that succeeds to pass through the origin indicates that the limiting step in the adsorption process is the intraparticle diffusion. However, it is clearly shown in Figure 5(b) that the plot of Bt against time is linear for each of the three adsorbents, but the lines failed to pass through the origin, which indicates that the process is controlled by film diffusion [34,36].

Equilibrium Isotherms.
Isotherms are the equilibrium relationship between the solid phase and the liquid phase solution concentration. The suitability of a particular adsorbent for extraction of a pollutant can be studied using adsorption isotherms, and the maximum adsorption capacity can also be obtained using various mathematical expressions, such as Langmuir [37], Freundlich [38], Temkin [39],   [40] isotherms. These are the most common isotherms used to describe a solid-liquid adsorption system. These isotherms are described by Equation (10), Equation (11), Equation (12), and Equation (13), respectively.
Log q e = Log K F + Log C e n , ð11Þ ln q e = ln q D - where K L is a constant related to the affinity of the binding sites (L/mg), a L is the Langmuir isotherm, K F is the Freundlich adsorption constant (mg/g (mg/L) −1/n ), 1/n is the intensity of adsorption in the system, K T is the constant of equilibrium binding (L/mg), T is the temperature (K), R is the universal gas constant (8.314 J/mol·K), b is the Temkin isotherm constant, C e is the adsorbate concentration at equi-librium (mg/L), q D is the theoretical capacity of the adsorbent at saturation (mg/g), B is a constant related to the mean free energy of adsorption per mole of the solute (mol 2 /J 2 ), and ε is the Polanyi potential which is related to equilibrium (kJ 2 mol 2 ). Figures 6(a)-6(d) represent plots of the experimental data using the linearized form of the isotherms mentioned above models. Each isotherm has specific assumptions regarding the conditions at which adsorption takes place. Freundlich isothermic adsorption assumes that adsorption occurs on a heterogeneous surface through multilayer adsorption processes, while Langmuir isotherm assumes maximum limiting adsorption at a given number of accessible sites on the adsorbent surface, with the same energy available at all adsorption sites [41,42]. Temkin isotherm assumes that the adsorption heat decreases linearly as the surface of the adsorbent is covered. The Dubinin-Radushkevich isotherm considers the influence of the porous structure of the adsorbent surface.
The corresponding correlation coefficients (R 2 ) were determined for each absorbent as shown in Figures 5(a)-5(d). The Langmuir isotherm better described the adsorption behaviour of ZnO and TiO 2 , while the adsorption behaviour of Al 2 O 3 was best described by the Dubinin-Radushkevich isotherm. Results of equilibrium studies on Pb(II) uptake using ZnO, TiO 2 , and Al 2 O 3 are summarized in Table 2. Several adsorbents from other studies were compared to the adsorbents used in this research concerning their adsorption capacities, and the comparison is presented in Table 3. According to this comparison, metal oxides have a promising future as an adsorbent in wastewater treatment, as approved by their excellent results.

Thermodynamics Studies.
Since temperature is one of the most important factors in the adsorption processes, the thermodynamics of the adsorption process shall be investigated. To determine the inherent energetic changes that are linked to the adsorption process, thermodynamic constants were investigated by varying the equilibrium constants with temperature to determine: (1) Gibbs free energy (ΔG°), (2) enthalpy (ΔH°), and (3) entropy (ΔS°). The distribution coefficient (K d ) is calculated using Equation (14), followed by Equation (15) [45]: ln The values of enthalpy (ΔH°) and entropy (ΔS°) are determined by plotting ln K d versus 1/T and then calculating the resulting slopes and determining the interception of a Van't Hoff chart as shown in Figure 7. The Gibbs energy (ΔG°) of adsorption is calculated using The thermodynamic parameters of Pb(II) adsorption onto    Table 4. The positive enthalpy values indicate that Pb(II) adsorption by ZnO, TiO 2 , and Al 2 O 3 is an endothermic process. In addition, the high enthalpy values (ΔH°> 20 kJ/mol) indicate that chemisorption is involved in the adsorption process as an ion exchange mechanism [46]. The positive values for entropy reflect excess randomness at the solid-liquid interface through the adsorption process. The negative values of the Gibbs energy indicate that the procedure is spontaneous and that this spontaneity increases as the temperature increases.  (Figures 8(a) and 8(c)) showed that both ZnO and TiO 2 exhibit tough and irregular surfaces with spongy openings before the adsorption process. In contrast, the SEM images for Al 2 O 3 , Figures 8(e), show solid and smooth surfaces with few porous openings. This may explain the low removal efficiency of Pb(II) using Al 2 O 3 compared to ZnO and TiO 2 . After the adsorption process (Figures 8(b)

10
Adsorption Science & Technology different magnification scales reveals a clear change in the morphology and emphasizes that adsorption occurred. EDX analysis was used to determine the elemental and chemical composition of a sample and measure the amount of trace elements. EDX analysis provided further affirmation of Pb(II) uptake on ZnO, TiO 2 , and Al 2 O 3 . Figures 9(a)-9(f) show the EDX spectra for ZnO, TiO 2 , and Al 2 O 3 before (unloaded) and after (loaded) the adsorption process. Figures 9(a), 9(c), and 9(e) show no characteristic signals for any metal ions except for Zn 2+ , Ti 2+ , and Al 3+ , respectively. On the other hand, Pb(II) signals were observed in Figures 9(b), 9(d), and 9(f). This shows the accumulation of Pb(II) onto the surface of ZnO, TiO 2 , and Al 2 O 3 particles.

Conclusions
In this research, a comprehensive study of lead removal by adsorption was conducted using three different metal oxide adsorbents, namely, zinc oxide (ZnO), titanium dioxide (TiO 2 ), and aluminium oxide (Al 2 O 3 ). The points of zero charge for the considered adsorbents were found to be 7.3, 7.1, and 9.0 for ZnO, TiO 2 , and Al 2 O 3 , respectively. The effect of pH of the solution, adsorbent dose, contact time, initial concentrations, and temperature on the adsorption process and removal efficiency was studied. The optimal values of these parameters were obtained using separate batch experiments. After applying the optimal values of the studied adsorption parameters for the adsorbents, the adsorption kinetics were investigated using different models, such as PFO, PSO, and Elovich equation, where PSO best fit the adsorption of lead for all of the adsorbents. Furthermore, the adsorption mechanism was investigated using the intraparticle adsorption and Boyd kinetic models. The intraparticle adsorption model showed that the adsorption of lead was a multistep controlled mechanism, while Boyd's kinetic model showed that the film diffusion mechanism is the limiting step of the adsorption process for all the considered adsorbents. In addition, the adsorption isotherms were studied at equilibrium. The results showed that the Langmuir isotherm model best described the adsorption of lead by ZnO and TiO 2 , while the adsorption behaviour of Al 2 O 3 was best fitted by the Dubinin-Radushkevich isotherm model. The thermodynamic studies showed that the adsorption of lead onto ZnO, TiO 2 , and Al 2 O 3 was endothermic and spontaneous. Finally, the morphological and chemical compositions of the adsorbents were tested using SEM and EDX analysis. The SEM images showed that both ZnO and TiO 2 surfaces had very porous holes compared to Al 2 O 3 . Changes in the adsorbent morphologies confirmed that adsorption occurred. The EDX spectra illustrated the presence of lead signals after the adsorption process, ensuring that lead ions had accumulated onto the surface of the ZnO, TiO 2 , and Al 2 O 3 particles. At the end, it can be concluded that under the optimal conditions adopted in this study, lead removal from aqueous solutions by adsorption is best obtained using ZnO followed by TiO 2 and finally comes Al 2 O 3 .