Lead and cadmium removal by adsorption process using hydroxyapatite porous materials

Water pollution has a major ecological impact on the environment and human health usually due to industrial effluents with toxic substances such as heavy metals (Lu et al. 2014). Many researches are focused on lead (Pb) and cadmium (Cd) ions which are extremely toxic at low concentrations and considered as carcinogenic heavy metals (Sönmezay et al. 2012). These heavy metals are poured into water and groundwater from mining and petrochemicals industries (Li et al. 2019). Therefore, numerous approaches such as chemical precipitation, adsorption, membrane process, and electrochemical treatment have been investigated to improve and purify the quality of water by reducing the concentration of heavy metals in wastewater (Mousa et al. 2016; Jing et al. 2018). Governments and environmental protection agencies fixed the limit values and guidelines recommended in terms of the allowable concentration for lead and cadmium in wastewaters from industrial effluents and drinking water. The permissible limits recommended by World health Organization (WHO 1993) are 0.01 mg/L and 0.003 mg/L for Pb²⁺ and Cd²⁺ ions respectively. Also, Algeria sets a regulatory framework, which considers lead and cadmium as toxic products and shows that the limit values must not exceed 0.01 mg/L (for Pb²⁺) and 0.003 mg/L (for Cd²⁺) for drinking water (Official Journal of the Algerian Republic 2006), and 0.5 mg/L (for Pb²⁺) and 0.2 mg/L (for Cd²⁺) for discharge in sewage (Official Journal of the Algerian Republic N°18 2011).

Many attempts have been made by researchers for the removal of heavy metals using mesoporous materials, and nanoporous particles such as zeolites, polymers, organic resins, silicate, clays, carbon nanotubes and apatites (Ullah et al. 2016; Song et al. 2016; Alexander et al. 2017; França de Oliveira et al. 2017; Kołodyńska et al. 2017; Pobi et al. 2019). The latter has shown remarkable heavy metal efficiencies. Hydroxyapatite (Ca₁₀ (PO₄)₆ (OH)₂) belongs to the crystallographic family of apatite with a hexagonal structure under Ca/P ratio between 1.67 and 1.5 (Song et al. 2016). Hydroxyapatite (Hap) is a natural important mineral that consists mainly of calcium and phosphate (Kemençe & Bölgen 2017) and can be found in human and animal hard tissues like vertebrate bones, mammalian teeth, fish scales, and the mature teeth of some chiton species (Zhao et al. 2012). Because of its abundance and low cost, hydroxyapatite (Hap) is attracting interest in biology, chemistry and in medical biomaterial such as implants for its chemical composition similarity to the natural bone (Zhao et al. 2012). Hydroxyapatite (Hap) can be synthesized and prepared via several methods, such as electro-deposition or sol-gel method (wet-chemical precipitation) (Ben-Arfa et al. 2017). Hydroxyapatite is one of the promising materials because of its high specific surface area, and high adsorption ability in drug delivery (Yan et al. 2016) dyes and proteins adsorption (Kim et al. 2016), and in particular removal of heavy metal ions from wastewater (Sönmezay et al. 2012). This research focuses on the synthesis of hydroxyapatite (Sy-Hap) from bovine bone as a low-cost and effective crystalline adsorbent, as well as a comparative study of the heavy metals removal efficiency between hydroxyapatite (Sy-HAP) and commercial (C-HAP).

Reagents and chemicals used in this study were of analytical grade Pb(NO₃)₂, Cd(NO₃)₂, and commercial Hydroxyapatite was purchased from Sigma Aldrich. HNO₃ and NaOH were purchased from Biochem. Stock solutions of Pb²⁺ and Cd²⁺ ions were prepared with concentration of 10⁻² mol/L by dissolving 3.318 g and 3.054 g of Pb (NO₃)₂ and Cd (NO₃)₂ salts respectively in demonized water (1,000 mg/L). The Pb and Cd ion concentrations were measured using a UV–Visible Spectrophotometer (UV-UV-2401PC of Schimadzu mark) at wavelengths of 650 nm and 600 nm respectively, according to a procedure described elsewhere (Murakami 1981).

The porous texture and the specific surface area of the samples were determined by nitrogen physisorption at 77 K, using a Micromeritics ASAP 2000 volumetric adsorption device after sample out-gassing at 383 K under vacuum for 4 h. The specific surface area (SBET) of both samples was obtained by adopting the BET equation (Brunauer-Emmett-Teller). The microporous and mesoporous volumes deduced from the data of the N₂ adsorption-desorption isotherms are calculated respectively, according to the methods of Dubinin-Radushkevich (DR) and Barrett, Joyner and Halenda (BJH). In order to determine the pH_PZC of the Hydroxyapatites, we used the methods described in previous studies (Herbache et al. 2016).

For SEM observations, the morphological analysis of the adsorbent samples was performed by Scanning Electron Microscopy (SEM), using a Hitachi S-4500 microscope in the secondary electron scanning mode, working at an operation voltage of 30 kV and coupled with an energy dispersive spectrometer (EDS, Thermofisher). The samples were deposited on a wafer containing silver lacquer and metallic gold or carbon. Fourier transform infrared spectroscopy (FTIR) was carried out using a Bruker Alpha Model Frontier/Multiscope spectrophotometer over a range of 350–3,800 cm⁻¹ with a resolution of 1 cm⁻¹.

The adsorbents used in this study were synthetic hydroxyapatite (Sy-HAP) from bovine bone and hydroxyapatite (C-HAP), which is a commercially available product as reference. To obtain Sy-HAP, the femur bone of the cow sample was washed with distilled water, hydrogen peroxide, nitric acid and bleach to remove impurities. After that, the cleaned bone was heated in an electric furnace (Nabertherm GmbH-LV9/11 B180 (from 30 °C to 3,000 °C)) under ambient conditions; at 850 °C using a heating rate of 58 C/min with 2 h holding time to remove organic substances and avoid any microbial contamination. Then, the obtained bone ash was dried at 100 °C for six hours. The final product was a white powder indexed as Sy-HAP to be used further for heavy metal removal from wastewater.

The FTIR spectra of both hydroxyapatites, C-HAP and Sy-HAP, samples are presented in Figure 1. The stretching vibration bands of -OH groups in both hydroxyapatites, C-HAP and Sy-HAP, are found between 3,500 cm⁻¹ and 3,570 cm⁻¹. The absorptions at 1,081 cm⁻¹ and 1,030 cm⁻¹ are attributed to the stretching vibrations of P = O. The bands at 570 cm⁻¹, 601 cm⁻¹ and 954 cm⁻¹ correspond to the bending vibrations of P-O in PO₄³⁻ (Song et al. 2016; Jing et al. 2018). Moreover, the weak bands located between 1,460 cm⁻¹ and 1,415 cm⁻¹ are related to the CO₃²⁻; this band is similar to that found in layered double hydroxide (Cengiz et al. 2008). In addition, C-HAP presents a band centered at 1,632 cm⁻¹ and assigned to the deformation vibrations of –OH attributed to H₂O molecules physisorbed. Sy-HAP involves two bands stretching and bending located at 1,500 cm⁻¹ and 1,650 cm⁻¹ attributed to C = O and N-H groups respectively (Cengiz et al. 2008). These two bands confirm the existence of organic matter.

The SEM micrographs of the Sy-HAP and C-HAP are given in Figure 2. C-HAP presents aggregation particles with cylindrical rod-like shapes, whose sizes are about 50–100 nm. However, the synthetic hydroxyapatite (Sy-HAP) presents the existence of different size of particles. These latter appear in the form of irregular spheroidal and amorphous grains, which indicated a reduction in crystallinity. Sy-HAP contains multiple pores created by decomposition of organic substances during calcinations at 800 °C (Khoo et al. 2015).

The comparison of the average pore diameter (Dp) is given by two methods, the geometrical one (4Vp/SS) and the BJH. The specific surface area (SS) and the pore volume (Vp) are obtained by the BET method. It can be deduced that the specific surface area of C-HAP and Sy-HAP are 40.8 m²/g and 46.8 m²/g respectively. The pore size distribution value of both biomaterials, namely Sy-HAP and C-HAP, calculated by BJH method are 13.9 nm and 25.5 nm respectively. However, by using the geometrical method, it is noticed a pore size distribution decrease for both biomaterials 10.1 nm for Sy-HAP and 24.6 nm for C-HAP respectively. The pore volume of C-HAP is around 0.02 cm³/g where, Sy-HAP is ca 0.18 cm³/g. On the basis of these results, it can be noted that both hydroxyapatites Sy-HAP and C-HAP are mesoporous materials. The pH of zero point charge (pHpzc) values of Sy-HAP and C-HAP were found 5.8 and 4.6, respectively.

After the adsorption of Cd²⁺ and Pb²⁺ ions, the two hydroxyapatites were dried at 80 °C overnight and their chemical compositions were determined by EDX (see data in Figure 3). The Ca/P molar ratio of HAP (1.657) was close to the stoichiometric data of 1.67 (Khoo et al. 2015; Jing et al. 2018). This result seems to be higher than that of C-HAP. The quantitative elemental analysis given by EDX present in the Sy-HAP and C-HAP samples are C, O, P, Ca, Mg and Na. The presence of oxygen (O), calcium (Ca) and phosphorus (P) in a substantial amount are the main compounds for the crystalline structure of Hydroxyapatite. The characterization of materials after adsorption by EDX confirms the retention of lead and cadmium ions (see Figure 3).

Figure 4 shows the effects of contact time (from 5 min to 120 min) on Sy-HAP and C-HAP for an initial concentrations of the Pb²⁺ and Cd²⁺ ions of 16.5 ppm and 15.4 ppm at ambient temperature respectively. It can be observed that the adsorption of cadmium and lead ions increased rapidly with the contact time at the initial stage and reached equilibrium in 15 min for both ions. The removal efficiency of Pb²⁺ is about 86.53% and Cd²⁺ is c.a 81.43% for C-HAP. However; an increase in the removal efficiency was noticed for Sy-HAP; where Pb²⁺and Cd²⁺ ions were removed up to 95.52% and 90.91%, respectively. The maximum adsorbed amounts of lead and cadmium ions onto Sy-HAP are 152.6 mg/g and 139.5 mg/g respectively and whereas, the maximum adsorbed amounts of C-HAP are 146.77 mg/g for Pb²⁺ ions and 131.5 mg/g for Cd²⁺ ions. The trend of the heavy metals removal by Hydroxyapatites therefore follows the order: Pb²⁺ (Sy-HAP) > Cd²⁺ (Sy-HAP) > Pb²⁺ (C-HAP)> Cd²⁺ (C-HAP). The synthetic hydroxyapatite (Sy-HAP) shows higher heavy metals removal efficiency and the adsorption capacity to be in agreement with its higher BET surface area. Jing et al. (2018) prepared super-small HA nanoparticles with an average diameter of 7 nm in the presence of glucose as an adsorbent to remove heavy metal ions from wastewater. The maximum adsorption capacities of HA nanoparticles compared to commercial HA to the metal ions are 3,289 mg/g (to Pb²⁺) and 2,784.8 mg/g (to Cd²⁺). Jiang et al. (2012) used HA hollow microspheres to remove Pb²⁺ and Cd²⁺ mixed ions from water and found that when the initial concentration of the mixed metal ions was 50 μg/mL, the adsorption capacity of Pb²⁺ and Cd²⁺ ions could reach to 99.79 mg/g and 38.78 mg/g, respectively.

Based on the obtained results, the adsorption capacity of Pb²⁺ ion was quite higher than that of cadmium ions. This difference can be attributed to the ionic properties such as ion radius, charge, ionic potential, enthalpy of hydration, electronegativity, and the electron configuration of each metal ion. It is known that, the surface complexation reaction is more influenced by the electrostatic attraction between the surface charge of hydroxyapatite and the dissolved ions (Sönmezay et al. 2012; Mousa et al. 2016). The electronegativity of Pb²⁺ ion (2.33) is slightly higher than that of Cd²⁺ ion (1.69), which leads to the PbII ion to interact more strongly electrostatically with the surface groups present on the surface of the adsorbent. On the other hand, the tendency to lose water molecules from the aquo-cations is stronger for PbII ion since the single ion hydration enthalpy is −1,481 kJ/mol for PbII ion and −1,807 kJ/mol for CdII ion. This also facilitates the interaction between the PbII ion and the adsorbent surface (Sönmezay et al. 2012).

In comparison with the pseudo-first-order model, the linearity of the pseudo-second-order model is good, and the data of R² corresponding to various conditions tabulated in Table 1 are above 0.99. In addition, the data of q_e,cal derived from the pseudo-second-order model were found to be close to the experimental results, which indicates the appropriateness of this model. Therefore, the pseudo-second-order model can well describe the adsorption kinetics process of metal ions on hydroxyapatites. Furthermore, according to the pseudo-second order model, the boundary layer resistance was not a rate limiting step. For that, the adsorption rate was controlled by intraparticle mass transport in the interior of the adsorbent (Lu et al. 2014). According to the experimental results from the representation of the linearized regression Morris and Weber model, the diffusion process occurs within the initial stage (about 15 min). The coefficients of determination are found to be R² = 0.983–0.994 with k_d = 34.64–39.37 (see Table 1). The high R² values for this model indicate that the diffusion process occurs fairly well during the initial stage.

The pH is an important parameter affecting the sorption process. The effect of initial pH on the equilibrium metal cation uptake is given in Figure 5. The cations were adsorbed effectively in the acidic range. Pb²⁺ and Cd²⁺ ions removal was greatly increased by increasing solution pH from 2 to 5 for both Sy-HAP and C-HAP. At higher pH values (pH > 5), a slow decrease in the adsorption of metal ions was observed. It is known from literature (Khoo et al. 2015; Mousa et al. 2016) that hydroxyapatites are characterized by the functional groups = PO and = CaOH₂⁺, which are the predominant sites for a pH value close to the pH_pzc. These groups are capable of adsorbing or releasing H⁺ ions and they become very significant for a lower pH value and higher than the pH_pzc.

At very low pH, the HAP surface would also be surrounded by predominantly H⁺ ions. Consequently, repulsive force and the destruction of crystalline structure for the HAP material result in their dissolution in an acidic medium, which results in a decrease in the adsorption of Pb²⁺ and Cd²⁺ ions onto hydroxyapatites (Mousa et al. 2016). Furthermore, when the pH increases, the competition between positive charges decreases as these surface active sites become more negatively charged, which increases the adsorption of the Pb²⁺ and Cd²⁺ ions through electrostatic force of attraction (Sönmezay et al. 2012; Mousa et al. 2016). However, beyond pH 5, no removal increase was observed. The effect of pH has not been studied at higher pH values due to the presence of PbOH⁺, Pb(OH)₂, [Pb₂(OH)₃]⁺, [Pb₃(OH)₄]²⁺, [Pb₆(OH)₈]⁴⁺, Pb(OH)₃, CdOH⁺, Cd(OH)₂, Cd(OH)₃⁻ and Cd(OH)₄²⁻ (Sönmezay et al. 2012). The formation of these hydroxide species disadvantages the adsorption of Cd²⁺ and Pb²⁺ ions on HAP (Mousa et al. 2016). Previous studies have also reported that the best adsorption efficiency of Pb²⁺ and Cd²⁺ ions has been observed at pH 4 and 5 (Sönmezay et al. 2012; Mousa et al. 2016). On the other hand, pH can have a substantial influence on the surface charge of solids, which behaved differently in adsorbing metal ions at different solution pHs. The pH_pzc of Sy-HAP and C-HAP powders was found to be 5.8 and 4.6 respectively. This means that the HAP surface was positively charged at solution pH below pH_pzc and Pb²⁺ ions were repulsed by the HAP material surface, resulting in the removal of Cd²⁺ and Pb²⁺ ions. At pH higher than pH_pzc, the surface of HAP can be deprotonated and typically charges negatively (Herbache et al. 2016; Mousa et al. 2016). Thus, it is advantageous that the process is a suitable application for heavy metals removal because of its neutral and clean effluent.

Figure 6 shows that the metal adsorption capacity of both biomaterials increased gradually with the adsorption concentration at equilibrium. The Sy-HAP presents the highest rate of adsorption for both metallic ions. For different heavy metal ions, at low initial concentrations, the adsorption equilibrium is quickly reached and saturation is achieved; while with high initial concentrations, the adsorption is slow and the driving force greater, forcing the solution to reach equilibrium easier. This indicates that HAP has a high affinity for the metals studied. Furthermore, the maximum adsorption is achieved with an adsorption capacity in the order of 166.67 mg/g Pb²⁺ ion, 138.89 mg/g Cd²⁺ ion, 142.86 mg/g Pb²⁺ ion and 125 mg/g Cd²⁺ ion for Sy-HAP and C-HAP, respectively. The parameters and coefficient of determination (R²) obtained are reported in Table 2. It can be concluded that the Langmuir, Freundlich and Temkin models are the most valid to describe and evaluate the adsorption processes for the Pb²⁺ and Cd²⁺ ion adsorption by the HAP material. Whereas the Dubinin − Radushkevich (D − R) model was found to be less favorable for elucidating these types of surface reactions. The validity of the isotherm models is in the following order (Table 2), Langmuir, Freundlich and Temkin > Dubinin-Radushkevich. Based on the slight difference in R² values, which exceed 0.984, and based on the pore size of both materials, which belong to mesoporous materials, the appropriate model should be Freundlich.

The values of standard enthalpy and entropy of the adsorption can be determined from the slope and intercept of a linear plot of ln K_d vs 1/T. The calculated values of the thermodynamic data are shown in Table 3. The overall lead and cadmium adsorption process seems to be endothermic (ΔH_ads Pb²⁺ (C-HAP) = 32.093 kJ/K*mol, ΔH_ads Pb²⁺ (Sy-HAP) = 14.783 kJ/K*mol, ΔH_ads Cd²⁺ (C-HAP) = 22.57 kJ/K*mol, and ΔH_ads Cd²⁺ (Sy-HAP) = 6.922 kJ/K*mol). Although not very high, these values of enthalpy ΔH_ads can be interpreted on the basis of considerably strong interaction between metal ions and hydroxyapatite surface under ion exchange process. Negative values of free energy ΔG_ads (−14.575 to −8.364 kJ/K*mol) confirm the spontaneous nature of the adsorption. Also, it can be noticed that for different adsorbents, ΔG_ads increases with the increase in the solution temperature. Indeed, the ΔS_ads values were positive (0.012–0.056 kJ/mol). This occurs as a result of redistribution of energy and an increased randomness at the adsorbent– adsorbate interface during the adsorption process. Similar observations were reported for the adsorption of lead and cadmium ions from aqueous solution onto different materials such as Schiff base-modified nanoparticles, amino-bacterial cellulose, modified biomass ash and manganese oxide minerals (Sönmezay et al. 2012; Lu et al. 2014; Herbache et al. 2016; Xu et al. 2017).

Adsorption performance of a low-cost and efficient adsorbent prepared from bovine bones, Sy-HAP, was studied for the removal of Pb²⁺ and Cd²⁺ ions from aqueous solutions, compared to commercial hydroxyapatite C-HAP. The experimental results indicated that the equilibrium time required for a maximum adsorption of Pb²⁺ and Cd²⁺ ions on both hydroxyapatites is 15 min. The obtained yields for Sy-HAP are 96.52% and 92.91% for Pb²⁺ and Cd²⁺ ions respectively, whereas C-HAP presents slightly lower removal efficiencies of 87.53% and 85.43% for Pb²⁺ and Cd²⁺ respectively. These removal efficiencies correspond to the maximal adsorption capacities of Pb²⁺ 152.6 mg/g and Cd²⁺ 139.5 mg/g, respectively for Sy-HAP and to Pb²⁺ 138.77 mg/g and Cd²⁺ 131.5 mg/g, respectively for C-HAP. The best adsorption efficiency of Pb²⁺ and Cd²⁺ ions has been observed at pH 5 for both hydroxyapatites. Isotherm adsorption studies indicated that Langmuir, Freundlich and Temkin are the most valid models to describe and evaluate the adsorption process. Indeed, the kinetic adsorption process followed the pseudo-second-order with intraparticle diffusion, which is significant at the initial adsorption stage. Negative values of ΔG_ads and positive values of ΔH_ads indicate that the adsorption process is spontaneous and endothermic with a physical nature. The results show that efficient and inexpensive hydroxyapatite minerals can be useful adsorbents for removing metal ions from aqueous solution.

The author would like to thank the European Institute of Membranes, University Montpellier, France, for the realization of characterization of materials in this present study.