Mesoporous carbonated hydroxyapatite/chitosan porous materials for removal of Pb(II) ions under flow conditions

Abstract

Pb(II) ions are highly harmful to the environment and human health because of their bioaccumulation, non-degradability and toxicity. Herein, we fabricated mesoporous carbonated hydroxyapatite (HAP)/chitosan (CS) porous materials (MHCMs) by using calcium carbonate/CS porous materials (CCPMs) as precursors. After soaking in phosphate buffer solution (PBS), the CaCO₃ particles in the CCPMs were converted to carbonated HAP plates with a preferred c-plane orientation via a dissolution–precipitation reaction. The MHCMs possessed interconnected macropores with pore sizes of 50–150 μm and mesopores with pore sizes of 3.97 nm, which caused them to have a great BET surface area of 111.3 m² g⁻¹ and a pore volume of 0.199 cm³ g⁻¹. The sorption performance of the MHCMs for Pb(II) ions was evaluated by the flow of 400 mg L⁻¹ lead solutions through the adsorbents, and HAP/CS porous materials (HCPMs) served as control samples. After adsorbing Pb(II) ions, the HAP particles in both the MHCMs and HCPMs were transferred into lead hydroxyapatite (PbHAP) rods. At the adsorption equilibrium, the adsorption amounts of the MHCMs and HCPMs arrived at 559.6 and 264.4 mg g⁻¹, respectively. The MHCMs had better adsorption properties than the HCPMs because of their large surface area, hierarchically porous structures, low Ca/P ratio, good degradability and plate-like carbonated HAP with a preferred c-plane orientation. The adsorption of Pb(II) ions on MHCMs exhibited higher compliance with the pseudo-second-order kinetic model than the pseudo-first-order kinetic model, suggesting that the adsorption process was mainly controlled by chemical adsorption for Pb(II) ions. Therefore, the MHCMs have great potential for the removal of Pb(II) ions from aqueous solutions even under flow conditions.

---

1. Introduction

Heavy metal ion pollution rapidly spreads throughout the word due to the expansion of industrial activities such as tanneries, electronics, mining, electroplating and petrochemical industry.^1,2 Heavy metal ions have posed significant threats to the environment and human beings even at a low concentration because of their bioaccumulation, non-degradability and toxicity.³ Notably, the Pb(II) ion is one of the common heavy metal ions, which may cause various diseases such as brain damage and anaemia.⁴

Adsorption technology is an effective method to remove Pb(II) ions due to its low-cost, flexibility and operation simplicity.^5–8 Recently, hydroxyapatite (HAP) [Ca₁₀(PO₄)₆(OH)₂] has been widely used as an adsorbent, because it not only possesses nontoxicity to environment,^9,10 but also has good sorption capacity for heavy metal ions.^11,12 Up to now, virus HAP-based composites have been prepared for removing heavy metal ions in waste water, as shown in Table 1.^13–20 The Ca(II) ions in the crystal lattices of HAP can be substituted by other divalent metal ions such as Cu(II), Pb(II), Zn(II), Cd(II) and Ni(II).^21–25 The chemical adsorption mechanisms of HAP for heavy metal ions include ion exchange reaction, dissolution–precipitation reaction.^26–28 Ciobanu et al. reported that after the removal experiment of Pb²⁺ ions from solutions HAP nanoparticles were transformed into lead HAP (PbHAP) via the adsorption of Pb²⁺ ions followed by a cation exchange reaction.²⁷ In contrast, Dong et al. suggested that the dissolution–precipitation reaction and surface complexation were mutually responsible for Pb(II) on HAP/magnetite composite adsorbent.²⁸ Therefore, the chemical adsorption mechanism of HAP for heavy metal ions should be further investigated.

Table 1 Comparison of maximum adsorption amounts of HAP and HAP-based composite adsorbents for heavy metal ions

Adsorbents	Metal ions	Type	c0 (mg L^−1)	pH	qmax (mg g^−1)	Ref.
HAP	Pb(II)	Powder	1000–8000	3.0–5.0	330.0–450.0	13
HAP/polyacrylamide composite hydrogels	Pb(II)	Hydrogel	50–300	2–5	123.0–209.0	14
HAP/polyurethane composite foams	Pb(II)	Porous materials	44–184	5.0	150.0	15
HAP/Fe3O4 microspheres	Pb(II)	Powder	600	3.0	440	16
NanoHAP–alginate composite adsorbents	Pb(II)	Powder	900	6.2–7.1	270.3	17
Nitrilotris(methylene)triphosphonate-modified HAP	Pb(II)	Powder	10–2000	5.0	640	18
Nitrilotris(methylene)triphosphonate-modified HAP	Zn(II)	Powder	10–2000	5.0	300	18
Nano-HAP	Pb(II)	Powder	100–400	5.0–6.0	700	19
Nano-HAP	Cd(II)	Powder	100–400	5.0–6.0	142	19
Nano-HAP	Ni(II)	Powder	100–400	5.0–6.0	36.25	19
Porous nano-HAP	Pb(II)	Powder	100	2.0–6.0	40.04	20
Porous nano-HAP	Cu(II)	Powder	100	2.0–6.0	99.94	20

---

The adsorption performances of HAP are mainly related to its crystallinity, crystallographic orientation and porous structure. Carbonated HAP with a low crystallinity possesses a large amount of lattice defects and high degradability; thus, the Ca²⁺ ions in HAP are easily replaced by other metal ions via a dissolution–precipitation reaction or an ion exchange reaction.^27,28 As we know, hexagonal HA crystals with a P6₃/m space group have two crystal planes including a(b)-plane (ac and bc crystal faces) and c-plane (ab crystal face).^29–31 The a(b)-planes of HA crystals are rich in calcium ions, while the c-planes are rich in phosphate and hydroxide ions.^29–31 Moreover, hierarchically porous HA provides large surface areas and pore volumes to accelerate the adsorption of heavy metal ions. Unfortunately, hierarchically porous carbonated HAP nanoplates with low crystallinity and c-plane orientation have rarely reported.

Besides HAP, chitosan (CS) as an environment-friendly material has been widely used as an adsorbents in the long-term disposal of heavy metal ions such as copper, lead, mercury, cadmium and chromium.^32,33 CS consists of β-(1,4)-2-acetamido-2-β-D-glucose and β-(1,4)-2-amido-2-β-D-glucose units. The amine groups in CS have strong affinity to various heavy metal ions, resulting in the formation of complexes.³³ Both HAP and CS possess good adsorption property to remove heavy metal ions, so many HAP/CS composite particles have been reported.^34,35 However, the difficulty in separating the HAP/CS composite particles from wastewater limits their industrial application.

Herein, we report for the first time the synthesis of mesoporous carbonated HAP/CS porous materials (MHCMs) for the removal of Pb(II) ions from aqueous solutions. The synthesis steps include: (i) the fabrication of CaCO₃/CS porous materials (CCPMs) by a freeze-drying technology; and (ii) the conversion of MHCMs from CCPMs by treatment with a phosphate buffer solution (PBS). The main aims of this work are to fabricate the MHCMs, investigate their morphology, porous structure, crystal orientation and formation mechanism, and study their adsorption properties and adsorption kinetics for Pb(II) ions.

2. Experimental

2.1 Materials and reagents

CS (90% deacetylated form) and absolute ethanol (AR) were purchased from Richjoint Chemical Reagent Co., Ltd. (Shanghai, China). Calcium carbonated (99.9%, AR), acetic acid (99.9%, AR), sodium dihydrogen phosphate dehydrate (99.9%, AR), disodium hydrogen phosphate dodecahydrate (99.9%, AR), nitric acid (65%, AR) and lead nitrate (99%, AR) were all purchased from Aladdin. Sodium hydroxide pellets (AR, 98%) was purchased from Merck. Deionized water obtained in our laboratory (Sartorius AG, Arium 611Ultra filter, Germany) was used throughout the experiment with a resistivity of 18.2 MΩ.

2.2 Preparation of MHCMs

A homogeneous CS solution (1.96 wt%, 50 mL) was prepared by the addition of CS powders (1.00 g) into an acetic acid solution (50.0 mL, 1.00 vol%). The solution was continuously stirred for 12 h and sonicated for 1 h to remove air bubbles. 25.5 g of CC (CaCO₃) powders were added into the above CS solution under continuous agitation for 2 h. The above mixtures were added into moulds and cooled at −20 °C in freezer for 5 h. The samples were rapidly transferred to a freeze-drying machine to deep freeze at −35 to −40 °C. The solidified mixtures were subsequently freeze-dried for 42 h. After the freeze-drying process, the products (CCPMs) were treated with a NaOH solution (100 mL per well, 10.0 wt%) for 24 h, washed with deionized water till the pH = 7.0, and dried under vacuum dryer again. Then, the CCPMs were treated with a PBS at 37 °C for 21 days. The concentration and pH value of PBS were 0.2 mol L⁻¹ and 7.4, respectively. Finally, MHCMs were obtained after washed with deionized water and dried under vacuum dryer.

The HA/CS porous materials (HCPMs) as control samples were synthesized under the same conditions except without PBS treatment. In brief, 25.5 g of HAP nanoparticles were added into the CS solution under continuous agitation for 2 h. The above mixtures were added into moulds and cooled at −20 °C in freezer for 5 h. The HCPMs were obtained by the freeze-drying process, washed with deionized water and dried under vacuum dryer, whose preparation method was similar to the CCPMs.

2.3 Adsorption experimental for Pb(II) ions

A Pb(II) aqueous solution (400 mg L⁻¹) was prepared after lead nitrate powders were dissolved in deionized water. The pH value of the solutions was adjusted at 2.5, 4.0 or 5.5 by a HNO₃ solution (0.1 M) or a NaOH solution (0.1 M) at 25 °C. The MHCMs (0.5 g) and HCPMs (0.5 g) were cut into a cylinder (Φ 12 mm), respectively. After the adsorbents were put into sample tube, the Pb(II) solutions were flowed at a flow rate of 0.8 mL min⁻¹. At the given time intervals, the concentrations of Pb(II) ions in the solution were determined by inductively coupled plasma/optical emission spectrometry (ICP/OES, Perkin Elmer, OPTIMA 3300 DV). The cumulative adsorption amount (q_t) was calculated according to the following equations:where c₀ (mg L⁻¹) was the initial concentration of Pb(II) ions, c_t was the concentration at a given time, m was the weight of the HCPMs or MHCMs, and V is the volume of outflow solutions.

2.4 Characterization

The morphologies of samples were investigated by scanning electron microscopy (SEM, Hitachi S-4800, CamScan) with energy-dispersive spectrometry (EDS), after they were sputter-coated with gold. The phases of samples were examined by X-ray powder diffraction (XRD, D/max-II B, Japan) using CuKα radiation (λ = 1.541874 Å) within the scanning range of 2θ = 10° to 80° at a scanning rate of 4° min⁻¹ and a step size of 0.02°. Fourier transform infrared spectra (FTIR, Nicolet 5DX) were carried out to analyze the functional groups of samples in the range of wavenumbers 4000–400 cm⁻¹ at a resolution of 2 cm⁻¹ (number of scans ∼60) by using the KBr pellet technique at room temperature. N₂ adsorption–desorption isotherms of samples were measured with an automatic surface area and porosity analyzer (AUTOSORB-1-C, Quantachrome). The pore size distributions were derived from desorption branches of isotherms by using the Barrett–Joyner–Halenda (BJH) method. The porosity (π) of the samples MHCMs were measured by liquid displacement method.³⁶ The samples were immersed into absolute ethanol under vacuum conditions until they were saturated by absorbing ethanol. The porosity of the porous materials was calculated according to the formula:where V₀ and W₀ were the volume and weight of dry MHCMs and HCPMs, respectively, V₁ and W₁ were the volume and weight of MHCMs and HCPMs saturated by absorbing ethanol, respectively, and ρ was the density of ethanol.

3. Results and discussion

3.1 Characterization of MHCMs before adsorption of Pb(II) ions

3.1.1 Morphology of MHCMs before adsorption of Pb(II) ions. Both HAP and CS possess excellent adsorption property for heavy metal ions.^34,35 Unfortunately, it is difficult to separate the HAP or CS particles from wastewater. Herein, we reported the synthesis of MHCMs by using CCPMs as sacrificial templates (Scheme 1a and b). The preparation processes of MHCMs included two steps: (i) the formation of the CCPMs by freeze-drying technology; and (ii) the conversion of the MHCMs from the CCPMs by PBS treatment.

Scheme 1 Illustration of MHCMs preparation and chemical adsorption of Pb(II) ions: (a) formation of CCPMs by freeze-drying process; (b) conversion of CCPMs into MHCMs after treatment with PBS solution; (c) the chemical adsorption of Pb(II) ions on MHCMs to form PbHAP.

At the first stage, the CCPMs were fabricated by the dispersion of CaCO₃ and CS powders in an acetic acid solution followed by the freeze-drying process. In addition, the HCPMs including HAP and CS were prepared under the same condition but the replace of CaCO₃ powders by HAP powders. The low-resolution SEM image indicated that both the HCPMs and CCPMs had possessed 3D interconnected macropores (Fig. 1a and e). The macropores were derived from the ice crystals, which were produced in the solidification process during cooling process. After the vaporization of the ice crystals, the macropores were formed. Interestingly, the macropore sizes of the HCPMs were 100–300 μm, while those of the CCPMs were only 50–150 μm. The smaller pore size of the CCPMs than the HCPMs was attributed to the bigger particle size of CaCO₃ powders than HAP powders. The high-resolution SEM image indicated that many HAP nanoparticles with particle sizes of ∼20 nm existed on and within the HCPMs films (Fig. 1b and c). In contrast, the calcium carbonate powders in the CCPMs had big particle sizes of 0.2–1.0 μm (Fig. 1f and g). The aggregates of the CaCO₃ powders barricade partly macropores in the CCPMs.

Fig. 1 SEM images and EDS spectra of the porous materials: (a–d) HCPMs; (e–h) CCPMs; (i–l) MHCMs.

At the second stage, the MHCMs were converted from the CCPMs after treatment with PBS at 37 °C for 21 days (Scheme 1a and b). The MHCMs possessed 3D interconnected porous structure with pore sizes of 50–150 μm (Fig. 1i), which was similar to that of the HCPMs (Fig. 1a). The total porosities of HCPMs and MHCMs were 91.76% and 90.89%, respectively. The high porosities were attributed to the interconnected macropores structure, which played an important role in adsorbing heavy metal ions. The high-resolution SEM images revealed that the MHCMs exhibited the different morphologies from the CCPMs (Fig. 1). After soaking the CCPMs in PBS for 21 days, the CaCO₃ particles were converted into HAP nanoplates by using the original CaCO₃ particles as active sites. The as-obtained carbonated HAP nanoplates were perpendicular to the particle surfaces, and coalesced to form a flower-like structure with an aperture size of 0.2–1.0 μm (Fig. 1j and k). The plate-like carbonated HAP aggregates were connected together (Fig. 1j), and the macropores in the CCPMs were remained (Fig. 1e and i). The chemical elements of the HCPMs, CCPMs and MHCMs were analyzed by the corresponding EDS spectra (Fig. 1d, h and l). The chemical elements of the HCPMs were mainly composed of Ca, P, C and O. The Ca and P elements were derived from the HAP particles in the HCPMs, the O element was derived from CS and HAP particles, and the C element was derived from CS (Fig. 1d). The average Ca/P molar ratio of the HCPMs was 1.67. The CCPMs were composed of calcium carbonate and CS. The Ca element was derived from CaCO₃ particles, and the C and O were derived from both CaCO₃ particles and CS. The chemical elements of the MHCMs were mainly composed of Ca, P, C, O and Na. The Ca and P elements were derived from the carbonated HAP in the MHCMs, the O element was derived from CS and carbonated HAP, the C element was attributed to the CS or the CO₃²⁻ in the carbonated HAP crystal lattices, and the Na element was attributed to either the part substitution of Ca²⁺ ions by Na⁺ ions in the carbonated HAP crystal lattices or the adsorbed Na⁺ ions from PBS. The average Ca/P molar ratio of the MHCMs was 1.51, which was lower than that of stoichiometric HAP (Ca/P ratio = 1.67). The above result suggested that the carbonated HAP in the MHCMs was calcium-deficient apatite.

The morphology and structure of the carbonated HAP in the MHCMs were characterized by the TEM images and corresponding ED pattern (Fig. 2a–c). The low-resolution TEM image revealed that the carbonated HAP nanoplates coalesced to form flower-like particles (Fig. 2a), which was in good agreement with the SEM images (Fig. 1j). Notably, the high-resolution TEM image indicated that these nanoplates were composed of many smaller nanocrystals with a particle size of 4–7 nm (Fig. 2c). The corresponding ED pattern showed the visible diffraction rings due to apatite structure (Fig. 2b, inset). The aggregates of the above-mentioned nanocrystals resulted in the formation of the mesopores among them. The mesoporous structure of the MHCMs was demonstrated further by the N₂ adsorption–desorption isotherm and corresponding BJH pore size distribution curve. The N₂ adsorption–desorption isotherm curve of the MHCMs was identified as type IV isotherms with type H3 hysteresis loop (Fig. 2d).³⁷ This H3 hysteresis loop did not exhibit any limiting adsorption at high P/P_o 0.9–1.0, which was attributed to the aggregation of particles giving rise to slit-shaped pores.³⁷ The corresponding BJH pore size distribution curve indicated that the mesopore size of the MHCMs was mainly distributed around 3.97 nm (Fig. 2d, inset). The steep increase of N₂ adsorption suggested the presence of macropores, which was consistent with the SEM images (Fig. 1i–k). The BET surface areas and the pore volumes of the MHCMs were 111.3 m² g⁻¹ and 0.199 cm³ g⁻¹, respectively.

Fig. 2 (a–c) TEM images; (d) N₂ adsorption–desorption isotherm of the sample MHCMs; the inset in (b) shows the ED pattern, and the inset in (d) shows the BJH pore size distribution.

3.1.2 Structure of MHCMs before adsorption of Pb(II) ions. The phases of the MHCMs, HCPMs and CS were characterized by XRD patterns (Fig. 3). The CCPMs were composed of CS and CaCO₃ in the phase of calcite. After soaking in PBS at 37 °C for 21 days, the CCPMs were converted into the MHCMs, and the characteristic peaks of only HAP without calcite were detected in the XRD pattern (Fig. 3c). The broad peaks suggested that the HAP in the MHCMs had low crystallinity. CS is a semi-crystalline material, and its characteristic peaks locate at 2θ = 20.30° and 28.11° (Fig. 3a). Although both the MHCMs and HCPMs consisted of HAP and CS, no characteristic peaks of CS were detected in the XRD patterns (Fig. 3b and c). The reasons may be attributed to the following aspects: (i) CS is a semi-crystalline material, so its peak strength is lower than HAP; (ii) the percentage of CS in the HCPMs and MHCMs is lower than that of HAP; and (iii) the HAP particles in porous materials may weaken the intermolecular interactions among CS chains, and thus reduce its crystallinity.

Fig. 3 XRD patterns of the porous materials before adsorption of Pb(II) ions: (a) CS powders; (b) HCPMs; (c) MHCMs.

The functional groups of the CS, HCPMs and MHCMs were detected in the FTIR spectra (Fig. 4). For the CS powders, the characteristic absorption bands at 3220 and 3420 cm⁻¹ corresponded to the elongation of N–H and O–H bonds.³⁸ The bands at 2920 and 2870 cm⁻¹ were ascribed to the C–H stretching vibration in –CH₂ groups.^39,40 The bands at 1655 and 1604 cm⁻¹ were assigned to the C=O bond stretching of amide I and N–H deformations of amide II, respectively. The bands at 1400, 1318 and 1261 cm⁻¹ were represented C–H bending vibration, CH₃ symmetric stretching vibration and C–O–H stretching vibration, respectively.⁴¹ The bands at 1154 and 1036 cm⁻¹ were assigned to C–O–C stretching vibration modes.⁴² For both the HCPMs and MHCMs, the intense adsorption peak at 1034 cm⁻¹ was ascribed to the stretching vibration (v₃) of the phosphate (PO₄³⁻) groups, and the absorption peaks at 563 and 604 cm⁻¹ were ascribed to the bending vibration (v₄) of the phosphate (PO₄³⁻) groups.⁴³ The C=O stretching band at 1655 cm⁻¹ shifted to a lower band 1637 cm⁻¹ and the band at 1604 cm⁻¹ disappeared, suggesting that the interaction may take place between CS and HAP. Interestingly, the bands at 874 cm⁻¹ (v₂) and 1476 cm⁻¹ (v₃) were ascribed to the B-type CO₃²⁻ substitution (Fig. 4c).⁴⁴ Moreover, the absorption band at around 1103 cm⁻¹ was corresponding to HPO₄²⁻ groups, indicating that the apatite in the MHCMs was a calcium-deficient HAP.⁴⁵ Thus, the general formula of apatite in MHCMs might be expressed as Ca_10−x−y/2(HPO₄)_x(PO₄)_6−x−y(CO₃)_y(OH)_2−x. Moreover, the characteristic peaks due to CS were also detected in Fig. 4b and c, although some peaks were overlapped by those of the HAP in the MHCMs and HCPMs.

Fig. 4 FTIR spectra of the porous materials before adsorption of Pb(II) ions: (a) CS powders; (b) HCPMs; (c) MHCMs.

3.1.3 Formation mechanism of MHCMs. After soaking in PBS at 37 °C for 21 days, the CCPMs were converted into the MHCMs (Scheme 1a and b). Since the logarithmic solubility product of calcite (pK_sp = 8.56) is lower than that of HAP (pK_sp = 58.63), HAP is more stable in solutions than CaCO₃.^46,47 The CaCO₃ particles in the CCPMs can be transformed to carbonated HAP nanoplates in PBS (Fig. 1–4). The remarkable differences of phases and morphologies between the CCPMs and MHCMs suggested that the conversion mechanism of the CCPMs to the MHCMs was a dissolution–precipitation reaction. After soaking in PBS, Ca²⁺ ions were dissolved from the surfaces of calcium carbonate particles according to the following reaction:

CaCO₃ → Ca²⁺ + CO₃²⁻ (3)

The Ca²⁺ and CO₃²⁻ ions released from the CaCO₃ particles increased the local ions concentrations around the crystals, and reacted with PO₄³⁻ and OH⁻ ions in PBS to form carbonated HAP particles as its activity product exceeded its thermodynamic solubility product.

(10 − x − y/2)Ca²⁺ + xHPO₄²⁻ + (6 − x − y)PO₄³⁻ + yCO₃²⁻ + (2 − x)OH⁻ → Ca_10−x−y/2(HPO₄)_x(PO₄)_6−x−y(CO₃)_y(OH)_2−x (4)

The carbonated HAP particles tended to deposit on the porous materials rather than in PBS because of the following reasons: first, the concentrations of Ca²⁺ and CO₃³⁻ ions around the porous materials were higher than those in the solution; second, the functional groups such as –OH and –NH₂ in CS could serve as active sites to promote the deposition of carbonated HAP crystals. Notably, the carbonated HAP particles exhibited plate-like shapes rather than rod-like shapes (Fig. 1k and 2a). As we know, hexagonal HA crystals with a P6₃/m space group have two crystal planes including a(b)-plane (ac and bc crystal faces) and c-plane (ab crystal face).^29–31 The a(b)-planes of HA crystals are rich in calcium ions, while the c-planes are rich in phosphate and hydroxide ions.^29–31 During the conversion process of the CCPMs to the MHCMs, the molar ratio of PO₄³⁻ ions in PBS to Ca²⁺ ions released from the CCPMs was over 0.6. The excessive PO₄³⁻ ions over Ca²⁺ ions tented to promote the formation of HAP plates with preferred c-plane orientation. The EDS spectrum confirmed the above conclusion (Fig. 1l). The Ca/P molar ratio of carbonated HAP in the MHCMs was about 1.51, which was lower than the stoichiometric HAP (Ca/P = 1.67). The 3D interconnected macropores and CS in the porous materials were remained as the CCPMs were converted into the MHCMs in PBS. In contrast, the HCPMs were originated from the mixed solutions including CS and HAP nanoparticles via the freeze-drying process. The HAP nanoparticles existed on and within the HCPMs films. Therefore, the morphologies of HCPMs are obviously different from the MHCMs (Fig. 1).

3.2 Characterization of MHCMs after adsorption of Pb(II) ions

3.2.1 Morphology of MHCMs after adsorption of Pb(II) ions. The MHCMs including carbonated HAP and CS were used as adsorbents for Pb(II) ions, and the adsorption experimental was carried out by the flow of a lead solution throughout the MHCMs (Scheme 1b and c). In addition, the HCPMs possess a great potential for the removal of Pb(II) ions from aqueous solutions, which served as a control samples.⁴⁸ The HCPMs possessed 3D interconnected porous structure, and the HAP nanoparticles aggregated on and within the CS films (Fig. 1a–c). After the adsorption of Pb(II) ions, the 3D interconnected macropores in the HCPMs were remained without obvious changes, and many lead hydroxyapatite [PbHAP, Pb₁₀(PO₄)₆(OH)₂] nanorods deposited on the films (Fig. 5a–d). The PbHAP nanorods possessed different particle sizes. Most of nanorods had large diameters of approximately 100 nm, and the others had small diameters of approximately 20 nm. Moreover, the morphologies of the MHCMs were obviously different between before and after the adsorption of Pb(II) ions. Before the adsorption of Pb(II) ions, the MHCMs possessed 3D interconnected porous structure, and the carbonated HAP plates were covered on the films (Fig. 1i–k). After the adsorption of Pb(II) ions, the carbonated HAP plates were converted into PbHAP nanorods with diameters of approximately 50 nm and lengths of several micrometers (Fig. 5a–c). The macropores in the porous materials were filled partly with the PbHAP nanorods because of their long lengths, resulting in decreasing the macropore size (Fig. 5d).

Fig. 5 (a–c) SEM images of the HCPMs after adsorption of Pb(II) ions for 7 days; (d–f) SEM images of the MHCMs after adsorption of Pb(II) ions for 7 days.

3.2.2 Structure of MHCMs after adsorption of Pb(II) ions. Both the MHCMs and HCPMs were composed of HAP and CS (Fig. 3 and 4). After adsorbing Pb(II) ions, the characteristic peaks due to PbHAP (JCPDS card no. 08-0259) were detected in the XRD patterns of both samples (Fig. 6). At the same time, the characteristic peaks of HAP were also observed, but their peak strength decreased as compared with those before adsorption of Pb(II) ions (Fig. 3), suggesting that part HAP particles in both the MHCMs and HCPMs were converted to PbHAP. However, the characterized peaks due to CS in the composite porous materials were not observed because of its low percentage and crystallinity.

Fig. 6 XRD patterns of samples after adsorption of Pb(II) ions: (a) HCPMs; (b) MHCMs.

The functional groups of the MHCMs and HCPMs after the adsorption of Pb(II) ions were characterized by FTIR spectra (Fig. 7). For both the MHCMs and HCPMs after adsorbing Pb(II) ions, the intense adsorption peak at 1034 cm⁻¹ was ascribed to the stretching vibration (v₃) of phosphate (PO₄³⁻) groups. The bands at 563 and 604 cm⁻¹ corresponded to the bending vibration (v₄) of phosphate (PO₄³⁻) groups.⁴³ Notably, for the HCPMs after adsorbing Pb(II) ions, the band at 1400 cm⁻¹ due to CS was shifted to 1384 cm⁻¹ (Fig. 6a and 7a), indicating that the Pb(II) ions may associate with CS to form CS–Pb complex. However, there were no characteristic peaks of CS–Pb complex in the MHCMs (Fig. 7b). The reason may be attributed to the absence of CS–Pb complex in the MHCMs. The adsorption of Pb(II) ions was mainly ascribed to the conversion of HAP into PbHAP. Taken together, the SEM images, XRD patterns and FTIR spectra revealed that the HAP in both the MHCMs and HCPMs were partly converted to PbHAP after adsorbing Pb(II) ions.

Fig. 7 FTIR spectra of samples after adsorption of Pb(II) ions: (a) HCPMs; (b) MHCMs.

3.3 Adsorption property and mechanism of Pb(II) ions on MHCMs

3.3.1 Adsorption performance of Pb(II) ions on MHCMs. The absorption experiment were carried out by the flow of a lead solution throughout the adsorbents, and the HCPMs served as control samples. The flow rate was controlled at 0.8 mL min⁻¹. The concentration and pH of the Pb(II) aqueous solutions were 400 mg L⁻¹ and 5.5, respectively. The time-dependent adsorption behaviors of Pb(II) ions on the HCPMs and MHCMs were presented in Fig. 8a. The adsorption processes of Pb(II) ions on the HCPMs included an initial steeply stage from 0 to 48 h and a relative slow stage from 48 to 144 h. After 168 h, the adsorption equilibrium were reached, and the corresponding maximum adsorption amount arrived at 264.4 mg g⁻¹. Interestingly, the adsorption amount of Pb(II) ions on the MHCMs were grater than those on the HCPMs at different time, especially after adsorption for 24 h (Fig. 8a). The cumulative adsorbed amount of Pb(II) on the MHCMs increased with time, and reached a maximum adsorption amount of 559.6 mg g⁻¹ after 456 h. In addition, the re-use abilities of MHCMs have been tested, as shown in Fig. S1.† The MHCMs exhibit excellent re-use abilities for adsorbing Pb(II) ions. Even after recycling adsorption four times, the MHCMs still possess high maximum adsorption amounts of 391.1 mg g⁻¹, although the adsorption amounts gradually decrease with increasing the cycle times (Fig. S1†).

Fig. 8 (a) Cumulative adsorption amounts of Pb(II) ions on HCPMs and MHCMs at different time; (b) adsorption rates of Pb(II) ions on HCPMs and MHCMs at different time; (c) release amounts of Ca(II) ions from HCPMs and MHCMs at different time; (d) comparison of adsorption amounts of Pb(II) ions and corresponding release amounts of Ca(II) ions on HCPMs and MHCMs at different time.

Although the cumulative adsorption amounts of Pb(II) ions on the HCPMs and the MHCMs increased as prolonging the adsorption time, the adsorption rates decreased gradually (Fig. 8b). Both the HCPMs and the MHCMs had a similar trend of adsorption rates. Unfortunately, the HCPMs had the bad adsorption property for Pb(II) ions, and the adsorption rate reduced to a constant after 168 h. For the MHCMs, the adsorption rate of Pb(II) ions decreased rapidly during the first 24 h, and then decreased slowly upon increasing further the time to reach the equilibrium (Fig. 8b). Notably, the cumulative adsorption amounts of Pb(II) ions on both the HCPMs and MHCMs had the similar trends to the cumulative release amounts of Ca(II) ions (Fig. 8a and c). The equilibrium release amounts of Ca(II) ions on the HCPMs and the MHCMs were 50.42 and 77.50 mg g⁻¹, respectively. The molar amount of the Ca(II) ions released from the HCPMs was equal to that of the adsorbed Pb(II) ions before 24 h, and the former was slightly lower than the latter with further increasing the adsorption time (Fig. 8d). However, the molar amounts of the Ca(II) ions released from the MHCMs were lower than that of the Pb(II) ions at different time, especially after adsorption for 24 h (Fig. 8d). The significant differences should be attributed to the different compositions and structures of HAP in the MHCMs and HCPMs.

3.3.2 Adsorption mechanism of Pb(II) ions on MHCMs. Both the HCPMs and MHCMs possessed the excellent adsorption properties for Pb(II) ions, and their maximum adsorption amounts were 264.4 and 559.6 mg g⁻¹, respectively (Fig. 8). The XRD patterns and FTIR spectra indicated that the main components of the HCPMs and MHCMs are CS and HAP (Fig. 3 and 4). Our previous work demonstrated that CS could adsorb Pb²⁺ ions via its binding sites such as amino (–NH₂) and secondary alcohol (–OH) functional groups.⁴⁸ The nitrogen in amino group and the oxygen in hydroxyl group had an available pair of electrons to form coordinated covalent bonds with Pb²⁺ ions. The oxygen atom in hydroxyl group had a stronger attraction to its electron lone pairs than the nitrogen atom in amino group, so the amino groups were more likely to donate the lone pairs to metal ions. However, the good adsorption properties of the HCPMs and MHCMs were mainly attributed to HAP particles. The above conclusion were demonstrated by the fact that the release molar amounts of Ca(II) ions from the HAP in the HCPMs were similar to the adsorption molar amounts of Pb(II) ions (Fig. 8d). The HAP nanoparticles were dispersed within or on the HCPMs, while the carbonated HAP nanoplates in the MHCMs covered the whole porous materials (Fig. 1). Interestingly, many PbHAP rods were formed on both the HCPMs and MHCMs (Fig. 5 and 6), after the Pb(II) aqueous solutions flowed through the above porous materials by using the macropores as channels. For both the HCPMs and MHCMs, there existed the remarkable changes of the morphologies and phases between before and after the adsorption of Pb(II) ions, suggesting that the HAP particles in the porous materials were converted into PbHAP rods via a dissolution–precipitation reaction. Since the logarithmic solubility product constant of PbHAP (pK_sp = 125.6) at 25 °C is lower than that of HAP (pK_sp = 76.3), HAP is easily converted into PbHAP.⁴⁹ After the lead solutions flowed through the porous materials, the dissolution reaction of HAP in the HCPMs and MHCMs took place. The released PO₄³⁻ and OH⁻ ions increased the local ion concentrations around the HAP particles in the porous materials, and reacted with Pb²⁺ ions to form PbHAP as its ionic activity product exceeded the thermodynamic solubility product. The formation mechanism of PbHAP could be expressed as the following equation:

10Pb²⁺ + 6PO₄³⁻ + 2OH⁻ → Pb₁₀(PO₄)₆(OH)₂ (5)

Although the HAP particles in both the HCPMs and MHCMs were converted into the PbHAP rods, the morphologies of the products are different between them (Fig. 5). After adsorption of Pb²⁺ ions, the as-obtained PbHAP rods in the MHCMs exhibited the smaller diameter and longer length than those in the HCPMs (Fig. 5). The reason may be attributed to the different morphologies, Ca/P ratios and degradabilities of the HAP particles in the MHCMs from those in the HCPMs. The carbonated HAP particles in the MHCMs exhibited the plat-like shape with a Ca/P ratio of 1.51, while the HAP particles in the HCPMs exhibited the spherical shape with a Ca/P ratio of 1.67 (Fig. 1). The PO₄³⁻ ions could serve as nucleation sites to chemically adsorb Pb(II) ions form PbHAP rods. As we know, the formation process of PbHAP crystals could be divided into two stages, including nucleus formation and nucleus growth. The more amounts of the PO₄³⁻ ions in the MHCMs than in the HCPMs promoted the nucleus formation rather than nucleus growth, resulting in the greater number and smaller diameter of the PbHAP rods in the MHCMs than those in the HCPMs (Fig. 5). Moreover, the degradability of the HAP particles had a great effect on the morphology of the PbHAP rods. Previous research demonstrated that the nucleation rate was proportional to the relative supersaturation, while nucleus growth rate was proportional to the absolute supersaturation.^50,51 The absolute saturation had a greater effect on the nucleation rate than on the nucleus growth rate because of the small solubility of PbHAP. The ions, which played important effects on the formation of PbHAP rods, were composed of Pb²⁺, PO₄³⁻, and OH⁻ ions. Since the concentrations of the Pb²⁺ and OH⁻ ions could be considered as constants in the Pb(II) solutions at pH 5.5, the absolute supersaturation was determined by the concentrations of PO₄³⁻ ions. The concentrations of PO₄³⁻ ions released from the MHCMs were higher than those from the HCPMs because the carbonated HAP particles in the former had greater degradability than the HAP particles in the latter. The higher absolute supersaturation around the MHCMs than around the HCPMs resulted in the greater nucleation rate of PbHAP rods. Therefore, the PbHAP rods in the MHCMs exhibited the smaller diameter and longer length than those in the HCPMs were ascribed to the more nucleation sites and greater nucleation rates for PbHAP rods in the former than the latter.

Although both the MHCMs and HCPMs had the good adsorption properties for Pb(II) ions, the former possessed the greater adsorption amount than the latter. The reasons were attributed to the different surface areas, porous structures, degradabilities, Ca/P ratios and morphologies of the HAP in the HCPMs and MHCMs. Firstly, the great surface areas and porous structures tented to improve the adsorption amount of the MHCMs. Both the HCPMs and MHCMs possess the three-dimensional porous structure with porosities of 91.76% and 90.89%, respectively. Although the HCPMs had three-dimensional macropores with a pore size of 100–300 μm, they did not possess mesoporous structure. Interestingly, the carbonated HAP plates in the MHCMs have mesoporous structure with a pore size of ∼3.97 nm (Fig. 2c and d). The macropores and mesopores in the MHCMs rapidly increased their surface areas, and thus provided many active sites to accelerate the adsorption of Pb²⁺ ions. Secondly, the good degradability of HAP in the MHCMs increased the chemical adsorption property for Pb²⁺ ions. The HAP in the HCPMs was stoichiometric ratio, while that in the MHCMs was nonstoichiometric B-type carbonated HAP. The nonstoichiometric B-type carbonated HAP had better degradability than the stoichiometric HAP, which was confirmed by the experimental result that the release amounts of Ca²⁺ ions from the MHCMs were greater than those from the HCPMs at different time (Fig. 8d). As compared with the HCPMs, the more PO₄³⁻ ions released from the MHCMs can attract more Pb²⁺ ions to form PbHAP rods. Thirdly, the Ca/P ratio of HAP had a great effect on the adsorption property of adsorbents. The chemical adsorption capacity of HAP for heavy metal ions were mainly ascribed to the PO₄³⁻ ions rather than the Ca²⁺ ions in HAP crystal lattices. After adsorbing Pb²⁺ ions, the HAP were converted into PbHAP. The EDS spectra indicated that the Ca/P ratios of HAP in the MHCMs and HCPMs were 1.51 and 1.67, respectively (Fig. 1d and l). The lower Ca/P ratio of HAP in the MHCMs than that in the HCPMs made the former have more PO₄³⁻ ions to chemically adsorb Pb²⁺ ions than the latter. Finally, the morphology of the HAP affected the adsorption property of adsorbents. The a(b)-planes of HA crystals are rich in calcium ions, while the c-planes are rich in phosphate and hydroxide ions. The carbonated HAP plates in the MHCMs exhibited preferred c-plane orientation, while the HAP nanoparticles in the HCPMs did not exhibit good crystal orientation. The rich phosphate ions in the c-planes of HAP plates provided more active sites to adsorb heavy metal ions.

In addition, the original concentrations and pH values of Pb(II) aqueous solutions have great effects on the adsorption activities of MHCMs. Fig. S2† indicates that as the original Pb(II) concentrations are 200, 400 and 800 mg L⁻¹, the cumulative adsorption amounts of Pb(II) ions on the MHCMs after adsorbing for 456 h arrive at 387.9, 559.6 and 848.30 mg g⁻¹, respectively. These results suggest that the increase of Pb(II) concentrations can improve the adsorption amounts, which may be attributed to the possibility that the high Pb(II) ion concentrations improve the concentration gradient from the Pb(II) solutions to the adsorbent surfaces, and thus facilitate the adsorption of Pb(II) ions on the MHCMs. Moreover, decreasing the pH values of Pb(II) aqueous solutions can enhance the adsorption properties of the MHCMs. The main adsorption mechanism of the MHCMs is the dissolution of HAP particles followed by the precipitation of PbHAP rods. The decrease of pH values from 5.5 to 2.5 may accelerate the dissolution of HAP and promote the formation of PbHAP rods on porous materials via a dissolution–precipitation reaction, so the Pb(II) adsorption amounts on the MHCMs increase at different time points (Fig. S3†).

3.3.3 Adsorption kinetics of Pb(II) ions on MHCMs. In order to demonstrate the adsorption behaviours of Pb(II) ions on the HCPMs and MHCMs, adsorption kinetics were investigated by Lagergren's pseudo-first-order kinetic model and chemisorption pseudo-second-order kinetic model.^26,52,53 The two kinetic models were used to describe the adsorption of solid/liquid systems, which could be simply expressed in the linear forms.

Pseudo-first-order model:

Pseudo-second-order model:

where q_t (mg g⁻¹) was the adsorption amount of Pb(II) ions on adsorbents at any time t (h), q_e (mg g⁻¹) and q_e,cal (mg g⁻¹) were the experimental equilibrium adsorption amount and theoretical equilibrium adsorption amount, respectively, k₁ (h⁻¹) and k₂ (g mg⁻¹ h⁻¹) were the adsorption rate constant for pseudo-first-order and pseudo-second-order kinetic model, respectively.

The kinetic adsorption results of pseudo-second-order and pseudo-first-order kinetic models were shown in Fig. 9, and the corresponding parameters were summarized in Table 2. For the MHCMs, the correlation coefficients (R²) of the pseudo-first kinetic model and pseudo-second-order kinetic model were 0.96179 and 0.98839, respectively. For the HCPMs, the correlation coefficients (R²) of the pseudo-first-order kinetic model and pseudo-second-order kinetic model were 0.92884 and 0.99648, respectively. The greater correlation coefficient (R²) of the pseudo-second-order kinetic model than the pseudo-first-order kinetic model suggested that both the HCPMs and MHCMs exhibited higher compliance with the pseudo-second-order kinetic model (Fig. 9a and b). The reason was attributed to the chemical adsorption of Pb(II) ions on the HCPMs and MHCMs. After the flow of Pb(II) ions solutions through the adsorbents, PbHAP rods deposited on the HCPMs and MHCMs. However, the pseudo-first-order kinetic model was mainly used to fit simple physical adsorption of heavy metal ions.

Fig. 9 (a) Pseudo-first-order kinetic simulation of Pb(II) ions adsorption on HCPMs and MHCMs; (b) pseudo-second-order kinetic simulation of Pb(II) ions adsorption on HCPMs and MHCMs.

Table 2 Pseudo-first-order kinetic data and pseudo-second-order kinetic data for the adsorption of Pb(II) ions on HCPMs and MHCMs

Sample	Experimental values qe (mg g^−1)	Pseudo-first-order kinetic model			Pseudo-second-order kinetic model
		qe,cal (mg g^−1)	k1 (h^−1)	R^2	qe,cal (mg g^−1)	k2 (g mg^−1 h^−1)	R^2	hi (mg g^−1 h^−1)
HCPMs	264.9	335.2	0.03627	0.92884	284.0	2.0180 × 10^−4	0.99648	16.28
MHCMs	559.6	551.6	0.00958	0.96179	609.8	3.0932 × 10^−5	0.98839	11.50

---

The kinetic processes for the adsorption of Pb(II) ions on both the HCPMs and MHCMs followed the characteristics of the pseudo-first-order and pseudo-second-order kinetic models. Based on the pseudo-first-order kinetic model, the theoretical equilibrium adsorption amounts (q_e,cal) of the HCPMs and MHCMs were 335.2 and 551.6 mg g⁻¹, which were calculated according to the slopes and intercepts in the curves of log(q_e − q_t) versus t. Based on the pseudo-second-order kinetic model, the theoretical equilibrium adsorption amounts (q_e,cal) of the HCPMs and MHCMs were 284.0 and 609.8 mg g⁻¹, respectively. Since the adsorption of Pb(II) ions on the HCPMs and MHCMs exhibited higher compliance with the pseudo-second-order kinetic model than the pseudo-first-order kinetic model, the theoretical equilibrium adsorption amounts (q_e,cal) of the HCPMs and MHCMs should be 284.0 and 609.8 mg g⁻¹. Notably, the theoretical equilibrium adsorption amounts from the pseudo-second-order kinetic model were slightly higher than the experimental equilibrium adsorption amounts (Table 2).

According to the pseudo-second-order kinetic model, the initial adsorption rate of Pb(II) ions could be expressed as h_i = k₂q_e,cal² (g mg⁻¹ h⁻¹). Table 2 indicated that the initial adsorption rates (h_i) of Pb(II) ions on the HCPMs and MHCMs were 16.28 g mg⁻¹ h⁻¹ and 11.50 g mg⁻¹ h⁻¹, respectively, suggesting that the initial adsorption rate of Pb(II) ions on the HCPMs was slightly greater than that on MHCMs. Moreover, the adsorption rate constant of Pb(II) ions on the HCPMs (k₂ = 2.0180 × 10⁻⁴ g mg⁻¹ h⁻¹) was greater than that on the MHCMs (k₂ = 3.0932 × 10⁻⁵ g mg⁻¹ h⁻¹). The adsorption types of Pb(II) ions on the adsorbents included physical adsorption and chemical adsorption. Table 2 indicated that the MHCMs possessed greater Pb(II) adsorption amounts than the HCPMs because of their better chemical adsorption property. The lower initial adsorption rate of Pb(II) ions on the and MHCMs than on the HCPMs suggested that the physical adsorption played the greater role in adsorbing Pb(II) ions than the chemical adsorption at initial stage. Since the MHCMs had better chemical adsorption ability for Pb(II) ions than the HCPMs, the as-formed PbHAP rods blocked more easily the diffusion of PO₄³⁻ ions from the MHCMs than from the HCPMs. For the HCPMs, the adsorption rate of Pb(II) ions decreased to a constant, and then reached the equilibrium stage at about 168 h. In contrast, for the MHCMs, the adsorption rate of Pb(II) ions had a decline process from 24 h to 432 h and then reached the equilibrium stage at about 456 h. Therefore, the initial adsorption rate and adsorption rate constant of Pb(II) ions on the MHCMs were lower than those on the HCPMs.

4. Conclusion

We reported a new strategy to prepare the MHCMs according to the following steps: (i) the fabrication of CCPMs by freeze-drying technology; and (ii) the conversion of MHCMs from CCPMs by PBS treatment. After soaking in PBS, the CaCO₃ particles in the CCPMs were converted in situ to carbonated HAP plates with a preferred c-plane orientation via a dissolution–precipitation reaction. The MHCMs possessed 3D interconnected macropores with a pore size of 50–150 μm and mesopores with a pore size of 3.97 nm. The adsorption experimental of Pb(II) ions indicated that the MHCMs and HCPMs have great adsorption amounts of 559.6 and 264.4 mg g⁻¹, respectively. During the Pb(II) adsorption process, the HAP particles in both the MHCMs and HCPMs were transferred into PbHAP rods. The better adsorption property of the MHCMs than the HCPMs was attributed to great surface area, low Ca/P ratio, good degradability, hierarchically porous structures and carbonated HAP with a preferred c-plane orientation. The adsorption of Pb(II) ions on the MHCMs exhibits higher compliance with the pseudo-second-order kinetic model than the pseudo-first-order kinetic model.

Acknowledgements

This research was supported by Natural Science Foundation of China (No. 51372152), Program of Shanghai Normal University (No. DZL124, DCL201303, SK201333), Innovation Foundation of Shanghai Education Committee (No. 14ZZ124).

Notes and references

1. T. S. Anirudhan, C. D. Bringle and P. G. Radhakrishnan, Chem. Eng. J., 2012, 200, 149.
2. F. Googerdchian, A. Moheb and R. Emadi, Chem. Eng. J., 2012, 200, 471.
3. Y. Feng, J. L. Gong, G. M. Zeng, Q. Y. Niu, H. Y. Zhang, C. G. Niu, J. H. Deng and M. Yan, Chem. Eng. J., 2010, 162, 487.
4. G. Gyananath and D. K. Balhal, Cellul. Chem. Technol., 2012, 46, 121.
5. W. Wei, J. Cui and Z. Wei, Chemosphere, 2014, 105, 14.
6. Y. B. Yan, X. L. Dong, X. L. Sun, X. Y. Sun, J. S. Li, J. Y. Shen, W. Q. Han, X. D. Liu and L. J. Wang, J. Colloid Interface Sci., 2014, 429, 68.
7. L. M. Cui, Y. G. Wang, L. H. Hu, L. Gao, B. Du and Q. Wei, RSC Adv., 2015, 5, 9759.
8. D. G. Liu, Z. H. Li, W. Li, Z. R. Zhong, J. Q. Xu, J. J. Ren and Z. S. Ma, Ind. Eng. Chem. Res., 2013, 52, 11036.
9. C. X. Li, A. K. Born, T. Schweizer, M. Zenobi-Wong, M. Cerruti and R. Mezzenga, Adv. Mater., 2014, 26, 3207.
10. Y. P. Guo, T. Long, S. Tang, Y. J. Guo and Z. A. Zhu, J. Mater. Chem. B, 2014, 2, 2899.
11. X. Y. Zhao, Y. J. Zhu, J. Zhao, B. Q. Lu, F. Chen, C. Qi and J. Wu, J. Colloid Interface Sci., 2014, 416, 11.
12. D. X. Liao, W. Zheng, X. M. Li, Q. Yang, X. Yue, L. Guo and G. M. Zeng, J. Hazard. Mater., 2010, 177, 126.
13. B. Sandrine, N. Ange, B. A. Didier, C. Eric and S. Patrick, J. Hazard. Mater., 2007, 139, 443.
14. S. H. Jang, Y. G. Jeong, B. G. Min, W. S. Lyoo and S. C. Lee, J. Hazard. Mater., 2008, 159, 294.
15. S. H. Jang, B. G. Min, Y. G. Jeong, W. S. Lyoo and S. C. Lee, J. Hazard. Mater., 2008, 152, 1285.
16. F. Zhuang, R. Tan, W. Shen, X. Zhang, W. Xu and W. Song, J. Alloys Compd., 2015, 637, 531.
17. G. Fahimeh, M. Ahmad and E. Rahmatollah, Chem. Eng. J., 2012, 200–202, 471.
18. S. Saoiabi, S. El Asri, A. Laghzizil, A. Saoiabi, J. L. Ackerman and T. Coradin, Chem. Eng. J., 2012, 211–212, 233.
19. I. Mobasherpour, E. Salahi and M. Pazouki, Arabian J. Chem., 2012, 5, 439.
20. C. M. Simonescu, R. E. Patescu, L. T. Busuioc, C. Onose, A. Melinescu and V. Lilea, Rev. Chim., 2016, 67, 1498.
21. C. Piccirillo, S. I. A. Pereira, A. P. G. C. Marques, R. C. Pullar, D. M. Tobaldi, M. E. Pintado and P. M. L. Castro, J. Environ. Manage., 2013, 121, 87.
22. S. Shanmugamn and B. Gopal, Ceram. Int., 2014, 40, 15655–15662.
23. D. P. Minh, N. D. Tran, A. Nzihou and P. Sharrock, Chem. Eng. J., 2013, 232, 128.
24. K. Zargoosh, H. Abedini, A. Abdolmaleki and M. R. Molavian, Ind. Eng. Chem. Res., 2013, 52, 14944.
25. Y. B. Yan, X. L. Dong, X. L. Sun, X. Y. Sun, J. S. Li, J. Y. Shen, W. Q. Han, X. D. Liu and L. J. Wang, J. Colloid Interface Sci., 2014, 429, 68.
26. W. Liang, L. Zhan, L. H. Piao and C. Rüssel, Mater. Sci. Eng., B, 2011, 176, 1010.
27. C. S. Ciobanu, S. L. Iconaru, C. L. Popa, A. Costescu, M. Motelica-Heino and D. Predoi, J. Nanomater., 2014, 2014, 361061.
28. L. J. Dong, Z. L. Zhu, Y. L. Qiu and J. F. Zhao, Chem. Eng. J., 2010, 165, 827.
29. Z. Zhuang, T. J. Fujimi, M. Nakamura, T. Konishi, H. Yoshimura and M. Aizawa, Acta Biomater., 2013, 9, 6732.
30. Z. Zhuang and M. Aizawa, J. Mater. Sci.: Mater. Med., 2013, 24, 1211.
31. Z. Zhuang, H. Yoshimura and M. Aizawa, Mater. Sci. Eng., C, 2013, 33, 2534.
32. Y. C. Lu, J. He and G. Luo, Chem. Eng. J., 2013, 226, 271.
33. N. Li and R. Bai, Ind. Eng. Chem. Res., 2006, 45, 7897.
34. N. Gupta, A. K. Kushwaha and M. C. Chattopadhyaya, J. Taiwan Inst. Chem. Eng., 2012, 43, 125.
35. R. Bazargan-Lari, M. E. Bahrololoom and A. Nemati, J. Food, Agric. Environ., 2011, 9, 892.
36. Y. Zhang and M. Zhang, J. Biomed. Mater. Res., 2001, 55, 304.
37. K. S. W. Sing, Pure Appl. Chem., 1982, 54, 2201.
38. J. R. Rangel-Mendez, R. Monroy-Zepeda, E. Leyva-Ramos, P. E. Diaz-Flores and K. Shirai, J. Hazard. Mater., 2009, 162, 503.
39. Y. Zhang, Y. Liu, X. B. Ji, C. E. Banks and W. Zhang, Chem. Commun., 2011, 47, 4126.
40. M. R. Gandhi and S. Meenakshi, J. Hazard. Mater., 2012, 203–204, 29.
41. T. C. Coelho, R. Laus, A. S. Mangrich, V. T. deFávere and M. C. M. Laranjeira, React. Funct. Polym., 2007, 67, 468.
42. L. E. Abugoch, C. Tapia, M. C. Villamán, M. Yazdani-Pedram and M. Díaz-Dosque, Food Hydrocolloids, 2011, 25, 879.
43. W. Chen, T. Long, Y. J. Guo, Z. A. Zhu and Y. P. Guo, J. Mater. Chem. B, 2014, 2, 1653.
44. J. D. Chen, Y. J. Wang, K. Wei, S. H. Zhang and X. T. Shi, Self-organization of hydroxyapatite nanorods through oriented attachment, Biomaterials, 2007, 28, 2275.
45. D. Walsh, T. Furuzono and J. Tanaka, Biomaterials, 2001, 22, 1205.
46. H. Elfil and H. Roques, AICHE J., 2004, 50, 1908.
47. H. McDowell, T. M. Gregory and W. E. Brown, J. Res. Natl. Bur. Stand., Sect. A, 1977, 81, 273.
48. Y. Lei, J. J. Guan, W. Chen, Q. F. Ke, C. Q. Zhang and Y. P. Guo, RSC Adv., 2015, 5, 25462.
49. H. Y. Xu, L. Yang, P. Wang, Y. Liu and M. S. Peng, J. Environ. Manage., 2008, 86, 319.
50. H. B. Weiser and A. P. Bloxsom, J. Phys. Chem., 1924, 28, 26.
51. H. B. Weiser and W. O. Milligan, J. Phys. Chem., 1932, 36, 1950.
52. R. Karthik and S. Meenakshi, Int. J. Biol. Macromol., 2015, 72, 235.
53. Y. Lei, W. Chen, B. Lu, Q. F. Ke and Y. P. Guo, RSC Adv., 2015, 5, 98783.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra20448a

‡ X. Wen and C. T. Shao contributed equally to this work.

---