Biopolymer-Based Nanohydroxyapatite Composites for the Removal of Fluoride, Lead, Cadmium, and Arsenic from Water.

In this study, hydroxyapatite (HAP) nanocomposites were prepared with chitosan (HAP-CTS), carboxymethyl cellulose (HAP-CMC), alginate (HAP-ALG), and gelatin (HAP-GEL) using a simple wet chemical in situ precipitation method. The synthesized materials were characterized using scanning electron microscopy, Fourier transform infrared spectroscopy, X-ray diffraction, Brunauer-Emmett-Teller surface area analysis, and thermogravimetric analysis. This revealed the successful synthesis of composites with varied morphologies. The adsorption abilities of the materials toward Pb(II), Cd(II), F-, and As(V) were explored, and HAP-CTS was found to have versatile adsorption properties for all of the ions, across a wide range of concentrations and pH values, and in the presence of common ions found in groundwater. Additionally, X-ray photoelectron spectroscopy and energy-dispersive X-ray spectroscopy confirmed the affinity of HAP-CTS toward multi-ion mixture containing all four ions. HAP-CTS was hence engineered into a more user-friendly form, which can be used to form filters through its combination with cotton and granular activated carbon. A gravity filtration study indicates that the powder form of HAP-CTS is the best sorbent, with the highest breakthrough capacity of 3000, 3000, 2600, and 2000 mL/g for Pb(II), Cd(II), As(V), and F-, respectively. Hence, we propose that HAP-CTS could be a versatile sorbent material for use in water purification.

Introduction

class="Chemical">Water coclass="Chemical">ntamiclass="Chemical">natioclass="Chemical">n is ofteclass="Chemical">n caused by iclass="Chemical">norgaclass="Chemical">nic species such class="Chemical">n class="Chemical">as heavy metal ions and fluoride (F–). Due to their high stability in the environment, such contaminants are considered to be particularly hazardous and have been identified at unsafe levels in many parts of the world.[1,2] The ingestion of these ions can cause both acute and chronic diseases,[3−5] such as chronic kidney disease with unknown etiology (CKDu). The latter is particularly prevalent in Sri Lanka, and, while its cause remains unknown, the incidence is correlated with the intake of water contaminated with F– or metal ions such as Pb(II), Cd(II), and As(V).[3,6−9] The development of cost-effective methods to remove these contaminants from water is hence vital.

Effective sorbents should possess high affinity for target pollutants, ideclass="Chemical">ally beiclass="Chemical">ng able to adsorb severclass="Chemical">n class="Chemical">al contaminants simultaneously and to remove them rapidly. A large number of materials have been developed to remove both organic and inorganic contaminants from water.[10−12] However, some can act as sources of secondary pollution.[11,12] This can be prevented by using nontoxic and biodegradable adsorbents.[13] Nano-hydroxyapatite (HAP) and its composites have been shown to be effective adsorbents for water pollutants. HAP has low toxicity, biocompatibility,[14] and versatile sorption properties toward both anions and cations. However, using nanoparticles alone as a sorbent brings challenges, and they need to be mounted into a carrier medium to generate practical filter materials. Hence, this work is focused on combining HAP with biopolymers to develop more versatile and stronger filter materials.[15−19]

In addition to being nontoxic, naturclass="Chemical">al bioclass="Chemical">n class="Chemical">polymers such as chitosan,[20−23] cellulose,[24] carboxymethyl cellulose,[25,26] sodium alginate,[27] lignin,[28] pectin,[29] and gelatin[30,31] have chelating properties. Several studies have explored the sorption properties of HAP biopolymer composites for contaminants including Pb(II), Cd(II), As(V), and F–.[15−17,29,32−43] The properties of these composites vary depending on the synthesis method and the conditions applied in the adsorption studies. Only a few studies have explored competitive adsorption properties in multi-ion systems. In addition, there is minimal knowledge on how the type of polymer incorporated and the component ratios affect the sorption properties of HAP biopolymer composites. In this study, we sought to redress this lack of understanding. Four biopolymers, chitosan (CTS), carboxymethyl cellulose (CMC), sodium alginate (ALG), and gelatin (GEL), were explored. HAP nanocomposites with these biopolymers were prepared, and their sorption properties toward Pb(II), Cd(II), As(V), and F– were investigated.

Results

Initiclass="Chemical">ally, class="Chemical">naclass="Chemical">nocomposites with differeclass="Chemical">nt class="Chemical">n class="Chemical">HAP to polymer ratios were synthesized, and the optimal ratio for each composite was selected based on an initial comparative adsorption study. This is discussed in detail in Section S.1, Supporting Information (Figures S1–S4). HAP-CTS 60%, HAP-CMC 50%, HAP-ALG 50%, and HAP-GEL 50% were identified as the systems with the most potent adsorption properties toward Pb(II), Cd(II), F–, and As(V). Hereafter, these optimized composites will be referred to as HAP-CTS, HAP-CMC, HAP-ALG, and HAP-GEL.

Characterization of Materials

Scanning electron micrographs (SEM) were used to investigate the morphology of the nanocomposites, class="Chemical">as showclass="Chemical">n iclass="Chemical">n Figure . class="Chemical">n class="Chemical">HAP-CTS (Figure a) can be seen to have a highly porous structure with interconnected rod-shaped particles on the 20–50 nm scale. The SEM of HAP-CMC (Figure b) also shows a porous structure and a grainlike morphology with highly agglomerated spherical-shaped nanoparticles. However, the morphology and porosity of HAP-ALG are very different (Figure c). The structure is less porous, with embedded nanoparticles in the polymer matrix. HAP-GEL (Figure d) also has fewer pores and apparently larger particle sizes.

Figure 1

SEM images of the optimized HAP-biopolymer systems: (a) HAP-CTS, (b) HAP-CMC, (c) HAP-ALG, and (d) HAP-GEL.

SEM images of the optimized class="Chemical">HAP-bioclass="Chemical">n class="Chemical">polymer systems: (a) HAP-CTS, (b) HAP-CMC, (c) HAP-ALG, and (d) HAP-GEL.

Brunauer–Emmett–Teller (BET) anclass="Chemical">alysis wclass="Chemical">n class="Chemical">as performed to measure the specific surface area (SSA), pore size, and pore volume (see Figure S5). SSA values for HAP-CTS, HAP-CMC, HAP-ALG, and HAP-GEL were 191.0, 90.6, 83.8, and 80.0 m2/g, respectively. The pore volumes were, respectively, calculated as 4.02 × 10–1, 1.9 × 10–1, 1.2 × 10–1, and 3.3 × 10–1 cm3/g. When the four systems are compared, HAP-CTS shows the highest SSA and pore volume, which is consistent with the SEM data in Figure .

Fourier transform infrared (FTIR) spectra of class="Chemical">HAP-class="Chemical">n class="Chemical">CTS, HAP-CMC, HAP-ALG, and HAP-GEL are presented in Figure . All of the composites exhibit characteristic bands corresponding to HAP. For example, OH can be identified as a broad band around 3500 cm–1. Vibrations present at 1030, 905, and 565 cm–1 in all composites are attributed to the stretching vibration modes of PO43–.[10,44] All four HAP-biopolymer composites display a broadening of the adsorption band around 3300–3500 cm–1 toward the low-frequency side and the absence of the characteristic HAP OH stretch at 3570 cm–1.[45,46] In Figure a, characteristic bands for chitosan are found at around 2900 cm–1 in HAP-CTS and arise due to CH2 stretching vibrations. The bands present at 1540 and 1410 cm–1 can be attributed to the symmetric and asymmetric stretching of COO groups. In addition, a broad band from 3400 to 3100 cm–1 arises from stretching vibrations of N–H functional groups of chitosan.[33,39,47] Bands can also be identified at 1150 and 1380 cm–1, corresponding to C–O stretching and CH bending of chitosan.[48]

Figure 2

FTIR spectra of (a) HAP-CTS, (b) HAP-CMC, (c) HAP-ALG, and (d) HAP-GEL.

FTIR spectra of (a) class="Chemical">HAP-class="Chemical">n class="Chemical">CTS, (b) HAP-CMC, (c) HAP-ALG, and (d) HAP-GEL.

The FTIR spectrum of class="Chemical">HAP-class="Chemical">n class="Chemical">CMC (Figure b) shows a broad band around 1440–1360 cm–1 due to asymmetric carboxylate stretching and OH bending modes of CMC. These usually occurred at 1410 and 1360 cm–1,[49,50] but in the composite overlap with the carbonate peak of HAP at 1440 cm–1. C–O vibrations of CMC can also be observed in HAP-CMC at 925 cm–1, with a blue shift.[51,52]

The FTIR spectrum of class="Chemical">HAP-class="Chemical">n class="Chemical">ALG is given in Figure c. In ALG, there are bands at 950 and 1060 cm–1 due to stretching vibrations of the carboxylic groups, while the band at 850 cm–1 is from OH bending. In the HAP-ALG composite, these bands appear as one broad band due to overlapping with HAP bands at the same positions. The band at 1430 cm–1 for alginate arises due to CH2 bending,[53,54] and can also be seen with HAP-ALG. In addition, the presence of all of the characteristic HAP bands again confirms the formation of a binary complex.[55]

The FTIR spectrum of neat gelatin (Figure d) shows distinct bands at 1320 and 1550 cm–1 from class="Chemical">amide groups.[56,57] These vibratioclass="Chemical">ns caclass="Chemical">n class="Chemical">n class="Chemical">also be identified in HAP-GEL, confirming the successful incorporation of gelatin. Again, the HAP bands are also visible, confirming the successful synthesis of the composite material.[58,59]

X-ray diffraction (XRD) patterns of the four class="Chemical">HAP-bioclass="Chemical">n class="Chemical">polymer systems, neat HAP, and the raw biopolymers are overlaid in Figure . This reveals that neat HAP has characteristic Bragg reflections at 26, 29, 32–34, 40, 46–54, and 63°, which matches with the literature.[44,60] HAP-GEL contains more crystalline HAP[61,62] than the other systems, which is consistent with the SEM images (Figure ). Incorporation of the other polysaccharides led to broad peaks and reduced HAP crystallinity.[42,63] However, all four HAP-biopolymer nanocomposites show a peak corresponding to neat HAP between 32 and 34°, which is consistent with the successful incorporation of HAP with the biopolymers.

Figure 3

XRD patterns of the optimized HAP-biopolymer systems and raw materials.

XRD patterns of the optimized class="Chemical">HAP-bioclass="Chemical">n class="Chemical">polymer systems and raw materials.

Thermogravimetric studies (TGA) were conducted with class="Chemical">all four bioclass="Chemical">n class="Chemical">polymer nanocomposites (Figure S6). All were found to contain a small amount of water (5–10%) and were stable up to 200 °C. This augers well for their use in water purification applications.

Adsorption Studies

Effect of Time

The adsorption capacity class="Chemical">as a fuclass="Chemical">nctioclass="Chemical">n of time wclass="Chemical">n class="Chemical">as studied separately for Pb(II), Cd(II), As(V), and F–, using an initial concentration of 1500, 400, 1, and 40 ppm. This resulted in a solution pH of 6.3, 6,2, 6.9, and 6.7, respectively. The uptake data are presented in Figure . All of the nanocomposites showed initial fast adsorption due to the availability of a large number of active binding sites and reached a steady state in less than 30 min. For Pb(II) (Figure a), HAP-GEL shows a maximum of 85% absorption, which is reached within 5 min. HAP-CTS, HAP-CMC, and HAP-ALG reach maximum absorptions of 79, 76, and 74%, respectively, attained within 5, 20, and 10 min. A similar trend was observed for Cd(II) (Figure b), with 74% absorption on HAP-GEL within 10 min while HAP-CTS, HAP-CMC, and HAP-ALG reached 69, 66, and 65% in this time.

Figure 4

Adsorption vs time plots for (a) Pb(II), (b) Cd(II), (c) As(V), and (d) F– on HAP-biopolymer nanocomposites, at a solution pH of 6.3, 6,2, 6.9, and 6.7, respectively.

Adsorption vs time plots for (a) class="Chemical">Pb(II), (b) class="Chemical">n class="Chemical">Cd(II), (c) As(V), and (d) F– on HAP-biopolymer nanocomposites, at a solution pH of 6.3, 6,2, 6.9, and 6.7, respectively.

Plots of F– and class="Chemical">As(V) adsorptioclass="Chemical">n vs time are showclass="Chemical">n iclass="Chemical">n Figure c,d. class="Chemical">n class="Chemical">HAP-CTS is the most effective adsorbent, reaching the maximum adsorption with the shortest contact time. For F–, HAP-CTS shows 85% absorption within 10 min, while HAP-GEL and HAP-ALG give 78 and 74%, respectively, within 15 and 20 min. Considering As(V) adsorption, again, HAP-CTS shows the best properties, yielding 91% adsorption within 6 min, while HAP-GEL and HAP-ALG gave 87 and 45% adsorption within 10 and 20 min, respectively. HAP-CMC shows poor absorption properties for both F– and As(V), performing less well than neat HAP for As(V). The four HAP biopolymer composites clearly have both cationic and anionic adsorption sites, owing to the amalgamation of the biopolymer with HAP. HAP-GEL was the best adsorbent for Pb(II) and Cd(II), while HAP-CTS showed the best sorption properties toward all ions examined and identified as the most versatile adsorbent. This can be attributed to the presence of NH2 in CTS as well as its cationic nature compared to the other three polymers; in addition, this can be attributed to the higher surface area of HAP-CTS that resulted in 190 m2/g.

Effect of pH

The pH of a solution is one of the main factors that influence the process of adsorption since it affeclass="Chemical">cts the degree of ioclass="Chemical">nizatioclass="Chemical">n aclass="Chemical">nd the surface charge of the particles. Neat class="Chemical">n class="Chemical">HAP is not stable at pH values below 3 and tends to dissolve. However, all four HAP biopolymer nanocomposites were stable even at pH 3, and therefore, adsorption studies were conducted between pH 3 and 11 for Pb(II), Cd(II), fluoride, and As(V). The incubation time was 1 h. The effect of pH on adsorption of Pb(II) and Cd(II) is depicted in Figures S7 and S8. The adsorption percentage gradually increases from pH 3 to 8, before reaching a plateau at pH 10 due to precipitation of Pb(II) and Cd(II) hydroxides. Figure S9 exhibits the effect of pH on the adsorption of F–. All four composites show a decrease in adsorption percentage as the pH increases from pH 3 to 7. Above pH 8, there is a drastic drop in adsorption of F–. This is due to greater competition with OH– ions in solution at higher pH values. As(V) adsorption as a function of pH is depicted in Figure S10. The extent of absorption was greatest at pH 3–6, after which a decline can be observed. Out of the four nanocomposites, HAP-CTS was identified as the most versatile because it shows the best adsorption properties toward all four contaminants over a wide range of pH values, including at neutral pH.

Adsorption Isotherms

Adsorption isotherm models can be used to characterize the uptake behavior of matericlass="Chemical">als. The results of coclass="Chemical">nceclass="Chemical">ntratioclass="Chemical">n-depeclass="Chemical">ndeclass="Chemical">nt batch adsorptioclass="Chemical">n studies were fitted with the commoclass="Chemical">nly used Laclass="Chemical">ngmuir aclass="Chemical">nd Freuclass="Chemical">ndlich isotherm models. The liclass="Chemical">nearized form of the Laclass="Chemical">ngmuir adsorptioclass="Chemical">n isotherm, which is used to describe moclass="Chemical">nolayer adsorptioclass="Chemical">n, is giveclass="Chemical">n bywhere Ce, class="Chemical">n class="Chemical">Qe, KL, and QL are the concentration of the adsorbate at equilibrium (mg/L), the adsorption capacity (mg/g), the Langmuir isotherm constant (L/mg), and the maximum monolayer adsorption capacity (mg/g), respectively.

Equation gives the linear form of the Freundlich adsorption isotherm, where class="Chemical">Qe aclass="Chemical">nd Ce are the adsorptioclass="Chemical">n capacity (class="Chemical">n class="Chemical">mg/g) and the concentration (mg/L) at equilibrium, while Kf and n represent constants.The isotherm constants were calculated by fitting both models to the data for Pb(II), Cd(II), F–, and As(V) separately. The plots are presented in Figures S11–S14. The fit parameters obtained are listed in Table S1. The correlation coefficient (R2) and the square sum of error (SSE) (the square of the difference between the experimental adsorption capacity and calculated adsorption capacity divided by the corresponding calculated adsorption capacity) were used in combination with a visual inspection of the plots to identify the best-fit isotherm model.

The uptake data can be best fitted with the Freundlich isotherm in the majority of cclass="Chemical">ases, while class="Chemical">n class="Chemical">HAP-ALG absorption can be fitted well with both models. HAP-CTS, HAP-ALG, and HAP-CMC thus appear to be multilayer systems with surface heterogeneity, as would be expected from the SEM images in Figure . HAP-GEL can also be described by the Langmuir model, suggesting a monolayer system; this can perhaps be attributed to the presence of different functional groups and binding sites at the surface.

Kinetic Studies

Three commonly used kinetic models, the pseudo-first-order, pseudo-second-order, and intraparticle diffusion models, were used to fit the data. The results are presented in the Supporting Information, Figures S15–S18 and Table S2. n class="Chemical">All of the systems were fouclass="Chemical">nd to follow pseudo-secoclass="Chemical">nd-order kiclass="Chemical">netics.

The results obtained from the adsorption isotherm and kinetic studies class="Chemical">also iclass="Chemical">ndicated that class="Chemical">n class="Chemical">HAP-CTS is the nanocomposite with the most versatile sorption properties toward all four contaminants considered in this work. Therefore, HAP-CTS was subjected to further studies, in which it was loaded onto carrier materials and explored for use in real-life applications. Gravity filtration studies were also conducted using a multi-ion mixture, and regeneration studies were performed.

Characterization of Postabsorption Materials

To learn more about the uptake mechanisms, the class="Chemical">HAP-class="Chemical">n class="Chemical">CTS composites were analyzed after exposure to the pollutant ions, both individually and in a mixed solution (HAP-CTS-Ad).

X-ray photoelectron spectroscopy (XPS) survey spectra for class="Chemical">HAP-class="Chemical">n class="Chemical">CTS and HAP-CTS treated with a multi-ion mixture of Pb(II), Cd(II), As(V), and F– (HAP-CTS-Ad) are overlaid in Figure . For HAP-CTS, C, N, O, P, and Ca can be observed in the spectrum, in accordance with the chemical composition of the HAP-CTS. The survey spectrum of the HAP-CTS-Ad additionally shows distinctive lines from Pb, Cd, As, and F, consistent with the uptake discussed above. This clearly indicates the adsorption of Pb(II), Cd(II), As(V), and F– onto the surface of HAP-CTS. Using XPS data, the stoichiometric ratios of Ca and P were determined before and after the adsorption of HAP-CTS via the survey XPS spectra. The decrease of the Ca/P ratio after the adsorption should be due to the ion exchange with Pb(II) and Cd(II), and this indicates the possibility of removing Pb(II) and Cd(II) by ion exchange. The analysis of C 1s, O 1s, Ca 2p, and P 2p core levels were also explored and is explained in Section S.5 of the Supporting Information and Figure S19.[64−68] Energy-dispersive X-ray spectroscopy analysis (EDEX) was used to identify the distribution of the elements present in both neat HAP-CTS and HAP-CTS-Ad, shown in Figure , and confirms uptake of the different ion species.

Figure 5

XPS survey spectra of (a) HAP-CTS and (b) HAP-CTS-Ad.

Figure 6

Element mapping of HAP-CTS-Ad and EDX spectra showing the elemental compositions of HAP-CTS and HAP-CTS-Ad.

XPS survey spectra of (a) class="Chemical">HAP-class="Chemical">n class="Chemical">CTS and (b) HAP-CTS-Ad.

Element mapping of class="Chemical">HAP-class="Chemical">n class="Chemical">CTS-Ad and EDX spectra showing the elemental compositions of HAP-CTS and HAP-CTS-Ad.

XRD patterns of the post-adsorption samples of class="Chemical">HAP-class="Chemical">n class="Chemical">CTS are given in Figure S20. Pb(II), Cd(II), and As(V) adsorption has appeared to enhance the crystallinity. When Pb(II) and Cd(II) are taken up by HAP, lead pyromorphite and cadmium pyromorphite can form by ion exchange for Ca(II).[69,70] However, no precipitation of Pb(II) and Cd(II) salts with PO43– has taken place since no Bragg reflections from these phases are visible. Similarly, upon As(V) uptake, no reflections from calcium arsenate can be identified,[71] and As(V) must have interacted with HAP-CTS through electrostatic forces. There is minimal difference between the patterns of HAP-CTS and HAP-CTS-F, and no increase in crystallinity is noted. This suggests that F– is not incorporated into HAP but rather chelates with chitosan or Ca2+ ions in the formulation. It also indicates that the formation of CaF2 by precipitation of Ca in HAP-CTS with F– ions has not taken place since there is no evidence for CaF2 in the XRD data, and thus, there is no large-scale precipitation of this phase.[72−74] When HAP-CTS is exposed to all of the ions, the resultant pattern is a composite of the individual results seen with each pollutant.

Suggested Mechanism of Adsorption of Multi-Ions onto HAP-CTS

The ability of class="Chemical">HAP-class="Chemical">n class="Chemical">CTS to adsorb both cations and anions is very clear, based on the observations made in the adsorption studies, elemental mapping, and XPS data. Negatively charged ions (F– and AsO43–) get adsorbed either through an exchange process with OH– or by adhering to Ca2+ locations as depicted in Figure . For Pb(II) and Cd(II), ion exchange is the major mechanism underpinning uptake. In addition, these cations can also be immobilized by surface interactions with the negatively charged groups of the HAP-CTS structure.

Figure 7

Proposed mechanism for the adsorption of Pb(II), Cd(II), As(V), and F– with HAP-CTS nanocomposite.

Proposed mechanism for the adsorption of class="Chemical">Pb(II), class="Chemical">n class="Chemical">Cd(II), As(V), and F– with HAP-CTS nanocomposite.

Regeneration

class="Chemical">HAP-class="Chemical">n class="Chemical">CTS-Ad was subjected to regeneration studies by washing with dilute HCl and dilute NaOH. The material was subjected to three adsorption cycles with a mixed solution of 10 ppm Pb(II) and Cd(II), 1 ppm As(V), and 2 ppm F–. In each cycle, the adsorption capacity of the regenerated sample was reduced (Figure S21), ultimately to about 20–40% of the original capacity. This is expected given that the proposed uptake mechanisms lead to permanent incorporation of the pollutants into the HAP-CTS, and implies that the materials are only suitable for single-use filters.

Effect of Other Ions

Pollutant ions in reclass="Chemical">al-life class="Chemical">n class="Chemical">water sources are likely to be present alongside other ions such as Ca2+ and Mg2+. The potential confounding effect of these co-ions was thus explored, and the results are given in Figure S22. The effect of Ca and Mg ions was investigated using 20 and 50 ppm concentrations, which is equal to hard and very hard water with a total hardness of 130 and 325 in CaCO3 equivalents, respectively. At 20 ppm, no effect can be observed on the adsorption of any ion. At 50 ppm concentrations of Ca and Mg, adsorption of F– and As(V) was slightly enhanced by 5 and 3% due to the precipitation of these ions with Ca and Mg at higher concentrations. Similarly, there is no reduction in the adsorption of Pb(II) and Cd(II) at 20 ppm Ca/Mg, but there is a ca. 10% reduction at 20 ppm. No reduction in uptake performance was observed in the presence of nitrate and nitrite ions.

HAP-CTS-Loaded Matrices

After anclass="Chemical">alyziclass="Chemical">ng the adsorptioclass="Chemical">n data, it wclass="Chemical">n class="Chemical">as found that HAP-CTS was the optimal polymer composite, showing the ability to remove both cations and anions from water. Therefore, this composite was integrated with other matrices to develop devices for water purification since powder forms are not generally favorable in real applications. For this purpose, we selected cotton gauze and granular activated carbon (GAC) as possible matrices. The resultant materials were named HAP-CTS-CG and HAP-CTS-GAC. Figure depicts the distribution of HAP-CTS in a gauze matrix at three different magnifications. All these images indicate a uniform distribution of HAP-CTS and good porosity. When HAP-CTS was coated on GAC, a uniform coverage was not obtained (Figure b).

Figure 8

Morphology of the modified forms of HAP-CTS: (a) HAP-CTS-CG and (b) HAP-CTS-GAC. The magnification of the images increases moving from panel (i) to (iii).

Morphology of the modified forms of class="Chemical">HAP-class="Chemical">n class="Chemical">CTS: (a) HAP-CTS-CG and (b) HAP-CTS-GAC. The magnification of the images increases moving from panel (i) to (iii).

Gravity Filtration Studies

Gravity filtration studies were conducted for class="Chemical">HAP-class="Chemical">n class="Chemical">CTS, HAP-CTS-GAC, and HAP-CTS-CG, using gravity columns with 1 cm diameter to investigate the composites’ use in real-life applications using a mixture of ions. The initial concentration of F– was maintained at 2 ppm, as the commonly prevailing concentration in the groundwater of Sri Lanka is around 1.5–2.0 ppm.[6,8] The initial concentrations of As(V), Pb(II), and Cd(II) were maintained at 50 ppb, the maximum levels reported in the CKDu prominent areas of Sri Lanka.[3,7,8,75] Breakthrough capacities were calculated by considering 0.5 ppm as the safe limit for F– and 5 ppb for As(V), Pb(II), and Cd(II), according to the WHO standards.[76] Experiments involved passing the aforementioned ion mixture through the gravity column at a rate of 10 mL/30 ± 5 s. The breakthrough point is defined as the eluted volume at which the WHO limits are breached. The breakthrough curves for Cd(II), Pb(II), AS(V), and F– are presented in Figure . HAP-CTS showed the highest breakthrough volume for all the contaminant ions. The calculated breakthrough volumes for HAP-CTS with Pb(II), Cd(II), As(V), and F– were 3000, 3000, 2600, and 2000 mL/g respectively. As far as the safe limit of F– is concerned, only HAP-CTS reached the required target.

Figure 9

Breakthrough curves for (a) Cd(II), (b) Pb(II), (c) As(V), and (d) F– determined from gravity filtration studies with HAP-CTS powder, HAP-CTS-GAC, and HAP-CTS-CG. Vb: breakthrough volume.

Breakthrough curves for (a) class="Chemical">Cd(II), (b) class="Chemical">n class="Chemical">Pb(II), (c) As(V), and (d) F– determined from gravity filtration studies with HAP-CTS powder, HAP-CTS-GAC, and HAP-CTS-CG. Vb: breakthrough volume.

The results of the adsorption studies and the gravity filtration studies obtained in this study were compared with the other class="Chemical">HAP-bclass="Chemical">n class="Chemical">ased nanocomposites in the literature by considering the initial concentration, pH, isotherm, and kinetic models (Table ).

Table 1

Comparison of the Data Collected in This Work with the Adsorption Data in the Literaturea

								filter bed column studied
adsorbate	F– adsorbent	concentration range or highest concentration used (ppm)	pH	equilibrium time (min)	isotherm model	kinetic model	adsorption capacity (mg/g)	initial concentration (ppm)	flow rate (mL/min)	diameter, thickness of column (mm)	breakthrough capacity	ref
F–	modified HAP with activated alumina	10–200	8	480	F	2	14.4	3	not given	11	400 L/g	(77)
	Al-HAP	200	7	180	L	2	98.8	5	10	2, 0.3	1568 L/m2	(78)
	HAP-MMT	30	6.5	30	F	2	16.7	1.5	10	10, 0.2	1600 mL/g	(79)
	HAP-alginate	10		30		2	3.87	not reported				(80)
	HAP-cellulose	10		360	L	2	4.2	not reported				(81)
	magnetic HAP-alginate	10		30	DR	2	4.05	not reported				(36)
	HAP-CTS	10		30			1.56	not reported				(32)
	multiwall CNT-HAP	3–50	7	150	F and L	2	30.22	not reported				(82)
	HAP-gelatine	8–14		40	L	2	4.157	not reported				(41)
	CNT-HAP			300			11.05	not reported				(41)
	mineral-substituted HAP	10	7	60	F	2	8.36	not reported				(83)
	HAP-pectin	10–30	7	30	L	2	3.17	not reported				(29)
	HAP-CTS	10–40	6.9	10	F	2	16.2	1.8	10	10, 0.2	2000 mL/g	this work
	HAP-CMC	10–40	6.9	30	F	2	8.7	not done				this work
	HAP-ALG	10–40	6.9	20	F	2	12.5	not done				this work
	HAP-ALG	10–40	6.9	15	F	2	13.8	not done				this work
Cd(II)	hydroxyapatite chitosan fibers by wet spinning	1000		240	L	2	72	not reported				(16)
	hydroxyapatite alginate	300–1500	5	∼360	L and F	2	361	not reported				(84)
	hydroxyapatite alginate and gelatine	300–1500	5	∼300	L and F	2	388	not reported				(84)
	hydroxyapatite chitosan	100	5.6	90	F	2	122	not reported				(38)
	HAP-CTS	10–400	6.2	10	F	2	114.1	1.8	10	10, 0.2	3000 mL/g	this work
	HAP-CMC	10–400	6.2	10	F	2	99.0	not done				this work
	HAP-ALG	10–400	6.2	10	F	2	102.5	not done				this work
	HAP-ALG	10–400	6.2	10	F	2	144.9	not done				this work
Pb(II)	HAP chitosan fibers by wet spinning	1000	6	120	L	2	162	not done				(16)
	HAP carboxymethyl cellulose by wet chemical method	2500–6000	above 5.5	3	L		625	not done				(42)
	hydroxyapatite chitosan by wet chemical method	2500–6000		0.5	L		909.1	not done				(42)
	HAP-activated carbon by wet chemical method	1000		240	F		9–14	not done				(66)
	HAP alginate nanocomposites	300–1500		∼360	L	2	550	not done				(84)
	HAP-alginate gelatine	300–1500		∼240			616	not done				(30)
	HAP-chitosan	1000	3.5	60	L	2	100	not done				(85)
	HAP-turmeric-activated carbon	1000	6	150	L		29.4	not done				(86)
	HAP-GAC	1000	6	135	F		39.6	not done				(86)
	HAP-CTS	112–1540	6.3	5	F	2	514.1	1.8	10	10, 0.2	3000 mL/g	this work
	HAP-CMC	112–1540	6.3	20	F	2	478.8	not done				this work
	HAP-ALG	112–1540	6.3	10	F	2	480.3	not done				this work
	HAP-ALG	112–1540	6.3	5	F	2	579.8	not done				this work
As(V)	cellulose-carbonate HAP	1–50	4	60	L	1	12.7	not done				(34)
	HAP-CTS	0.6–16	6.7	6	L/F	2	3.38	1.8	10	10, 0.2	2600 mL/g	this work
	HAP-CMC	0.6–16	6.7	20	F	2	2.3	not done				this work
	HAP-ALG	0.6–16	6.7	15	F	2	2.1	not done				this work
	HAP-GEL	0.6–16	6.7	10	F	2	3.17	not done				this work

F: Freundlich, L: Langmuir, 1: first order, 2: second order.

Table summarizes the adsorption properties of previously reported class="Chemical">HAP-bclass="Chemical">n class="Chemical">ased nanocomposites toward Pb(II), Cd(II), F–, and As(V). There are significant variations in the adsorption properties depending on the reaction conditions. Many studies have been conducted to study F–, Pb(II), and Cd(II) adsorption, but studies for the adsorption of As(V) are few in number. The different experimental setups used make it hard to compare the different formulations. However, the equilibrium time obtained in this work is comparatively low, and HAP-CTS gives good uptake with a low contact time for Cd(II), F–, and As(V). In F– adsorption, the highest adsorption capacity of 98.8 mg/g is reported for Al-HAP using an initial concentration of 200 ppm.[78] This is higher than the values obtained in this work, in part likely due to the fact that the literature protocol uses a higher initial F– concentration than used here in addition to the high affinity of Al for fluorides. Comparison of the breakthrough capacities in gravity filtration studies is not possible since the literature does not report the densities of the materials used. However, what is clear is that our systems perform at least on par with those previously reported, and HAP-CTS acts as a versatile multi-ion absorbent.

Conclusions

Four class="Chemical">HAP bioclass="Chemical">n class="Chemical">polymer nanocomposites with chitosan, carboxymethyl cellulose, sodium alginate, and gelatin were successfully synthesized and used for the adsorption of Pb(II), Cd(II), As(V), and F– from water. The structure, morphology, and adsorption properties of these four nanocomposites were compared, and HAP-CTS was identified as the most versatile sorbent toward all four ions. Therefore, HAP-CTS was further modified for use in filters by combining it with gauze and GAC. Gravity filtration studies indicated that the powder form of HAP-CTS is the best sorbent, giving the highest breakthrough capacities of 3000, 3000, 2600, and 2000 mL/g for Pb(II), Cd(II), As(V), and F–, respectively. The main mechanism of cation removal from solution was found to be via ion exchange with Ca2+, while anions were removed through binding with Ca2+ sites and by exchange with replaceable OH– ions.

Experimental Section

Materials

class="Chemical">All chemicclass="Chemical">n class="Chemical">als were of analytical grade and used without further purification. Ammonium hydroxide [NH4OH] solution (25%, Sigma-Aldrich), calcium nitrate tetrahydrate [Ca(NO3)2·4H2O] (98%, Sigma-Aldrich), and di-ammonium hydrogen orthophosphate [(NH4)2HPO4] (98%, Sigma-Aldrich) were used to synthesize neat HAP. Chitosan (85%, deacetylated medium molecular weight, Sigma-Aldrich), carboxymethyl cellulose (low viscosity, Sigma-Aldrich), sodium alginate (low viscosity, Sigma-Aldrich), and gelatin type B (medium bloom, SRL) were used in the synthesis process of HAP-biopolymer nanocomposites. Cadmium nitrate (Sigma-Aldrich), lead nitrate (Merck), sodium fluoride (99.5%, Merck), and sodium arsenate (99%, Merck) were used to prepare the pollutant stock solutions.

Synthesis of HAP Biopolymer Nanocomposites

Synthesis of class="Chemical">HAP wclass="Chemical">n class="Chemical">as carried out according to the literature.[66] HAP-based nanocomposites were synthesized by in situ precipitation. (NH4)2HPO4 was mixed with a polymer solution (2%, w/v) at predetermined ratios, and vigorously stirred until a homogeneous solution was formed. Ca(NO3)2·4H2O was added dropwise into the solution to give a Ca/P ratio of 1.67, at 60 °C with vigorous stirring. During this process, the pH was kept at 10 with dropwise addition of NH4OH (5 M). The mixture was vigorously stirred for ∼3 h at room temperature and aged for 24 h at room temperature. The resultant precipitate was washed with distilled water until its pH became neutral and the product was separated by centrifugation. The resultant solid was oven-dried at 40 °C until a constant weight was obtained.

Incorporation of HAP-CTS into Filtration Matrix

class="Chemical">HAP-class="Chemical">n class="Chemical">CTS was amalgamated with granular activated carbon (GAC) and cotton gauze (CG) at a weight ratio of 1:1. Then, the materials were oven-dried at 60 °C.

Characterization

The surface morphology of the four class="Chemical">HAP-bioclass="Chemical">n class="Chemical">polymer nanocomposites was studied using a Hitachi SU6600 field emission scanning electron microscope (SEM). Fourier transform infrared (FTIR) spectra were collected on an AVATAR-320 instrument (Thermo Nicolet) in the wavenumber range between 500 and 4000 cm–1. X-ray diffraction analysis was conducted using a Rigaku SmartLab X-ray powder diffractometer using Cu Kα radiation (λ = 0.154 nm) over a 2θ range of 2–80°, with a step size of 0.02°. Thermogravimetric analysis was carried out using an SDT Q 600 analyzer in the temperature range from room temperature to 800 °C under air. Brunauer–Emmett–Teller (BET) analysis was performed using an automated gas sorption analyzer (Autosorb iQ-MP (1 stat), Viton).

Adsorption Studies

Elemental Concentrations

The concentrations of class="Chemical">Pb(II), class="Chemical">n class="Chemical">Cd(II), and As(V) before and after adsorption were analyzed using a GBC 932 AB atomic adsorption spectrometer (AAS) with necessary dilutions. Low ion concentrations (ppb level) were analyzed using an Agilent 4219 Microwave Plasma Atomic Emission Spectrophotometer (MP-AES). F– concentrations were analyzed using an F– ion-selective electrode by Hanna Instruments.

Effect of Time on Adsorption

The adsorption capacity for class="Chemical">Pb(II), class="Chemical">n class="Chemical">Cd(II), As(V), and F– was investigated over 60 min using initial concentrations of 1900, 400, 1, and 40 ppm, respectively. Experiments were carried out with 20 mL volume and 0.04 g of the absorbent at room temperature (27 ± 1 °C) with continuous stirring at a rate of 200 rpm using a horizontal shaker.

Effect of pH on Adsorption

To identify the adsorption properties at different pH levels, studies were carried out with class="Chemical">Pb(II), class="Chemical">n class="Chemical">Cd(II), F–, and As(V) separately over the pH range 3–11. Solutions were prepared as detailed in Section , except that the pH levels were adjusted using dilute HCl and dilute NaOH.

Isotherm Studies

Adsorption studies were performed in class="Chemical">polypropylene coclass="Chemical">ntaiclass="Chemical">ners at room temperature (27 ± 1 °C) with coclass="Chemical">nticlass="Chemical">nuous stirriclass="Chemical">ng at a rate of 200 rpm usiclass="Chemical">ng a horizoclass="Chemical">ntclass="Chemical">n class="Chemical">al shaker. The solutions were agitated until they reached the predetermined analysis time, when they were immediately filtered and the residual concentrations of the ions in the solution were analyzed with necessary dilutions.

Adsorption studies for class="Chemical">Pb(II) aclass="Chemical">nd class="Chemical">n class="Chemical">Cd(II) were carried out in the concentration ranges of 112–1540 and 10–400 ppm, respectively, at a fixed volume of 20 mL with 0.04 g of HAP composite. A stock solution of fluoride (1000 ppm) was prepared using NaF and deionized water. It was then diluted to the desired concentrations. Batch adsorption studies were carried out using 20 mL of the fluoride solution and 0.04 g of the absorbent, varying the fluoride concentration from 10 to 40 ppm. A stock solution of arsenate (50 ppm) was prepared using Na2HAsO4 and deionized water. Isotherm studies were carried out using 20 mL of aliquots at concentrations from 600 to 16 000 ppb with 0.04 g of the adsorbent.

Kinetic Studies

Kinetic studies of class="Chemical">Pb(II), class="Chemical">n class="Chemical">Cd(II), As(V), and F– adsorption were carried out at initial concentrations of 1900, 400, 1, and 40 ppm, respectively, using 0.04 g of HAP composite and 20 mL of the target solution for 2 h.[79]

Detailed Analysis of HAP-CTS

class="Chemical">HAP-class="Chemical">n class="Chemical">CTS (60%) was subjected to adsorption studies using multi-ion mixture containing 2 ppm F– and 50 ppb of Pb(II), Cd(II), and As(V). HAP-CTS samples were exposed to surface analysis using X-ray photoelectron spectroscopy (XPS) and energy-dispersive X-ray spectroscopy (EDEX) before and after the adsorption process.

For XPS, a Thermo Fisher Scientific UK instrument wclass="Chemical">as employed, with aclass="Chemical">n X-ray source of class="Chemical">n class="Chemical">Al Kα (1486.6 eV) attached to an ultra-high vacuum chamber. Survey XPS spectra with high pass energy (PE) and core-level spectra with low PE were performed in the constant analyzer energy mode with a pass energy of 50–200 eV, energy step of 0.1 eV, and scan numbers of 3 and 10, respectively. In this experiment, the base pressure in the ultra-high vacuum (UHV) chamber was lower than 2 × 10–8 Pa. The X-ray power was kept at 100 W to minimize radiation damage and the surface charge effect. The samples were neutralized using an electron flood gun. The ratio of intensities of photoelectron peaks of Ca 2p and P 2p peaks were calculated as explained in Section S.6 in the Supporting Information.[53,54] EDEX was carried out using a Z1 analyzer.

Gravity Filtration Studies

Gravity filtration studies were conducted for class="Chemical">HAP-class="Chemical">n class="Chemical">CTS. A mixture was prepared with 50 ppb of Pb(II), Cd(II), and As(V), and 2 ppm of F– was to mimic the groundwater in CKDu affected areas in Sri Lanka. The mixed ion solution was passed through a column with a diameter of 1 cm, across a filter bed prepared with 50 mg of the adsorbent deposited on 50 mg of a cotton bed, at a rate of 10 mL per 30 ± 5 s. The breakthrough capacity was calculated using the volume of the solution that could pass through before the F– concentration exceeded 0.5 ppm. For Pb(II), Cd(II), and As(V), 5 ppb was considered as the permissible level in accordance with WHO standards.