The objectives of this study are to optimize the synthesis conditions of metals such as aluminum (Al), iron (Fe) and cerium (Ce) impregnated activated carbon composites (AC-M) for fluoride removal and to evaluate the sorption behavior of fluoride by the composite under varying solution chemistry. To achieve the objectives, several composites were prepared with different combinations of Al, Ce and Fe at different temperatures. The fluoride removal by different composites was evaluated, and the best performing composite was selected for further study. Equilibrium sorption experiments and kinetic tests were carried out. The effect of pH and the presence of different co-ions on the removal of fluoride were assessed. Based on the performances, the composites can be ranked as AC-Ce > AC-AlCe > AC-CeFe > AC-AlCeFe > AC-AlFe. The maximum sorption capacity by the AC-Ce composites is in the range of 4.1–4.6 mg/g. The equilibrium sorption follows the Freundlich isotherm model whereas the kinetics is better explained by a pseudo-second-order kinetics model (0.018–0.029 g/mg/min). The presence of bicarbonate and phosphate has a significant effect on fluoride removal efficiency. The novel AC-Ce composites have a strong buffering effect under a wide range of pH, which can make it suitable for treating drinking water.

INTRODUCTION

Fluoride contamination in drinking water is a well-known problem. Application of nanoparticles for the treatment of water and wastewater has enormous potential (Miretzky & Cirelli 2011; Raychoudhury & Scheytt 2013), given that they have significantly high specific surface area (SSA). However, the NPs tend to aggregate which make them less efficient. On the other hand, porous granular activated carbon (AC) has high SSA and good mechanical resistance. Several studies have explored the possibility of impregnation of metal within the pores of granule material for removal of arsenic from drinking water (Chang et al. 2010; Nieto-Delgado & Rangel-Mendez 2012; Raychoudhury et al. 2015). However, only a few studies have explored the potential of impregnation of single metal within the porous material for removal of fluoride from water (Velazquez-Jimenez et al. 2015; Yu et al. 2015a; Kalidindi et al. 2017). Impregnation of layered double oxides/hydroxides of metal within the pores of AC (AC-M) is not explored widely, especially for removing fluoride from water. Furthermore, the efficiency of a composite is dependent on the synthesis conditions (Raychoudhury et al. 2015), which needs to be assessed for AC-M composites. Fluoride sorption by any adsorbent can act differently in natural conditions, where the variation in solution chemistry is significant. Thus, it is important to estimate the effect of solution pH and the presence of co-ions on the removal of fluoride.

The objectives of this study are, therefore, to optimize the synthesis conditions for AC-M (metal are Al, Fe, and Ce) composites for removal of fluoride and to evaluate the sorption behavior of fluoride by the best performing composite under varying solution chemistry. To identify the suitable combinations of metals, several composites were prepared with a different combination of Al, Ce, and Fe. All these composites were synthesized under a wide range of temperature. The fluoride removal by different composite was evaluated, and the best performing composite was selected for further study. Equilibrium sorption experiments and kinetic tests were carried out, and the data were compared with available models to evaluate the sorption behavior of fluoride by the selected composites. Furthermore, the effect of pH and the presence of different co-ions on fluoride removal were assessed.

MATERIALS AND METHODS

Material

AC of Darco 12–20 mesh (size: 850–1,000 μm), and sodium fluoride (NaF) were obtained from Sigma-Aldrich. Salts of different metals such as cerium nitrate [Ce(NO3)3, 6H2O], aluminum sulfate [Al2(SO4)3, 16 H2O], ferric nitrate [Fe(NO3)3, 9H2O], potassium sulfate (K2SO4), potassium phosphate (KH2PO4), sodium bicarbonate (NaHCO3) are reagent grade and were obtained from Hychem Pvt. Ltd (SDFCL). Reagent grade sodium hydroxide (NaOH) or hydrochloric acid (HCl) were used for pH adjustment. Millipore de-ionized (DI) water was used throughout the experiments.

Synthesis of the composite

Metal impregnated granular activated carbon (AC-M) composites were synthesized using a similar method to that adopted in the literature (Raychoudhury et al. 2015; Kalidindi et al. 2017). In this study the composite material was prepared under varying temperatures, ranging over an order of magnitude (60 °C–600 °C) and with a different combination of Al, Ce, and Fe metals. Solutions of Ce-salt or combinations of Al-Ce, Ce-Fe, Al-Fe and Al-Ce-Fe salts (detailed proportions are given in Table 1) were poured into a series of conical flasks, each containing AC. The volume of the added solution was 250 mL, where the concentration of total metal and AC were kept 0.03 mol/L and 40 g/L, respectively. All the samples, containing different proportions of metals, were heated to 60 °C, 120 °C, 240 °C, 360 °C and 600 °C temperature for 24 h, 24 h, 2 h, 1 h, and 1 h, respectively. Details of the composites and their synthesis conditions are summarized in Table 1.

Table 1

Synthesis conditions and fluoride removal efficiency by different AC-M composites

Sr No.Different compositesSynthesis temperature (oC)Concentration of Al (mol)Concentration of Ce (mol)Concentration of Fe (mol)Fluoride sorption (mg/g)
AC – – –  0.18 ± 0.04 
AC-Ce-1 60 – 0.030 – 2.30 ± 0.05 
AC-AlCe-1 60 0.015 0.015 – 1.26 ± 0.04 
AC-AlFe-1 60 0.015 – 0.015 1.43 ± 0.13 
AC-CeFe-1 60 – 0.015 0.015 1.51 ± 0.15 
AC-AlCeFe-1 60 0.010 0.010 0.010 0.92 ± 0.23 
AC-Ce-2 120 – 0.030 – 2.44 ± 0.20 
AC-AlCe-2 120 0.015 0.015 – 1.45 ± 0.32 
AC-AlFe-2 120 0.015 – 0.015 0.54 ± 0.09 
10 AC-CeFe-2 120 – 0.015 0.015 1.00 ± 0.16 
11 AC-AlCeFe-2 120 0.010 0.010 0.010 1.16 ± 0.05 
12 AC-Ce-3 240 – 0.030 – 2.46 ± 0.22 
13 AC-AlCe-3 240 0.015 0.015 – 1.50 ± 0.30 
14 AC-AlFe-3 240 0.015 – 0.015 0.71 ± 0.05 
15 AC-CeFe-3 240 – 0.015 0.015 1.25 ± 0.16 
16 AC-AlCeFe-3 240 0.010 0.010 0.010 0.98 ± 0.01 
17 AC-Ce-4 360 – 0.030 – 2.51 ± 0.13 
18 AC-AlCe-4 360 0.015 0.015 – 1.42 ± 0.41 
19 AC-AlFe-4 360 0.015 – 0.015 0.67 ± 0.04 
20 AC-CeFe-4 360 – 0.015 0.015 1.33 ± 0.08 
21 AC-AlCeFe-4 360 0.010 0.010 0.010 0.99 ± 0.05 
22 AC-Ce-5 600 – 0.030 – 0.29 ± 0.02 
23 AC-AlCe-5 600 0.015 0.015 – 0.48 ± 0.02 
24 AC-AlFe-5 600 0.015 – 0.015 0.21 ± 0.06 
25 AC-CeFe-5 600 – 0.015 0.015 0.30 ± 0.15 
26 AC-AlCeFe-5 600 0.010 0.010 0.010 0.36 ± 0.10 
Sr No.Different compositesSynthesis temperature (oC)Concentration of Al (mol)Concentration of Ce (mol)Concentration of Fe (mol)Fluoride sorption (mg/g)
AC – – –  0.18 ± 0.04 
AC-Ce-1 60 – 0.030 – 2.30 ± 0.05 
AC-AlCe-1 60 0.015 0.015 – 1.26 ± 0.04 
AC-AlFe-1 60 0.015 – 0.015 1.43 ± 0.13 
AC-CeFe-1 60 – 0.015 0.015 1.51 ± 0.15 
AC-AlCeFe-1 60 0.010 0.010 0.010 0.92 ± 0.23 
AC-Ce-2 120 – 0.030 – 2.44 ± 0.20 
AC-AlCe-2 120 0.015 0.015 – 1.45 ± 0.32 
AC-AlFe-2 120 0.015 – 0.015 0.54 ± 0.09 
10 AC-CeFe-2 120 – 0.015 0.015 1.00 ± 0.16 
11 AC-AlCeFe-2 120 0.010 0.010 0.010 1.16 ± 0.05 
12 AC-Ce-3 240 – 0.030 – 2.46 ± 0.22 
13 AC-AlCe-3 240 0.015 0.015 – 1.50 ± 0.30 
14 AC-AlFe-3 240 0.015 – 0.015 0.71 ± 0.05 
15 AC-CeFe-3 240 – 0.015 0.015 1.25 ± 0.16 
16 AC-AlCeFe-3 240 0.010 0.010 0.010 0.98 ± 0.01 
17 AC-Ce-4 360 – 0.030 – 2.51 ± 0.13 
18 AC-AlCe-4 360 0.015 0.015 – 1.42 ± 0.41 
19 AC-AlFe-4 360 0.015 – 0.015 0.67 ± 0.04 
20 AC-CeFe-4 360 – 0.015 0.015 1.33 ± 0.08 
21 AC-AlCeFe-4 360 0.010 0.010 0.010 0.99 ± 0.05 
22 AC-Ce-5 600 – 0.030 – 0.29 ± 0.02 
23 AC-AlCe-5 600 0.015 0.015 – 0.48 ± 0.02 
24 AC-AlFe-5 600 0.015 – 0.015 0.21 ± 0.06 
25 AC-CeFe-5 600 – 0.015 0.015 0.30 ± 0.15 
26 AC-AlCeFe-5 600 0.010 0.010 0.010 0.36 ± 0.10 

Fluoride removal efficiency

Fluoride removal efficiency by AC-M composites (such as AC-Ce, AC-AlCe, AC-AlFe, AC-CeFe, AC-AlCeFe) was assessed. The batch experiments were carried out by adding 50 mL of 10 mg/L fluoride solutions (prepared from NaF), into a series of polyethylene tubes, each containing 2 g/L AC-M composites. The pH of the fluoride solution was in the range of 6.0 ± 0.2. The tubes were kept in a rotator at 50 rpm for 3 h at a temperature of 25 °C and then the samples were analyzed.

The AC-Ce composites prepared by impregnating only 0.03 mol/L of Ce and treated at a temperature ranging from 60 °C to 360 °C [denoted as AC-Ce-1 to AC-Ce-4] showed the most promising performance. Thus, these AC-Ce composites were selected as reference adsorbent for further study. Cerium concentration was measured using inductively coupled plasma atomic emission spectroscopy (ICP-AES) (SAIF, Cochin University) at the end of the sorption test to examine the possibility of leaching of cerium from the composite into the treated water. Furthermore, sorption–desorption for all AC-Ce composites was carried out to understand the nature of fluoride sorption.

Sorption kinetics

In these batch experiments, 50 mL of 10 mg/L of fluoride solution were placed in a series of tubes containing 2 g/L of selected AC-Ce composites (pH = 6.0 ± 0.2). The samples were withdrawn at different time intervals (i.e. 5, 10, 20, 30, 45, 60, 90, 120 and 180 minutes) from individual bottles for measuring fluoride concentration.

Sorption isotherm

For the equilibrium sorption experiment, the concentrations of fluoride in the solution were varied over an order of magnitude ranging from 5 mg/L to 58 mg/L, whereas the adsorbent dose was kept fixed at 2 g/L. The sampler tubes were shaken thoroughly with a rotator at room temperature for 3 h, given that the duration is sufficient to reach equilibrium concentration.

Effect of solution chemistry

The effect of the presence of co-ions such as the presence of bicarbonate (HCO3), sulfate (SO4), and phosphate (HPO4) on the removal of fluoride, was assessed individually within a concentration range of 0 mmol/L to 50 mmol/L. To evaluate the effect of pH on the removal of fluoride, 10 mg/L fluoride solutions were prepared with varying pH value ranging from 4.0 to 10. The pH was adjusted by adding 0.1 mol/L HCl or 0.1 mol/L NaOH. It was also ensured that the variation of ion concentration and pH did not affect the sensitivity of the electrodes.

Regeneration of the composites

The composite was filtered out and rinsed with DI water after sorption experiment. Then 50 mL of 0.5 mol/L NaOH solution was added to the fluoride containing AC-Ce composite and was mixed for 3 h to accelerate fluoride desorption (Biswas et al. 2010). After desorption, the composite was washed with 0.1 mol/L HCl followed by DI water. The washed composites were dried overnight (15 h), and the fluoride removal efficiency by the regenerated composites was assessed.

For sample analysis, fluoride concentration was measured following method EPA-340.2, where total ionic strength adjustable buffer solution was mixed with the samples (1:1 ratio) to maintain the desirable pH range. The fluoride concentration of samples was measured using a fluoride-Ion Selective Electrode (FL700, Extech).

RESULT AND DISCUSSION

Fluoride removal efficiency under different synthesis conditions

In total 25 AC-M composites were prepared with varying combinations of metals and under different temperatures. Fluoride removal efficiency by AC-M composites (presented in Figure 1(a) and Table 1) are reasonably high compared to that by unmodified-AC (0.18 ± 0.04 mg/g). Detailed characterization of the AC-M composites was carried out in a previous study (Kalidindi et al. 2017). The scanning electron microscopy images of the Ac-Ce composite are also presented in the Supplementary material (Figure S1, available with the online version of this paper).
Figure 1

(a) Fluoride removal efficiency by different AC-M composites. (b) Sorption of fluoride with respect to time and with pseudo-first-order and pseudo-second-order kinetics model fit. Initial fluoride and the composite concentrations were 10 mg/L and 2 g/L, respectively. (c) Sorption isotherm of fluoride by different AC-Ce composites. The initial fluoride concentrations varied from 5 to 58 mg/L, and the composite dose was 2 g/L.

Figure 1

(a) Fluoride removal efficiency by different AC-M composites. (b) Sorption of fluoride with respect to time and with pseudo-first-order and pseudo-second-order kinetics model fit. Initial fluoride and the composite concentrations were 10 mg/L and 2 g/L, respectively. (c) Sorption isotherm of fluoride by different AC-Ce composites. The initial fluoride concentrations varied from 5 to 58 mg/L, and the composite dose was 2 g/L.

Combination of different metals

The composites containing 0.03 mol/L of Ce (AC-Ce) and synthesized in the temperature range 60 °C–360 °C, show the most promising performance. Figure 1(a) indicates fluoride removal efficiency of the composites reduced when Al and/ Fe were also impregnated along with Ce. Different AC-M composites, synthesized within a temperature range of 120 °C to 360 °C are in proximity (either they have increased slightly or have not changed much with an increase in the temperature) and form a cluster. Identification of these clusters among the bi-metal impregnated AC-M composites indicates that the absence of cerium in AC-AlFe composites (0.54–0.67 mg/g) reduces their performance significantly. Overall, the performance of the composites can be ranked as follows AC-Ce > AC-AlCe > AC-CeFe > AC-AlCeFe > AC-AlFe.

Effect of synthesis temperature

Temperature has different effects on different types of AC-M composites. For AC-Ce, AC-AlCe, and AC-AlCeFe composites, where Al and Ce are present, fluoride removal efficiency increases up to a certain extent with the increase in synthesis temperature from 60 °C to 120 °C. In contrast, for the composites AC-AlFe and AC-CeFe, where iron is present, a sharp decrease in fluoride removal efficiency was observed with an increase in the temperature beyond 60 °C. At lower temperature, surface active iron-hydroxyl group is present, which is capable of better sorption of fluoride. However, with an increase in the temperature the proportion of iron oxides increases, which is probably responsible for a decrease in the fluoride removal efficiency (Chen et al. 2012). In general, a significant reduction in fluoride sorption capacity was observed for all the composites treated at 600 °C. Elimination of the hydroxyl surface active group by formation of irreversible metal oxides, and development of the crystalline structure of metal oxides at high temperature is probably responsible for the reduction in the fluoride removal efficiency (Wu et al. 2007; Chen et al. 2012; Kang et al. 2013).

AC-Ce composites synthesized at a temperature of 60 °C, 120 °C, 240 °C, and 360 °C denoted as AC-Ce-1, AC-Ce-2, AC-Ce-3, and AC-Ce-4, respectively show the most promising performance and thus are selected for further studies. Cerium concentration for these four composites was measured at the end of the sorption test. The concentration of Ce in the treated solution was below the detection limit (detection limit: 20 μg/L), which suggests that leaching of Ce in the treated water is insignificant.

Sorption kinetics

Sorption kinetics data were compared with pseudo-first-order, pseudo-second-order and the Webber–Morris model and presented in Figure 1(b). The model parameters are estimated by least square error minimization approach, and the parameters are specified in Table 2.

Table 2

Sorption kinetics and equilibrium isotherm results for the AC-Ce composites

 Pseudo-first order
Pseudo-second order
Webber–Morris
Samplesqe (mg/g)kt (1/min)rft2qe (mg/g)k't (g/mg/min)rst2k''t (mg/g/min1/2)rst2
AC-Ce-1 2.21 0.041 0.87 2.53 0.021 0.93 0.15 0.91 
AC-Ce-2 2.26 0.038 0.93 2.62 0.018 0.97 0.15 0.93 
AC-Ce-3 2.34 0.054 0.79 2.62 0.029 0.92 0.14 0.93 
AC-Ce-4 2.38 0.040 0.75 2.64 0.022 0.84 0.14 0.94 
 Pseudo-first order
Pseudo-second order
Webber–Morris
Samplesqe (mg/g)kt (1/min)rft2qe (mg/g)k't (g/mg/min)rst2k''t (mg/g/min1/2)rst2
AC-Ce-1 2.21 0.041 0.87 2.53 0.021 0.93 0.15 0.91 
AC-Ce-2 2.26 0.038 0.93 2.62 0.018 0.97 0.15 0.93 
AC-Ce-3 2.34 0.054 0.79 2.62 0.029 0.92 0.14 0.93 
AC-Ce-4 2.38 0.040 0.75 2.64 0.022 0.84 0.14 0.94 
 Freundlich model
Langmuir model
Samples1/nrf2qmax (mg/g)kl (L/mg)rl2
AC-Ce-1 1.49 0.28 0.91 4.37 0.27 0.70 
AC-Ce-2 1.58 0.26 0.89 4.10 0.44 0.68 
AC-Ce-3 1.53 0.27 0.87 4.14 0.37 0.75 
AC-Ce-4 1.35 0.32 0.90 4.62 0.25 0.71 
 Freundlich model
Langmuir model
Samples1/nrf2qmax (mg/g)kl (L/mg)rl2
AC-Ce-1 1.49 0.28 0.91 4.37 0.27 0.70 
AC-Ce-2 1.58 0.26 0.89 4.10 0.44 0.68 
AC-Ce-3 1.53 0.27 0.87 4.14 0.37 0.75 
AC-Ce-4 1.35 0.32 0.90 4.62 0.25 0.71 

qt: adsorbed mass at time t, qe1 and qe2 are the fitted value of adsorbed mass at equilibrium for pseudo-first-order and second-order kinetics models, respectively; kt, kt and kt’’ are pseudo-first, second-order and intra-particular diffusion rate constants, respectively; ce: equilibrium concentration; kf and 1/n: Freundlich sorption constants; qmax: maximum sorption capacity; kl: Langmuir sorption constant; rft2, rst2, rf2 and rl2, are the r2 values for pseudo-first-order, second-order kinetics, Freundlich and Langmuir isotherm models, respectively.

Fluoride sorption by all four AC-Ce composites (with a fluoride concentration of 10 mg/L and an adsorbent dose of 2 g/L) is better explained by a pseudo-second-order kinetics model (r2: 0.84–0.97) compared to pseudo-first-order kinetics model (r2: 0.75–0.93). Fluoride removal by metal–nonmetal granular composite such as aluminum–lanthanum–scoria (Zhang et al. 2014), alginate-Fe-Zr composite (Swain et al. 2013), MgO2 coated AC (Ma et al. 2009), etc. is well explained by a pseudo-second-order kinetics model (Supplementary material, Table S1, available with the online version of this paper). The sorption rate of fluoride and the equilibrium sorption capacity of the AC-Ce composites estimated from the second-order kinetics model are varying within a range of 0.018 to 0.029 g/mg/min and 2.53–2.64 mg/g, respectively (Table 2). Although the rate of sorption is in proximity to the values reported in other studies (0.022–0.088 g/mg/min), the equilibrium sorption capacities in those studies were in the lower range (Supplementary material, Table S1) (Ma et al. 2009; Swain et al. 2013; Zhang et al. 2014). It is important to note that sorption rate is dependent on adsorbent dose and equilibrium sorption. The adsorbent doses used in those studies are much higher (adsorption dose: 5–30 g/L) with less equilibrium sorption capacity (equilibrium sorption: 0.112–0.98 mg/g) compared to that of this study (adsorption dose: 2 g/L, equilibrium sorption: 2.53–2.64 mg/g). Thus, it could be stated that the sorption rate by the AC-Ce composite is promising with a high capacity of sorption and a smaller dose of adsorbent.

Fitting the data with the Webber–Morris model (r2: 0.91–0.94) indicates that (Supplementary material, Figure S2, available online) intra-particular diffusion and boundary layer diffusion are probably the rate limiting factors. Intra-particular diffusion rates for those AC-Ce composites are (0.14–0.15 mg/g/min1/2) reasonably high.

Sorption isotherm

The Langmuir and Freundlich isotherm models (Table 2) were fitted to the data using a least square error minimization approach. The experimental data along with model fit are presented in Figure 1(c). For all the four AC-Ce composites, the equilibrium sorption behavior is better explained by the Freundlich isotherm model (r2: 0.87–0.91) compared to the Langmuir isotherm model (r2: 0.68–0.75). The sorption behavior suggests that the sorption-active sites within the composite are heterogeneous in nature. Equilibrium sorption of fluoride by a few metal–non-metal granular composites such as alginate-Fe-Zr (Swain et al. 2013), MnO2 coated AC (Ma et al. 2009) are also explained well by the Freundlich isotherm model. The value of the constant, 1/n, estimated in this study is in the range of 0.26 to 0.32 (Table 2), which indicates higher sorption energy exists at lower fluoride concentration range. This phenomenon is extremely useful for treating drinking water from natural sources, where the fluoride content in water is not highly concentrated but of existing drinking water quality standard.

Estimation of Langmuir parameters suggests that the maximum sorption capacity of the composite is in the range of 4.1–4.6 mg/g, which is comparatively high (Supplementary material, Table S2, available online). Furthermore, a major proportion of fluoride sorption by the AC-Ce composites is likely to be governed by strong chemical interaction. This interpretation is supported by the fact that only 11 ± 1.78% of fluoride is desorbed from the AC-Ce composites (Supplementary material, Figure S3, available online). Moreover, Fourier transform infrared spectroscopy (FTIR) analysis (result not shown) confirms the occurrence of chemical bonding between Ce and F. The above facts imply that the composite is capable of removing fluoride efficiently, and that desorption of fluoride into the drinking water is unlikely.

Effect of solution chemistry

Effect of presence of co-ions on fluoride removal

The effect of concentration of different co-ions such as bicarbonate, sulfate, and phosphate on fluoride removal efficiency was evaluated. Figure 2 suggests different AC-Ce composites behave similarly in the response to presence of co-ions.
Figure 2

Effect of the presence of (a) bicarbonate, (b) sulfate and (c) phosphate ions on the removal of fluoride by different AC-Ce composites. (d) Effect of pH on fluoride removal by different AC-Ce composites. The initial concentration of fluoride and the composites were 10 mg/L and 2 g/L, respectively.

Figure 2

Effect of the presence of (a) bicarbonate, (b) sulfate and (c) phosphate ions on the removal of fluoride by different AC-Ce composites. (d) Effect of pH on fluoride removal by different AC-Ce composites. The initial concentration of fluoride and the composites were 10 mg/L and 2 g/L, respectively.

Fluoride removal efficiency decreases drastically from 2.30 mg/L to 0.72 mg/L in the presence of 1 mmol/L bicarbonate (Figure 2(a)). No fluoride removal was observed with the background NaHCO3 content of 50 mmol/L. The pH of the system increases from 6.15 to 9.72 with an increase in bicarbonate concentration from 0 to 50 mmol/L (Supplementary material, Figure S4, available online). Several studies have observed a significant effect of bicarbonate on the removal of fluoride by different metal-based adsorbents (He et al. 2014; Wang et al. 2015; Yu et al. 2015a). The release of OH ions due to hydrolysis in the presence of HCO3 can result in an increase in pH. The combined effect of increased pH and competition for the adsorption-active sites between OH and F ion can reduce the fluoride removal efficiency, which is attributed to the extensive reduction in fluoride removal efficiency in the presence of HCO3 ions.

It can be observed from Figure 2(b) that the presence of sulfate up to 10 mmol/L does not reduce the fluoride removal efficiency significantly (from 2.43 ± 0.09 mg/g to 2.03 ± 0.13 mg/g) for different composites. However, fluoride sorption capacity reduces approximately half (from 2.43 ± 0.09 mg/g to 1.06 ± 0.17 mg/g) with an increase in sulfate concentration from 0 to 50 mmol/L.

The presence of phosphate has a significant effect on fluoride removal efficiency (Figure 2(c)). In an average the fluoride removal capacity decreases from 2.43 ± 0.09 mg/g to 0.56 ± 0.06 mg/g with an increase in phosphate concentration from 0 mmol/L to 1 mmol/L. Fluoride removal was consistent with a phosphate concentration range of 1 mmol/L to 10 mmol/L. An increase in phosphate concentration up to 50 mmol/L resulted in further decrease in the fluoride removal (0.3 ± 0.03 mg/g). The initial pH of the solution decreased (from 6.15 to 4.77, Supplementary material, Figure S4) with an increase in phosphate concentration. A significant effect of phosphate concentration on the removal of fluoride is reported in many other studies (He et al. 2014; Wang et al. 2015; Yu et al. 2015a). Competition for sorption-active sites and formation of surface complexation are likely to be responsible for the reduction in fluoride removal efficiency with increase in phosphate concentration.

Effect of pH on fluoride removal

Fluoride removal from water under a pH range of 4 to 10 was evaluated and presented in Figure 2(d). At a pH value of 4, all the composites show the best performances. With an increase in pH from 4 to 8, fluoride removal efficiency decreases to some extent. Increase in pH value up to 10 resulted in a steep decrease in fluoride removal efficiency (Figure 2(d)). Several studies have made a similar observation (Ma et al. 2009; Chai et al. 2013; He et al. 2014). Decrease in fluoride removal at higher pH (>8) is probably attributed to the combined effect of both competition between hydroxyl (OH) and F ions for sorption-active sites (He et al. 2014) as well as a reduction in surface charge at high pH (Chai et al. 2013; Yu et al. 2015b).

The pH value of the solution was measured at the initial stage and the end of the experiment. The result (Figure 2(d)) indicates that although the initial pH value varies within a wide range of 4 to 10, the final value of pH is restricted within a narrow range of 5.7 to 6.3 in the presence of AC-Ce composites. This behavior indicates that all the AC-Ce composites has an extremely good buffering capacity, and thus it could be used efficiently for treating drinking water from different sources having a wide range of pH.

Regeneration

AC-Ce composites were regenerated by desorption of fluoride in the presence of NaOH and successively washing with 0.1 mol/L HCl solution and DI water, respectively. However, the performance of the regenerated composite was not very promising. The fluoride removal capacity has reduced almost five times (from 2.3–2.51 mg/g to 0.5 mg/g, Supplementary material, Figure S5, available online). Chemical interaction between fluoride and the oxides/hydroxides of cerium is probably inhibiting desorption of fluoride and consequently regeneration of the composite.

CONCLUSION

Overall, impregnation of cerium in granular AC increases the fluoride removal efficiency significantly. Synthesis temperature (ranging between 60 °C and 360 °C) has minimal effect on fluoride sorption capacity (4.1–4.6 mg/g) by the composite, which implies the composite treated at the lowest temperature is the most energy efficient. Desorption of fluoride and leachate of cerium from AC-Ce composites is minimum, which can make the composite reliable for using as a filter media. The equilibrium sorption follows the Freundlich isotherm model with high sorption capacity at low fluoride concentration. Furthermore, the rate of sorption of fluoride by AC-Ce composites is reasonably high (0.018–0.029 g/mg/min). Sorption behavior of the composite is suitable for treating drinking water, where the expected concentration of fluoride is not very high but conforms to the existing drinking water standard. The presence of sulfate does not have much effect on the fluoride removal. However, the presence of bicarbonate and phosphate can reduce the performance of the composite significantly, which needs to be addressed before using it as filter media.

REFERENCES

REFERENCES
Chai
L. Y.
Wang
Y. Y.
Zhao
N.
Yang
W. C.
You
X. Y.
2013
Sulfate-doped Fe3O4/Al2O3 nanoparticles as a novel adsorbent for fluoride removal from drinking water
.
Water Research
47
(
12
),
4040
4049
.
Chang
Q.
Lin
W.
Ying
W.-c.
2010
Preparation of iron-impregnated granular activated carbon for arsenic removal from drinking water
.
Journal of Hazardous Materials
184
(
1–3
),
515
522
.
Kalidindi
S.
Vecha
M.
Kar
A.
Raychoudhury
T.
2016
Aluminum-cerium double-metal impregnated activated carbon: a novel composite for fluoride removal from aqueous solution
.
Water Science and Technology: Water Supply
17
(
1
),
115
124
.
Kang
D.
Yu
X.
Tong
S.
Ge
M.
Zuo
J.
Cao
C.
Song
W.
2013
Performance and mechanism of Mg/Fe layered double hydroxides for fluoride and arsenate removal from aqueous solution
.
Chemical Engineering Journal
228
,
731
740
.
Ma
Y.
Wang
S.-G.
Fan
M.
Gong
W.-X.
Gao
B.-Y.
2009
Characteristics and defluoridation performance of granular activated carbons coated with manganese oxides
.
Journal of Hazardous Materials
168
(
2–3
),
1140
1146
.
Miretzky
P.
Cirelli
A. F.
2011
Fluoride removal from water by chitosan derivatives and composites: a review
.
Journal of Fluorine Chemistry
132
(
4
),
231
240
.
Raychoudhury
T.
Schiperski
F.
Scheytt
T.
2015
Distribution of iron in activated carbon composites: assessment of arsenic removal behavior
.
Water Science and Technology: Water Supply
15
(
5
),
990
998
.
Swain
S. K.
Patnaik
T.
Patnaik
P. C.
Jha
U.
Dey
R. K.
2013
Development of new alginate entrapped Fe(III)-Zr(IV) binary mixed oxide for removal of fluoride from water bodies
.
Chemical Engineering Journal
215–216
,
763
771
.
Velazquez-Jimenez
L. H.
Vences-Alvarez
E.
Flores-Arciniega
J. L.
Flores-Zuñiga
H.
Rangel-Mendez
J. R.
2015
Water defluoridation with special emphasis on adsorbents-containing metal oxides and/or hydroxides: a review
.
Separation and Purification Technology
150
,
292
307
.
Wu
X.
Zhang
Y.
Dou
X.
Zhao
B.
Yang
M.
2007
Fluoride adsorption on an Fe-Al-Ce trimetal hydrous oxide: characterization of adsorption sites and adsorbed fluorine complex species
.
Chemical Engineering Journal
223
,
364
370
.
Yu
Y.
Wang
C.
Guo
X. J.
Chen
P.
2015a
Modification of carbon derived from Sargassum sp. by lanthanum for enhanced adsorption of fluoride
.
Journal of Colloid and Interface Science
441
,
113
120
.
Zhang
S.
Lu
Y.
Lin
X.
Su
X.
Zhang
Y.
2014
Removal of fluoride from groundwater by adsorption onto La(III)- Al(III) loaded scoria adsorbent
.
Applied Surface Science
303
,
1
5
.

Supplementary data