Simultaneous adsorptive removal of fluoride and phosphate by magnesia–pullulan composite from aqueous solution

Abstract

This study is based on the use of MgOP to treat solution containing fluoride and phosphate for adsorption. The characteristics of MgOP before and after adsorption were analyzed by XRD, FTIR, SEM and EDX. The experimental parameters included adsorbent dose, contact time, initial pH of solution, initial concentration of ions and the adding order of ions. The optimized adsorbent dose and contact time are 2 g L⁻¹ and 60 min respectively. It is observed that the removal efficiency of fluoride and phosphate ions is hardly affected at the broad range of pH (2 to 12). The presence of fluoride and phosphate ions affects the adsorption rates of another one and the main existing forms of phosphate in domestic wastewater have no significant effects on the defluoridation capacity of MgOP. Besides, it is seen that the adding order of ions greatly influences the adsorption capacities of MgOP for fluoride and phosphate ions. The Langmuir isotherm and pseudo second-order kinetic models are observed to describe the adsorption processes of fluoride in the single- and multi-adsorbate systems. Intra-particle diffusion model with greater R² is showed to explain the process of the fluoride adsorption. Also, the negative values of ΔG⁰ reveal that the adsorption processes of fluoride and phosphate ions in the single- and multi-adsorbate systems are feasible and spontaneous.

---

1. Introduction

Fluoride plays an indispensable role in human health as a crucial element.^1,2 Its low concentration contributes to children's enamel. However, if people are exposed to an excessive intake of fluoride for a long time, they may suffer from diseases such as dental fluorosis.^3–5 Fluoride in sewage comes from the industries that use fluoride as raw material and some rocks and minerals which desorb fluoride by natural weathering and rain.⁶ Different methods have been used for fluoride removal such as ion exchange, adsorption, precipitation, membrane and electrodialysis.^7–9 Phosphate is essential to the photosynthetic creatures and is used widely in many industries especially in agricultural industries. But the excessive phosphate can cause eutrophication and decrease the amount of aquatic animals by deteriorating the water quality.¹⁰ Dephosphorization can be done by biological treatment, ion exchange resin, crystallization, membrane, chemical precipitation and adsorption.^11–13 However, some methods can be used to remove fluoride and phosphate ions from water simultaneously such as a hybrid precipitation–microfiltration process, RO/NF treatment, ion exchange, and adsorption.^14–17 Among the removal technologies, adsorption is viewed as the one of the most promising methods and the reasons are low cost, and simple design and operation.

In previous study, magnesia–pullulan (MgOP) composite has been used only for the fluoride removal as the adsorbent.¹⁸ The large specific surface area (32.8992 m² g⁻¹) might contribute to the fluoride adsorption. Over 90% of fluoride is adsorbed in the first 30 min by 2 g L⁻¹ of MgOP. The high defluoridation capacity of MgOP was observed at pH range 2 to 12. Besides, it was also reported that the presence of coexisting ions (i.e., Cl⁻, SO₄²⁻ and NO₃⁻) has negligible effects on defluoridation capacity.

The aim of present study is the application of MgOP for both the removal of fluoride and phosphate ions simultaneously, while investigating the parameters that affect the adsorption of fluoride and phosphate ions. Other objective of the study is to draw a conclusion on the process of simultaneous adsorption on the basis of our experiment.

2. Experimental procedure

2.1. Preparation of MgOP

MgOP was synthesized by magnesia and pullulan in 2 : 3 ratios. At first, the chemicals were mixed in adequate deionized water (ddH₂O) and then shaken for 24 h at 25 °C to guarantee its full mixing. Subsequently, the mixture was dried for 12 h in a drying oven at 105 °C to obtain the solid sample with primrose surface. After that, the solid sample was calcined for 2 h in a muffle furnace at 450 °C to get the final solid, namely MgOP composite. At last, MgOP was ground into powder and packaged into plastic bags with detailed label.

2.2. Analytical method

The concentration of fluoride ions was tested by ion selective electrode method with an ion meter (pXS-215, Tianda Instrument Shanghai, Co., Ltd. China). For the concentration of phosphate ions, the ammonium molybdate spectrophotometric method was used (721-spectrophotometer, Yidian Analytical Instrument Shanghai, Co., Ltd. China). The standard solutions of fluoride and phosphate were prepared by dissolving NaF and KH₂PO₄ respectively and initial concentrations of fluoride ions and phosphorus were both 1000 mg L⁻¹. All the solutions of fluoride and phosphate used in the study were diluted by the standard solutions. 0.1 M NaOH and HCl were used to adjust pH of solution. All the chemicals used in the study were of analytical grade.

2.3. Simultaneous adsorption of fluoride and phosphate on MgOP

MgOP was introduced to the test tubes containing fluoride and/or phosphate solutions and then the mixture was shaken in a thermostatic shaker at 150 rpm at 25 °C. After that, the suspensions were centrifuged at 3000 rpm for 30 min and then filtered through a 0.45 μm membrane filter (Tianjin Jinteng Experimental Equipment Co., Ltd. China). The procedure was conducted with the adsorbent dose (0.5–4.0 g L⁻¹), contact time (0–60 min), initial pH values (2–12), initial concentrations of ions (0–1000 mg L⁻¹) and the main forms of phosphate ions existing in the sewage at 25 °C. The adsorption isotherms, kinetics and thermodynamics were also studied.

The effects of the presence of fluoride or phosphate on the adsorption of another one were carried out with different adding orders of each of adsorbate. When the fluoride and phosphate solutions were brought with MgOP simultaneously, the system was named as F–P–MgOP. In another case, fluoride solution was firstly added to MgOP and then the suspension was shaken for 24 h at 25 °C. After that, the phosphate solution was brought and the mixture was shaken for another 24 h. This system was named as (F–MgOP)–P. Similarly, when phosphate solution was added with MgOP first, which was followed by fluoride solution, the system was named as (P–MgOP)–F.

All the studies were repeated several times in the same condition and the standard deviations were also introduced to the study to check the accuracy of the experimental data. The amount of fluoride ions and phosphorus adsorbed on MgOP was calculated by the following formula.

where q_t (mg g⁻¹) is the adsorption capacity of MgOP at time t; C₀ (mg L⁻¹) is the initial concentration of fluoride ions or phosphorus; C_t (mg L⁻¹) is the remaining concentration of fluoride ions or phosphorus at time t; V (L) and W (g) are the volume of the solution and the weight of MgOP respectively. The software Origin 8.0 was used to process experimental data.

2.4. Properties of MgOP

The properties of MgOP before and after the simultaneous adsorption of fluoride and phosphate were analyzed by X-ray diffraction (XRD), Fourier Transform Infra Red (FTIR), Scanning Electron Microscope (SEM) and Energy Dispersive analysis of X-ray (EDX). XRD patterns (PAN analytical B.V X'Pert PRO) were taken with CuKα radiations. FTIR spectrophotometer (Equinox 55 Bruker) was used to obtain the IR spectrum of MgOP. SEM images were obtained by FEI Quanta 200 as well as the analysis of EDX.

3. Results and discussion

The abbreviations of “F” and “P” show fluoride ion and phosphorus respectively. The abbreviations of “F only” and “F (F + P)” indicate the condition of fluoride ion in the single- and multi-adsorbate systems respectively. The patterns are similar for phosphorus.

3.1. Properties of MgOP before and after the simultaneous adsorption of fluoride and phosphate

XRD patterns were used to analyze the phase and crystal texture of MgOP based on the powder diffraction standards. The patterns of MgOP before and after adsorption of fluoride and phosphate are shown in Fig. S1.† It can be seen that more peaks appeared in Fig. S1b as compared to Fig. S1a,† which is ascribed to the adsorption of F, P and water molecules.¹⁹ Fluoride and phosphate ions adsorbed by MgOP appears as MgF₂ (JCPDS 70-2744) and Mg₃(PO₄)₂ (JCPDS 88-0413) respectively, revealing that the adsorption mechanism of F and P is ion exchange. The other peak appears in Fig. S1b† is indexed as Mg(OH)₂ (JCPDS 84-2164) which may come from water molecule.

The SEM images of MgOP before and after simultaneous adsorption of fluoride and phosphate ions are depicted with their corresponding EDX spectra shown in Fig. 1. The SEM image of virgin MgOP shows that flake-shaped morphology of the particles, which is similar from the SEM images of MgOP after adsorption, and its corresponding EDX spectra demonstrate the existence of Mg, O and C, thus confirming the synthesis of MgOP. EDX spectra also indicate the existence of F and P element in MgOP after adsorption, confirming the XRD patterns.

Fig. 1 SEM images and corresponding EDX spectra of (a) MgOP, (b) MgOP after the adsorption of fluoride and phosphate.

FITR spectra were used to investigate the functional groups of MgOP and analyze the mechanism of the simultaneous adsorption. FITR spectra of MgOP before and after adsorption are shown in Fig. 2. It can be seen that the bands at 3430 cm⁻¹ of MgOP after the adsorption, slightly shifts with an increase of intensity as compared to spectra of MgOP before adsorption. This band could be ascribed to the formation of the O–H–F bond, revealing the existence of hydrogen boding.¹⁸ The peaks at 1438 cm⁻¹ are ascribed to O–H & C–O bending vibration bands while the bands at 1631 cm⁻¹ are attributed to C=C with the stretching vibration disturbed by the water molecules. The FITR spectra of MgOP after adsorption of phosphate in the single- and multi-adsorbate systems at 1057 cm⁻¹ & 1054 cm⁻¹, reflect the antisymmetric stretching vibration of PO₄³⁻ and the authenticity of the adsorption of phosphate. The absorbance intensity of peak due to phosphate in the multi-adsorbate system has no significant change compared with that in the single-adsorbate system, indicating that the presence of fluoride has negligible effects on the adsorption capacity of phosphate in the multi-adsorbate system.

Fig. 2 FTIR spectra of (a) MgOP, (b) MgOP after the adsorption of fluoride, (c) MgOP after the adsorption of phosphate, (d) MgOP after the adsorption fluoride and phosphate.

3.2. Effect of adsorbent dose on simultaneous adsorption of fluoride and phosphate

The effect of the adsorbent dose on the simultaneous adsorption of fluoride and/or phosphate at given concentration (10 mg L⁻¹) of F and P, is depicted in Fig. S2.† The results reveal that the optimized doses of MgOP are 2 and 1.2 g L⁻¹ for F and P respectively both in the single- and multi-adsorbate systems. It is observed that the adsorption capacities of MgOP for F and P have no significant changes at the constant adsorbent dose. MgOP may have adequate vacant adsorption sites at these given adsorbent doses and with the increase of the adsorbent dose, the removal efficiencies of F and P increase in the single- and multi-adsorbate systems. It could be seen that high removal efficiency of fluoride ions observed at 2.0 g L⁻¹ of MgOP is 92.5% and 93.5% in the single- and multi-adsorbate systems respectively. As for P, the optimized adsorbent dose is exhibited at 1.2 g L⁻¹ while nearly 99.05% and 98.0% of phosphorus adsorbed in the single- and multi-adsorbate systems respectively. As it has been proved that the removal efficiencies of ions rise with the increase of the adsorbent dose.²⁰ So 2.0 g L⁻¹ of MgOP was used in the subsequent experiment for enhanced removal.

The better affinity of MgOP for phosphate than fluoride may be explained as that monovalent anions can be adsorbed difficultly than multivalent anions.²⁰ However, it was also reported that the affinity order of F > P had been observed while adsorption by LDH.²¹ So the affinity of adsorbent for ions may be related to the structural properties of the adsorbent and its adsorption mechanism of F and P.

3.3. Effect of initial pH on simultaneous adsorption of fluoride and phosphate

The influence of pH on the adsorption of F and/or P was investigated with 2 g L⁻¹ of MgOP and 10 mg L⁻¹ of each of the adsorbate. As shown in Fig. S3,† the defluoridation and dephosphorization capacities of MgOP are negligibly affected by the initial pH with fluctuations of 4.60 and 4.94 mg g⁻¹ respectively in the single-adsorbate system. Also, the adsorption capacities for F and P have negligible changes in the single- and multi-adsorbate systems at given concentration of F and P.

According to surface complexation, the solid surface is considered as polymeric acid which can adsorb the dissociative ions and the hydroxyl groups on the solid surface are amphoteric, which could react with acid and alkaline chemicals.²² MgOP might adsorb H⁺ and OH⁻ from acid and alkaline solutions respectively. The reactions are presented below.

≡SOH + H⁺ ↔ SOH₂⁺ (2)

≡SOH + OH⁻ ↔ SO⁻ + H₂O (3)

The amount of the surface active sites and the surface charges both could influence the adsorption of ions.²³ Though the adsorption of H⁺ and OH⁻ ions could change the surface charge of MgOP, it would not affect the amount of the surface active sites. Also, the surface charges only influence the adsorption rate of ions rather than the equilibrated adsorption capacity. Therefore, different initial pH would not affect the adsorption equilibrium of F and P in the single- and multi-adsorbate systems. Besides, the P adsorption is more evident than F adsorption as shown in Fig. S3† as MgOP may have better affinity for multivalent phosphate anions than monovalent fluoride anions.

3.4. Effect of contact time on simultaneous adsorption of fluoride and phosphate

The F and P adsorption in the single- and multi-adsorbate systems as a function of contact time were investigated with 2 g L⁻¹ of MgOP and 10 mg L⁻¹ of each of adsorbate at pH 7.0. From Fig. 3, the adsorption capacities of MgOP for F and P rise rapidly in the first 0–35 min where about 90% of F and P are adsorbed in the single-adsorbate system. The possible reason is that the vacant adsorption sites are abundant and the concentration gradient of F or P between the adsorbent surface and solution is high at the beginning of the adsorption. However, after 35 min, the adsorption rates decrease gradually due to the decrease of vacant adsorption sites and the concentration gradient. It is also indicated that the adsorption rates of F and P both are affected by the presence of another one. The equilibrated time for the P adsorption is about 5 min in the multi-adsorbate system while it is observed around 50 min in the single-adsorbate system. The presence of fluoride might accelerate the uptake rate of P and the possible reasons are (a) more repulsion between the two negatively charged ions (b) heavy molecular weight and (c) bigger size of phosphate causes it to attach faster than smaller F. In contrast, the presence of P suppresses the adsorption of fluoride slightly with negligible changes in the equilibration time for the F adsorption. As it is observed that the attainment of equilibrium of F and P in adsorption are all less than 60 min in the two systems and thereby the contact time of the adsorption was set as 60 min for the rest of the studies.

Fig. 3 Effect of contact time on simultaneous adsorption of fluoride and phosphate (C_0(F) = C_0(P) = 10 mg L⁻¹, pH 7.0, reaction temperature 298 K).

3.5. Effect of initial concentrations of F and P

The adsorption rates of F and P as a function of initial concentration of each of adsorbate was studied with 2 g L⁻¹ of MgOP and 60 min of contact time at pH 7.0 at 298 K. Fig. 4a demonstrates that the adsorption of fluoride is suppressed at higher concentrations of P and the equilibrated defluoridation capacity is within 4.46 to 4.80 mg g⁻¹ range in the presence of P (0–10 mg L⁻¹) at given initial concentration of F (10 mg L⁻¹). The pseudo second-order kinetics was proved to agree with the kinetic data of the fluoride adsorption in the single-adsorbate system,¹⁸ so the model was used to analyze this kinetic process. The expression of the model is shown below.where q_e (mg g⁻¹) is the equilibrated defluoridation capacity of MgOP; q_t (mg g⁻¹) is the amount of fluoride ions adsorbed on MgOP at time t (min); h (mg g^−1.min) is the initial adsorption rate; k (g mg^−1.min) is the rate constant of the pseudo second-order kinetic model.

Fig. 4 Effect of initial concentrations of (a) phosphorus on fluoride adsorption (C_0(F) = 10 mg L⁻¹) (b) fluoride ions on phosphate adsorption (C_0(P) = 10 mg L⁻¹) (c) nitrate on fluoride adsorption (C_0(F) = 10 mg L⁻¹) (d) chloridion on phosphate adsorption (C_0(P) = 10 mg L⁻¹).

The parameters presented in Table 1 reflect the goodness of pseudo second-order kinetics for the adsorption of fluoride in the presence of P where R² > 0.995. In the value of k, the adsorption rates of fluoride ions with different initial concentrations of P are in this order, 0 mg L⁻¹ (0.042 g mg⁻¹ min⁻¹) > 2 mg L⁻¹ (0.090 g mg⁻¹ min⁻¹) > 4 mg L⁻¹ (0.101 g mg⁻¹ min⁻¹) > 6 mg L⁻¹ (0.031 g mg⁻¹ min⁻¹) > 8 mg L⁻¹ (0.022 g mg⁻¹ min⁻¹) > 10 mg L⁻¹ (0.018 g mg⁻¹ min⁻¹). Thus, lower concentration of P promotes the adsorption of fluoride while its higher concentration has opposite effects in the multi-adsorbate system. However, Fig. 4b presents that the adsorption of P is greatly promoted at the concentration range of F (2–10 mg L⁻¹) without significant changes in the equilibrated dephosphorization capacity of MgOP when the initial concentration of P is set as 10 mg L⁻¹.

Table 1 Pseudo second-order kinetic parameters for the adsorption of fluoride in the multi-adsorbate system

Adsorption kinetic models	Kinetic parameters	The initial concentration of P (mg L^−1)
		0	2	4	6	8	10
Pseudo second-order kinetics	qe (mg g^−1)	5.056	4.728	4.810	5.133	5.451	5.611
	h (mg g^−1 min^−1)	1.080	2.018	2.328	0.823	0.658	0.570
	k (g mg^−1 min^−1)	0.042	0.090	0.101	0.031	0.022	0.018
	R^2	0.998	1.000	1.000	0.999	0.995	0.995

---

3.6. Effect of other ions on the adsorption of fluoride and phosphate

To study the phenomenon mentioned, different concentrations of NaNO₃ and NaCl were added to the single-adsorbate system respectively with 2 g L⁻¹ of MgOP and 10 mg L⁻¹ of F or P at pH 7.0 at 298 K. As a matter of fact, N and P are in the same group and have similar properties and same as with Cl and F. Fig. 4c shows the fluoride adsorption is promoted and obtain its equilibrium in the first 5 min in the presence of NO₃⁻ (2–10 mg L⁻¹). Similarly, the equilibrated time of P adsorption is reduced to be less than 10 min in the presence of Cl⁻ (2–10 mg L⁻¹) as shown in Fig. 4d. Also, the presence of NO₃⁻ and Cl⁻ would not influence the equilibrated adsorption capacities of F and P.

As mentioned above, the vacant adsorption sites and the concentration gradient of the adsorbate between the adsorbent surface and the solution could affect the adsorption rate of the adsorbate directly. It shows that the presence of NO₃⁻ and Cl⁻ can increase the vacant adsorption sites. Likewise, the presence of F and P might also increase the vacant adsorption sites in the multi-adsorbate system.

When different initial concentrations of P are added to the fluoride solution, this might increase the vacant adsorption sites and cause the competition between F and P for the adsorption sites. As it has been illustrated previously that MgOP has comparatively better affinity for phosphate than fluoride ions, phosphate ions would be more easily and quickly adsorbed. At low concentration of phosphate, the fluoride adsorption might not compete with the phosphate adsorption as the equilibrated time of the phosphate adsorption is quite short and thereby the competition has less opportunity to occur. In this case, the adsorption rate of fluoride would increase due to the added vacant adsorption sites derived from the addition of P. In contrast, the presence of high concentration of phosphate ions might decrease the uptake rate of fluoride on MgOP. In such situation, MgOP needs some time to adsorb most of P so P might compete with F for the vacant adsorption sites.

However, the adsorption rate of P is accelerated at broad concentration range of F (2–10 mg L⁻¹) because the addition of F increases the vacant adsorption sites and phosphate ions are more easily adsorbed by MgOP as indicated above.

3.7. Effect of the existing forms of phosphate on simultaneous adsorption of fluoride and phosphate

Normally, the total concentration of phosphate is about 10 mg L⁻¹ in the domestic wastewater,^24,25 mainly consisting of 5 mg L⁻¹ of orthophosphate, 3 mg L⁻¹ of pyrophosphate and 1 mg L⁻¹ of polyphosphates. The effects of main forms of phosphate on the fluoride adsorption were explored with 2 g L⁻¹ of MgOP and 60 min of contact time. As shown in Fig. S4,† the removal efficiencies of fluoride ions are more than 90.20% at the given concentration of F (10 mg L⁻¹) no matter which phosphate is brought to the solution. Besides, MgOP can adsorb over 96.8% of P from the main forms of phosphate.

3.8. Adsorption isotherms

The adsorption isotherms of F and P in the single- and multi-adsorbate systems were studied with 2 g L⁻¹ of MgOP and 60 min of contact time at pH 7.0. Langmuir and Freundlich isotherms were used to analyze the adsorption isotherms of F and P in the single- and multi-adsorbate systems at the different temperatures (i.e., 298, 308 and 318 K). According to Langmuir isotherm, the adsorption process is monolayer, and the active sites have uniform energy and exist on the surface of the absorbent with the constant adsorption energy.^11,26 As far as Freundlich isotherm is concerned, it states that the adsorbent surface energy is heterogeneous and more occupations of the active sites cause the decrease of the other sites' binding strength with adsorbate, and the sites with stronger binding ability are more easily occupied.^11,27 The equations of models are presented below.where C_e (mg L⁻¹) is the equilibrated concentration of F or P; q_e (mg g⁻¹) is the equilibrated defluoridation and dephosphorization capacities respectively; Q₀ (mg g⁻¹) is the theoretical monolayer capacity calculated by Langmuir isotherm; b (L mg⁻¹) and k_F (L g⁻¹) are the constants of Langmuir and Freundlich isotherms respectively; n is a dimensionless constant related to the heterogeneity factor.

Apart from R², chi-square (χ²) was also used to assess the goodness-of-fit of the model for the data. The equation of χ² is presented below.

where q_e,m and q_e (mg g⁻¹) are the equilibrated adsorption capacity of MgOP, which are deduced by the model and experiment respectively. Lower χ² reflect the better applicability of the adsorption isotherm model for data.

The calculated adsorption isotherms parameters are presented in Table 2, demonstrating that Langmuir isotherm agrees with the adsorption isotherm data better than Freundlich isotherm both in the single- and multi-adsorbate systems. The possible reason may be that R² for Langmuir isotherm are closer to 1 and χ² for Langmuir isotherm are smaller compared with those for Freundlich isotherm. Based on the assumed theory of Langmuir isotherm, the adsorption processes of F and P in the single- and multi-adsorbate systems are controlled by monolayer adsorption. This may be ascribed to the uniform energy of sites on the surface of MgOP. The increase of the temperature causes the decrease of the theoretical monolayer capacity because the activity of adsorption sites could be weakened at higher temperature. Besides, k_F rise with the increase of temperature, confirming that the adsorption of F and P in the two systems are endothermic.

Table 2 Adsorption isotherm parameters for the adsorption of F and P in the single- and multi-adsorbate systems

Systems	Temperature (K)	Langmuir isotherm				Freundlich isotherm
		Q^0 (mg g^−1)	b (L mg^−1)	R^2	χ^2	kF (L g^−1)	n	R^2	χ^2
F (F + P)	298	15.36	0.8029	0.999	0.005	6.569	2.052	0.993	0.041
	308	16.32	0.8274	0.999	0.010	7.063	1.928	0.988	0.086
	318	15.87	0.9334	0.998	0.013	7.295	1.989	0.984	0.111
	298	14.16	5.9435	0.993	0.364	13.196	2.689	0.886	0.788
P (F + P)	308	14.39	7.5063	0.994	0.333	14.811	2.629	0.898	0.748
	318	14.06	10.5441	0.994	0.464	15.480	2.867	0.890	0.846
	298	15.82	0.8863	0.994	0.050	7.090	1.973	0.970	0.189
F only	308	16.06	0.9261	0.995	0.049	7.361	1.961	0.971	0.180
	318	15.66	1.0541	0.993	0.082	7.645	2.051	0.964	0.204
	298	14.63	6.9946	0.991	0.421	14.983	2.560	0.882	0.827
P only	308	15.24	7.4060	0.991	0.317	16.940	2.354	0.913	0.673
	318	13.91	13.6051	0.995	0.498	16.159	3.004	0.889	0.865

---

R_L is used to determine the features of the Langmuir isotherm model.²⁶ R_L reflects different conditions of the adsorption process, i.e., unfavorable (R_L > 1), linear (R_L = 1), favorable (0 < R_L < 1) and irreversible (R_L = 0).²⁸ The formula of R_L is shown below.

where C₀ (mg L⁻¹) is the initial concentration of F and P.

According to eqn (8), R_L range from 0.04 to 0.12, indicating that the adsorption of F and P in the single- and multi-adsorbate systems are favorable. Only the plots of Langmuir isotherms for the adsorption of F and P in the two systems are depicted in Fig. 5.

Fig. 5 The plots of Langmuir isotherm for the adsorption of F and P in the single- and multi-adsorbate systems.

3.9. Adsorption thermodynamics

Adsorption thermodynamic parameters were used to assess the characteristics of the adsorption of F and P in the single- and multi-adsorbate systems, including standard free energy (ΔG⁰), standard enthalpy change (ΔH⁰) and standard entropy change (ΔS⁰). The parameters are presented in the following equations.²⁹

ΔG⁰ = −RT ln k_F (9)

ΔG⁰ = ΔH⁰ − TΔS⁰ (10)

ΔH⁰ and ΔS⁰ can be inferred by the equation below.

where R (J mol⁻¹ K⁻¹) is the gas constant, at 8.314; T (K) is the absolute temperature.

The thermodynamic parameters are summarized in Table 3. The positive values of ΔH⁰ show that the adsorption processes of F and P on MgOP in the two systems are endothermic, confirming the findings of the isotherm experiment. The positive values of ΔS⁰ indicate that the adsorption of F and P on MgOP cause the displacement of water molecules at the solid–liquid interface. The negative values of ΔG⁰ illustrate that the defluoridation and dephosphorization in the systems are feasible and unsolicited. The similar results are reported in other literature.^30–32

Table 3 Thermodynamic parameters for the adsorption of F and P in the single- and multi-adsorbate systems

System	ΔH^0 (kJ mol^−1)	ΔS^0 (J mol^−1 K^−1)	ΔG^0 (kJ mol^−1)
			298 K	308 K	318 K
F (F + P)	1.72	12.58	−2.03	−2.17	−2.27
P (F + P)	0.41	8.49	−2.78	−3.00	−3.15
F only	0.40	8.39	−2.10	−2.22	−2.34
P only	1.29	11.40	−2.91	−3.15	−3.19

---

3.10. Adsorption kinetics

As the adsorption of P in the multi-adsorbate system obtains its equilibrium in the first 5 min and the fluoride adsorption process in the single-adsorbate system has been studied in the previous paper,¹⁸ the present paper would only study the adsorption kinetics of fluoride in the multi-adsorbate system. The kinetic experiment was conducted with 2 g L⁻¹ of MgOP, 60 min of contact time, different F concentrations (10, 15, 20 and 25 mg L⁻¹) and fixed initial concentration of P (10 mg L⁻¹) at pH 7.0 at temperatures (298, 308, 318 K). Fig. 6 demonstrates the plots of contact time (t) against the remaining concentration of fluoride ions (C_t) at given concentration of P (10 mg L⁻¹) at 298 K, indicating the remaining concentration of fluoride ions decreases with contact time.

Fig. 6 The adsorption kinetics of fluoride adsorption in the multi-adsorbate system (C_0(P) = 10 mg L⁻¹; temperature: 298 K; pH, 7.0).

Reaction-based and diffusion-based models³³ were applied to explore the adsorption kinetics of F in the presence of P. The pseudo first-order and second-order kinetic models are the reaction-based models, which are used to predict the solid–liquid adsorption process. The expressions of the models are presented in the following equations.^34,35

where q_e (mg g⁻¹) is the equilibrated defluoridation capacity; q_t (mg g⁻¹) is the defluoridation capacity of MgOP at time t; t (min) is contact time; h (mg g⁻¹ min⁻¹) is the constant of the initial adsorption rate; k_ad (min⁻¹) and k (g mg⁻¹ min⁻¹) are the adsorption rate constants of the pseudo first-order and second-order kinetic models respectively.

The parameters of the reaction-based models are shown in Table 4, demonstrating that pseudo second-order kinetics could stimulate the adsorption process of F in the multi-adsorbate system due to greater R². The adsorption rate increases with temperature and thereby the adsorption process of fluoride is endothermic in the presence of P. Only the plots of pseudo second-order kinetics for the fluoride adsorption in the multi-adsorbate system are depicted in Fig. 7a.

Table 4 The parameters of the pseudo first-order and second-order kinetic models for the fluoride adsorption in the multi-adsorbate system

Adsorption kinetic models	Kinetic parameters	10 mg L^−1			15 mg L^−1
		298 K	308 K	318 K	298 K	308 K	318 K
Pseudo first-order kinetics	kad (min^−1)	0.087	0.072	0.049	0.088	0.055	0.062
	R^2	0.873	0.834	0.671	0.864	0.608	0.768
Pseudo second-order kinetics	k (g mg^−1 min^−1)	0.031	0.037	0.040	0.013	0.016	0.016
	h (mg g^−1 min^−1)	0.798	0.948	1.023	0.837	1.007	0.973
	qe (mg g^−1)	5.101	5.071	5.061	7.961	7.834	7.911
	R^2	0.998	0.997	0.995	0.991	0.992	0.992

---

Adsorption kinetic models	Kinetic parameters	20 mg L^−1			25 mg L^−1
		298 K	308 K	318 K	298 K	308 K	318 K
Pseudo first-order kinetics	kad (min^−1)	0.097	0.069	0.080	0.082	0.078	0.073
	R^2	0.952	0.719	0.983	0.842	0.835	0.852
Pseudo second-order kinetics	k (g mg^−1 min^−1)	0.007	0.008	0.008	0.005	0.006	0.009
	h (mg g^−1 min^−1)	0.781	0.886	0.938	0.902	1.065	1.340
	qe (mg g^−1)	10.602	10.623	10.602	12.936	12.864	12.527
	R^2	0.990	0.990	0.991	0.989	0.989	0.989

---

Fig. 7 (a) The plots of pseudo second-order for fluoride adsorption on MgOP in the multi-adsorbate system at 298 K (b) the plots of intra-particle diffusion for fluoride adsorption on MgOP in the multi-adsorbate system at 298 K.

The particle diffusion and intra-particle diffusion models were applied to explain solute transfer as the diffusion-based models.³⁶ The models are presented in the following equations.

Intra-particle diffusion, q_t = k_it^0.5 (15)

where C_e (mg L⁻¹) is the equilibrated concentration of F; C_t (mg L⁻¹) is the remaining concentration of F at time t; t (min) is contact time; K_P (min⁻¹) and k_i (mg g⁻¹ min^−0.5) are the adsorption rates of the particle diffusion and intra-particle diffusion models respectively.

The calculated parameters are shown in Table 5, revealing that the goodness-of-fit of intra-particle diffusion is better for the adsorption kinetic data. Only the plots of the intra-particle diffusion model for the fluoride adsorption in the multi-adsorbate system are shown in Fig. 7b, indicating that the adsorption process includes two steps. Firstly, the sharp slope of the curve illustrates that fluoride is adsorbed to the external surface of MgOP quickly between 0 and 35 min. Subsequently, the adsorption rate declines with contact time until equilibrium, which is attributed to the decrease of the vacant adsorption sites and the concentration gradient of F between the MgOP surface and solution.

Table 5 The parameters of particle diffusion and intra-particle diffusion for the fluoride adsorption in the multi-adsorbate system

Adsorption kinetic models	Kinetic parameters	10 mg L^−1			15 mg L^−1
		298 K	308 K	318 K	298 K	308 K	318 K
Particle diffusion	KP (min^−1)	0.0056	0.0053	0.0053	0.0076	0.0068	0.0070
	R^2	0.6642	0.5925	0.5098	0.5973	0.5663	0.5810
Intra-particle diffusion	ki (mg g^−1 min^−0.5)	0.2955	0.2767	0.2591	0.5186	0.4789	0.4925
	R^2	0.8144	0.7575	0.6814	0.7781	0.7476	0.7582

---

Adsorption kinetic models	Kinetic parameters	20 mg L^−1			25 mg L^−1
		298 K	308 K	318 K	298 K	308 K	318 K
Particle diffusion	KP (min^−1)	0.0089	0.0085	0.0081	0.0092	0.0085	0.0075
	R^2	0.6785	0.6494	0.6333	0.6823	0.6318	0.5727
Intra-particle diffusion	ki (mg g^−1 min^−0.5)	0.7677	0.7462	0.7218	0.9461	0.8930	0.7947
	R^2	0.8374	0.8217	0.8213	0.8458	0.8132	0.7658

---

3.11. Effect of adding order of fluoride and phosphate on adsorption of another one

The effects of the adding order of F and P on the adsorption of another one were studied with 2 g L⁻¹ of MgOP and 24 h of contact time at pH 7.0 at 298 K. Fig. 8a presents the fluoride adsorption with different adding orders of adsorbate at different initial concentration of F (100–1000 mg L⁻¹). The more suppression of the fluoride adsorption is observed at fixed initial concentration of P (1000 mg L⁻¹) when the phosphate solution is added first. The similar phenomenon is observed in the phosphate adsorption in the presence of F (1000 mg L⁻¹) and P (100–1000 mg L⁻¹) as shown in Fig. 8b.

Fig. 8 Effect of the adding order on the (a) fluoride adsorption on MgOP (C_0(P) = 1000 mg L⁻¹; temperature 298 K; pH 7.0) (b) phosphate adsorption on MgOP (C_0(F) = 1000 mg L⁻¹; temperature 298 K; pH 7.0).

In the simultaneous adsorption of F and P on MgOP, fluoride and phosphate ions compete for the same vacant adsorption sites on the surface of MgOP. So no matter which ions are added to the solution first, they would occupy the adsorption sites. As the vacant adsorption sites are limited, the ion exchange may exist between fluoride and phosphate ions. But it is difficult for ions to be displaced after the one is adsorbed by MgOP. In the multi-adsorbate system, P seems to be more difficult to be displaced than F as MgOP has better affinity for phosphate ions. So when only fluoride ions are added with MgOP first, they are adsorbed by MgOP and then the fluoride adsorption reaches the equilibrium. After the addition of phosphate ions, a part of phosphate ions has ion exchange with fluoride ions adsorbed and thereby the amount of fluoride ions adsorbed would decrease. Similarly, when only phosphate ions are brought with MgOP first, the addition of fluoride ions would cause the decrease of the phosphate adsorption. In another case, when F and P are added simultaneously with MgOP, F and P would be adsorbed by MgOP and compete for the adsorption sites at the same moment. As MgOP adsorbs phosphate more easily than fluoride ions, the former ions may obtain the equilibrium earlier. Some free fluoride ions would have ion exchange with partial phosphate ions which have been adsorbed by MgOP. The findings also explain why F–P–MgOP system is somewhat similar to (P–MgOP)–F system. The adsorption capacities of MgOP for F and P are higher in the single-adsorbate system compared with other system, which also verifies the competition between F and P on MgOP for the adsorption sites.

4. Conclusions

MgOP was used to remove fluoride and phosphate simultaneously from water. In the multi-adsorbate system for the adsorption of fluoride and phosphate, the optimized adsorbent dose is 2 g L⁻¹ and 60 min is shown as equilibrated time, while the adsorption capacity of MgOP for fluoride and phosphate ions presents negligible effect at broader pH range (2 to 12). In the multi-adsorbate system, the adsorption rates of fluoride and phosphate ions are affected by the presence of another one. Besides, the main forms of phosphate existing in sewage can hardly influence the defluoridation capacity of MgOP. In the single- and multi-adsorbate systems, the data of F and P are well fitted by Langmuir isotherm and the adsorption process of fluoride in the presence of P follows the pseudo second-order kinetic and the intra-particle diffusion models. On the basis of our findings, it can be concluded that the mechanism of fluoride and phosphate adsorption involves ion exchange and hydrogen bonding. The values of ΔH⁰ and ΔG⁰ present that the adsorption of F and P are endothermic, practicable and spontaneous in nature. The size of the ion does matter the adsorption rate hence phosphate adsorbed faster than fluoride ion in the multi-adsorbate system.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (21177045).

References

1. A. D. Atasoy and M. O. Sahin, CLEAN – Soil Air Water, 2014, 42, 415–420.
2. M. R. Boldaji, A. H. Mahvi, S. Dobaradaran and S. S. H. B. Sc, Int. J. Environ. Sci. Technol., 2009, 6, 629–632.
3. S. M. Maliyekkal, A. K. Sharma and L. Philip, Water Res., 2006, 40, 3497–3506.
4. E. Kumar, A. Bhatnagar, M. Ji, W. Jung, S. H. Lee, S. J. Kim, G. Lee, H. Song, J. Y. Choi and J. S. Yang, Water Res., 2009, 43, 490–498.
5. V. J. Litza Halla, R. H. Hurt, M. Juan and R. M. Jose Rene, Environ. Sci. Technol., 2014, 48, 1166–1174.
6. P. Loganathan, S. Vigneswaran, J. Kandasamy and R. Naidu, J. Hazard. Mater., 2013, 248–249, 1–19.
7. A. Meenakshi and R. C. Maheshwari, J. Hazard. Mater., 2006, 137, 456–463.
8. N. Drouiche, S. Aoudj, H. Lounici, H. Mahmoudi, N. Ghaffour and M. F. A. Goosen, Desalin. Water Treat., 2011, 29, 96–102.
9. N. Kabay, Ö. Arar, S. Samatya, Ü. Yüksel and M. Yüksel, J. Hazard. Mater., 2008, 153, 107–113.
10. A. C. Edwards and P. J. A. Withers, Soil Use Manage., 2007, 23, 133–143.
11. H. Chen and K. Vaughan, Wetlands, 2013, 34, 325–334.
12. K. Yang, L. G. Yan, Y. M. Yang, S. J. Yu, R. R. Shan, H. Q. Yu, B. C. Zhu and B. Du, Sep. Purif. Technol., 2014, 124, 36–42.
13. J. Cheng, L. Jia, Z. Bo, Y. He and G. Kirumba, J. Environ. Sci., 2014, 26, 466–477.
14. D. Dolar, K. Košutić and B. Vučić, Desalination, 2011, 265, 237–241.
15. N. C. Lu and J. C. Liu, Sep. Purif. Technol., 2010, 74, 329–335.
16. A. Eskandarpour, K. Sassa, Y. Bando, H. Ikuta, K. Iwai, M. Okido and S. Asai, Transactions of the Iron & Steel Institute of Japan, 2007, 47, 558–562.
17. C. Peng, Z. Hong, W. Chong, H. Ma, J. Hu, Y. Pu and L. Peng, J. Hazard. Mater., 2012, 213–214, 100–108.
18. J. Kang, L. Bo, S. Jing, D. Li, Y. Jing, Z. Wei and D. Liu, Chem. Eng. J., 2011, 166, 765–771.
19. Z. Hussain, D. Li, X. Li and J. Kang, RSC Adv., 2015, 5, 43906–43916.
20. L. Liang, H. Jing, W. Min, D. G. Evans and D. Xue, J. Hazard. Mater., 2006, 133, 119–128.
21. M. A. Ulibarri and M. Hermosin, Layered Double Hydroxides: present and future, 2001, pp. 251–284.
22. P. W. Schindler and H. Gamsjäger, Kolloid-Zeitschrift und Zeitschrift für Polymere, 1972, 250, 759–763.
23. J. S. Geelhoed, T. Hiemstra and W. H. V. Riemsdijk, Environ. Sci. Technol., 1998, 32, 2119–2123.
24. H. Roques, Tec & doc Lavoisier d, Paris, 1990.
25. M. A. Fulazzaky, M. H. Khamidun, M. F. M. Din and A. R. M. Yusoff, Chem. Eng. J., 2014, 258, 10–17.
26. T. W. Webi, R. K. Chakravort, T. W. Webi and R. K. Chakravort, AIChE J., 1974, 20, 228–238.
27. R. I. Yousef, B. El-Eswed and A. A. H. Al-Muhtaseb, Chem. Eng. J., 2011, 171, 1143–1149.
28. K. R. Hall, L. C. Eagleton, A. Acrivos and T. Vermeulen, Ind. Eng. Chem. Fundam., 2002, 5, 212–223.
29. Y. Shan and H. Guo, Chem. Eng. J., 2013, 223, 183–191.
30. A. Ghosh, S. Chakrabarti, K. Biswas and U. C. Ghosh, Appl. Surf. Sci., 2014, 307, 665–676.
31. M. N. Sepehr, V. Sivasankar, M. Zarrabi and M. S. Kumar, Chem. Eng. J., 2013, 228, 192–204.
32. L. A. Ramírez-Montoya, V. Hernández-Montoya, A. Bonilla-Petriciolet, M. A. Montes-Morán, R. Tovar-Gómez and M. R. Moreno-Virgen, J. Anal. Appl. Pyrolysis, 2014, 105, 75–82.
33. Y. S. Ho, J. C. Y. Ng and G. Mckay, Sep. Purif. Rev., 2011, 29, 189–232.
34. W. Xiang, G. Zhang, Y. Zhang, D. Tang and J. Wang, Chem. Eng. J., 2014, 250, 423–430.
35. S. M. Maliyekkal, K. R. Antony and T. Pradeep, Sci. Total Environ., 2010, 408, 2273–2282.
36. D. Wankasi, M. H. Jnr and A. I. Spiff, J. Chil. Chem. Soc., 2005, 50, 691–696.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra07175a

---