Preparation of hydroxyapatite-based porous materials for absorption of lead ions

Hydroxyapatite (HA) is the main inorganic constituent of vertebrate hard tissues such as bones and teeth, which has been chosen to study the system of biomolecule/biocompatible-surface interactions (Rimola et al. 2008). As the major constituent of bone (Akram et al. 2014; Liu et al. 2016), HA was coupled with inorganics (Fu et al. 2015; Anitha et al. 2017) and organics (Lv et al. 2013; Lin et al. 2016) for bone tissue engineering application. Since excellent biocompatibility, biodegradability, non-toxicity and high stability, it was extensively used in biomedical fields including dental restoration (Lee et al. 2017), controlling transport of proteins (Zhu et al. 2017), genes (Escudero et al. 2013; Deshmukh et al. 2016) and other applications (Fujita et al. 2016; Zhang et al. 2017). Furthermore, HA materials are of great significance in many other fields such as removing harmful volatile organic compounds (VOCs) (Fang et al. 2016), fuel cells (Wei & Matthew 2012) and corrosion inhibitors (Snihirova et al. 2010).

Due to Ca²⁺ exchange performance, HA has adsorption and immobilization effect on metal ions, such as Cd²⁺, Pd²⁺. Because of this feature, HA has been successfully used in heavy metal-contaminated soil (Wei et al. 2016) and water treatment (Meski et al. 2011; Wijesinghe et al. 2018). To improve absorption efficiency, biopolymer-hydroxyapatite composite was employed to remove heavy metals from aqueous solutions. Vila et al. (2011) have developed 3D-macroporous biopolymer-coated hydroxyapatite (HA) foams as potential devices for the treatment of Pb²⁺, cadmium and copper contamination of consumable water (Vila et al. 2011). Narwade et al. (2018) used in-situ synthesis method to prepare hydroxyapatite in the presence of cellulose nanofibrils (CNF) or TEMPO-oxidized CNF (TCNF) by wet chemical precipitation process, which combined with a low-cost biosorbent to remove heavy metal ion. Moreover, scientists found that HA particles with different morphologies, such as fibrous, plate-like, agglomerate-like, rodlike, needle-like and spherical, highly affected the efficiency of remove heavy metal ions (Qi et al. 2016; Wang et al. 2016).

The absorbent's surface structure and morphology deeply influence remove rate. We found that polymer modified inorganic clay could efficiently remove heavy metal ions (He et al. 2012), and plant-protein-modified semiconductor photocatalyst increased activity in photodegradation of dyes (Wang et al. 2018). Recently, increasing emphases were turned to natural polymers in the development of environmentally-friendly products (Wang et al. 2017; Al-Othman et al. 2012; Naushad & ALOthman 2015). As the soy protein isolate (SPI) with reproducible resource was used as template, this work could provide an efficient new treatment without high cost. In addition, HA and SPI are multifunctional biomaterials with biocompatibility and environmental friendliness. Herein, we prepared a novel HA-based hybrid with special structure and morphology by using SPI as bio-template. The obtained hybrid material was used as adsorbent to remove heavy metal ions from aqueous solutions to investigate its adsorption property. The influence of time, initial concentration of heavy metal ions, and the amount of adsorbent was tested. These hybrids have been studied in terms of isotherm models and sorption kinetics analysis at neutral medium in order to consider their application as heavy metal removers from aqueous solutions.

SPI (Shanghai Yuanye Bio-Technology Co. Ltd), urea (Tianjin Zhiyuan Chemical Reagent Co., Ltd), (NH₄)₂HPO₄ (Shanghai Zhongqin Chemical Co., Ltd), Ca(NO₃)₂ (Tianjin Kexin Chemical Industry Co., Ltd), and Pb(NO₃)₂ (Tianjin Daming Chemical Reagent Factory) were obtained commercially and used without further purification. Distilled water was used in all experiments.

In 20 mL of urea solution (8 mol/L), 1.000 g SPI was dissolved by stirring at room temperature, and then stirred for 12 hrs. Subsequently, the SPI solution was treated for 4 h and denatured in a heating bath at 90 °C. After adding 5 mL of (NH₄)₂HPO₄ aqueous solution (1.195 mol/L), the mixture was stirred for 30 min. Next, 5 mL of Ca(NO₃)₂ aqueous solution (1.952 mol/L) was added dropwise into the mixture. Subsequently, adjusting pH value to 7.4 by the ammonia, the mixture was reacted for 3 h with stirring at 90 °C. Then, the system was continued to stir for 12 h at 25 °C. The product was centrifugally separated, and washed with distilled water for removing urea. After freeze-drying, HA@SPI composite was obtained.

The HA@SPI composites obtained by 3 g were calcined at the temperatures of 200 °C, 400 °C, 500 °C, 600 °C, 700 °C, 800 °C for 60 min by temperature-programmed calcination in muffle furnace to afford the hybrids and was named as SPI/HA-200, SPI/HA-400, PI/HA-500, SPI/HA-600 (HApM), SPI/HA-700, SPI/HA-800, respectively (Figure S1, Supplementary Information). The result showed that HApM (calcined at 600 °C) showed the best adsorption ability.

Fourier transform infrared (FTIR) spectra of the adsorbents were recorded between 4,000 and 400 cm⁻¹ through the KBr method with a FTS3000 spectrophotometer (FTS3000, DigiLAB Merlin, USA). Thermogravimetric analysis (TGA) was carried out at a condition of N₂ protection and temperature range being 25 °C–800 °C, heating rate being 10 °C/min (PerkinElmer, model Pyris Diamond, USA). The crystal structures and photocurrent responses were characterized by X-ray diffraction (XRD) (D/max-2400, Rigaku Corporation, Japan). A morphology analysis of certain features was visualized by using scanning electron microscopy (SEM) (ULTRA PLUS, ZEISS, Germany).

As we know, SPI is a kind of edible, low-cost and plant-based protein whose chains show a loop domain containing several helices structure (Maruyama et al. 2004). Here, using SPI as biopolymer template, the microcrystal of hydroxyapatite was crystallized along with loop protein chains and HA@SPI composite was synthesized. Subsequently, the HA@SPI was calcined at 600 °C, in the process of calcination, hydroxyapatite was mineralized and a part of SPI was carbonized. Finally, HA based hybrids (HApM) were successfully prepared (Figure 1).

The obtained hydroxyapatite-based porous materials (HApM) were characterized by FTIR spectra, and compared with HA@SPI, templates materials (SPI), pure HA,and SPI/HA-Tx (Figure S2, Supplementary Information). The results are shown in Figure 2. It indicates that HA@SPI displayed characteristic absorption peaks both of SPI (1,649 cm⁻¹, 1,535 cm⁻¹ and 1,244 cm⁻¹ are corresponding to -CONH-) and HA (PO₄³⁻ vibration bands at 1,033, 603, and 566 cm⁻¹ and -OH band at 3,441 cm⁻¹). Characteristic absorption peaks of HApM are similar to HA@SPI, but the peaks of HA in HApM are more obvious. The amide two new characteristic peaks appeared at 2,050 cm⁻¹ and 2,450 cm⁻¹ which could be caused by carbonization. These results indicated that SPI of HA@SPI was turned to activated carbon at heating condition, and hydroxyapatite-based porous material was obtained (HApM). At the same time, a sharp peak at 1,628 cm⁻¹ corresponded to H–O–H bending band, being also representative of the strongly bonded –OH groups in the matrix. The O–H stretching bands merged together and shifted to lower frequency in the spectra of HApM. This was due to the possibility of H-bonding. It is also apparent from Figure 2 that after adsorption of Pb²⁺ (HApM-after), the intensity and position of 1,628 cm⁻¹ peak was changed, and the intensity of the 3,441 cm⁻¹ peak weakens slightly, which demonstrated the involvement of –OH groups in the adsorption of Pb²⁺.

TG curves are shown in Figure 3. Compare HA@SPI with SPI and pure HA, main weight loss of HA@SPI is due to SPI degradation, but for SPI itself, there was 22% remained at 800 °C which was according the degradation of SPI, its carbonization product. And it also can be deduced that the proportion of HA and SPI is 1:1. Although the calculated temperature was as high as 800 °C, the weight of HA@SPI is only 1.5% which owed to further carbonization of HA@SPI. These results suggested that the main interaction in preparing HApM progress was due to SPI carbonization of HA@SPI (Figure S3, Supplementary Information).

The XRD patterns of biopolymer-template SPI, pure HA, composite HA@SPI and HApM were given in Figure 4. It can be observed from the figure that when 2θ is 32°, the powder of SPI/HA-200, SPI/HA-400, SPI/HA-500, SPI/HA-600 and SPI/HA-800 all have (211) crystal plane diffraction peak, and when it is near 25°, The diffraction peak of (002) crystal plane appears, which is the most important characteristic diffraction peak of HA, indicating that SPI/HA, SPI/HA-200, SPI/HA-400, SPI/HA-500, HApM and SPI/HA-800 all formed HA crystals. At the same time, it can be seen from the diagram that the diffraction peaks of (102), (210), (300) and (202) planes tend to be sharp with the increase of calcination temperature, and the diffraction peaks of each crystal plane of SPI/HA-800 are sharp. The position and intensity are consistent with the characteristic diffraction peak of HA, indicating that the crystallization degree of HA in SPI/HA-800 complex is the best and the purity of the product is very high (Figure S4, Supplementary Information).

By observing the morphology of SPI/HA-200, SPI/HA-400, SPI/HA-500, HApM, SPI/HA-700 and SPI/HA-800 under scanning electron microscope (as shown in Figure S5, Supplementary Information), it can be seen that when the calcination temperature is 200 °C, the morphology of the composite SPI/HA-200 is close to that of HA@SPI, but the pore size increases slightly. When the calcination temperature is 400 °C, 500 °C and 600 °C, the pores of SPI/HA-400, SPI/HA-500 and HApM continue to increase, and there are different degrees of agglomeration at the same time. With the increase of temperature, when the calcination temperature rises to 700 °C, the pores of the composite SPI/HA-700 collapse and have obvious agglomeration, and the microscopic morphology is closer to that of HA. When the calcination temperature rises to 800 °C, the SPI/HA-800 agglomeration of the composite is serious. From the microscopic morphology of SPI/HA-Tx series, it can be seen that the SPI in the composite is gradually degraded with the increase of calcination temperature, which is consistent with the results of TG analysis. When the calcination temperature is 600 °C, the pore structure of SPI/HA-600(HApM) is the best. At the same time, the SEM pictures of HApM before and after adsorption of Pb²⁺ were compared. It was found that after the adsorption of Pb²⁺ ions, the porous morphology of HApM was filled in different degrees because of the adsorption of Pb²⁺on the surface and pores. The energy dispersive spectroscopy (EDS) elemental analysis of HApM after adsorption of Pb²⁺ was carried out, and the results were verified (Figure 5).

The Brunauer–Emmett–Teller (BET) specific surface area (S_BET) and pore-size distribution of the products were investigated. As shown in Figure 6(a) and 6(b), the specific surface area of HApM is 123.954 m²/g, which is higher than that of HApM-after (118.317 m²/g). The results show that due to the adsorption of Pb²⁺ on the surface and pores of HApM porous materials, the specific surface product of Pb²⁺ adsorbed decreases. The results were in good agreement with EDS analysis.

Heavy metal ions in water can cause long-term and serious harm to people and the environment. In particular, Pb²⁺ can lead to endocrine and reproductive system disorders, teratogenicity and even carcinogenesis. Here, we used Pb²⁺ as model heavy metal ion, the respective adsorption ability of HA, SPI, HA@SPI and HApM were evaluated and shown in Table 1. The results showed that removal efficiency of Pb²⁺ by HApM was higher than HA, SPI and HA@SPI which implied that HAPM was an effective adsorbent.

To investigate influence of adsorption conditions, the amount of adsorbent (HApM) (0.2, 0.1 and 0.05 g), adsorption time (30, 60 and 90 min), initial concentration of Pb²⁺ (30, 50, 60 mg/L) were selected. Removal rate (%) and adsorption capacity (mg/g) were calculated by orthogonal adsorption experiments. As shown in Table 2, in the condition of adsorbent dosage was 0.1 g, initial concentration of Pb²⁺ was 60 mg/L and adsorption time was 30 min, the adsorbent achieved its maximum adsorption efficiency of 96.25% removal rate and 28.88 mg/g adsorption capacity. If the adsorption capacity was taken into account, the removal rate (90.02%) and the adsorption capacity (45.01 mg/g) were relatively high in the condition of 0.05 g adsorbent, 50 mg/L initial concentration of Pb²⁺ and 60 min adsorption time.

As shown in Table 3 and Figure 7, the parameters of Langmuir and Freundlich for the adsorption of Pb²⁺ by the HApM hybrid are obtained, which is 0.9616 and 0.9853 for Langmuir and Freundlich. It can be concluded that the Freundlich isotherm is more suitable to describe the adsorption processes of Pb²⁺ onto HApM than the Langmuir isotherm. In addition, the adsorption index 1/n of this experiment is 0.4282, which is between 0.1 and 0.5, indicating that HApM is easy to adsorb Pb²⁺.

As shown in Table 4 and Figure 8, the R² of the pseudo-first-order and pseudo-second-order kinetic parameters is 0.9269 and 0.9984, respectively. Moreover, the values of qe (25.84 mg/g) calculated from pseudo-second-order kinetics are in agree with the experimental values of qe(exp) (25.18 mg/g). It indicates that the pseudo-second-order model is appropriate for the adsorption phenomena, and chemical adsorption plays a major role in the adsorption process.

Based on above results, we considered it belongs to chemisorption. After absorption of Pb²⁺, HApM-after was characterized by SEM, FTIR and BET, and results are shown in Figures 2, 5 and 6, respectively. Comparing with HApM, in HApM-after (Figure 2), the intensity and position of 1,628 cm⁻¹ peak was changed, and the intensity of the 3,441 cm⁻¹ peak weakens slightly that demonstrated the involvement of –OH groups in the adsorption of Pb²⁺. From SEM images, it was found that, the surface morphology of HApM changed after adsorption of Pb²⁺ ions (Figure 5(e)). BET analysis also confirmed that the surface area of HApM decreased due to the adsorption of lead ion on the surface of HApM (Figure 6). Therefore, we speculated that lead ions react with hydroxyl groups so that lead ions were firmly adsorbed on the surface of hydroxyapatite.

The adsorption HApM was also compared with reported results (Table 5) (Mittal et al. 2016; Naushad 2014; Gnanasekaran et al. 2018; Pugazhendhi et al. 2018). Ghasemi et al. (2014a; 2014b) reported that the Ash, nFe-A and FSAC adsorbents have excellent adsorption capacity of Pb²⁺. Separately, and the maximum adsorption capacities are 588.24 mg/g, 833.33 mg/g, 80.645 mg/g. Naushad et al. (2015) found that the removal of Pb²⁺ got to 95% using TIV as adsorbent, and the maximum adsorption capacity was 18.8 mg/g. Le et al. (2019) used porous hydroxyapatite granules as an effective adsorbent for the Removal of Aqueous Pb(II) Ions which the removal of Pb²⁺ got to 95.9%, and the maximum adsorption capacity was 7.99 mg/g. Here, using HApM as adsorbent, the removal of lead got to 96.25%, while the adsorption capacity was 28.88 mg/g.

In summary, a kind of HA based hybrid has been successfully fabricated and tested as Pb²⁺ removers from aqueous solutions. Morphology and structure of the HApM suggest that protein chains were shank and carbonized in the calcined process and formed a highly porous HA based hybrid which have proved an excellent adsorption capacity of 96.25% for Pb²⁺. The adsorption data fitted well with Freundlich isotherm model, and being followed by the chemisorption according to the pseudosecond-order. All of those suggest that the HA based hybrid will be an ideal Pb²⁺ adsorbent for the treatment of liquid wastes containing toxic Pb²⁺ metal ion, which could provide high effectiveness with low-cost and environmental friendliness.

The project was supported by the National Natural Science Foundation of China (Grant No. 21865030; 21364012).

The Supplementary Material for this paper is available online at http://dx.doi.org/10.2166/wst.2019.370.